[webkit-changes] [107476] trunk/LayoutTests

2012-02-11 Thread ossy
Title: [107476] trunk/LayoutTests








Revision 107476
Author o...@webkit.org
Date 2012-02-11 00:56:19 -0800 (Sat, 11 Feb 2012)


Log Message
[Qt] Unreviewed weekend gardening.

* platform/qt/Skipped: Skip tests because ENABLE(MUTATION_OBSERVERS) is disabled.

Modified Paths

trunk/LayoutTests/ChangeLog
trunk/LayoutTests/platform/qt/Skipped




Diff

Modified: trunk/LayoutTests/ChangeLog (107475 => 107476)

--- trunk/LayoutTests/ChangeLog	2012-02-11 08:33:57 UTC (rev 107475)
+++ trunk/LayoutTests/ChangeLog	2012-02-11 08:56:19 UTC (rev 107476)
@@ -1,3 +1,9 @@
+2012-02-11  Csaba Osztrogonác  o...@webkit.org
+
+[Qt] Unreviewed weekend gardening.
+
+* platform/qt/Skipped: Skip tests because ENABLE(MUTATION_OBSERVERS) is disabled.
+
 2012-02-10  Edward O'Connor  eocon...@apple.com
 
 Change values for WEBKIT_KEYFRAMES_RULE, WEBKIT_KEYFRAME_RULE


Modified: trunk/LayoutTests/platform/qt/Skipped (107475 => 107476)

--- trunk/LayoutTests/platform/qt/Skipped	2012-02-11 08:33:57 UTC (rev 107475)
+++ trunk/LayoutTests/platform/qt/Skipped	2012-02-11 08:56:19 UTC (rev 107476)
@@ -230,6 +230,17 @@
 fast/dom/shadow/content-element-api.html
 fast/dom/shadow/content-element-outside-shadow.html
 
+# ENABLE(MUTATION_OBSERVERS) is disabled
+fast/mutation/cross-document.html
+fast/mutation/database-callback-delivery.html
+fast/mutation/inline-event-listener.html
+fast/mutation/mutation-observer-constructor.html
+fast/mutation/observe-attributes.html
+fast/mutation/observe-characterdata.html
+fast/mutation/observe-childList.html
+fast/mutation/observe-exceptions.html
+fast/mutation/observe-subtree.html
+
 # === #
 #   Feature not yet supported.#
 # === #






___
webkit-changes mailing list
webkit-changes@lists.webkit.org
http://lists.webkit.org/mailman/listinfo.cgi/webkit-changes


[webkit-changes] [107477] trunk/Source/WebCore

2012-02-11 Thread kling
Title: [107477] trunk/Source/WebCore








Revision 107477
Author kl...@webkit.org
Date 2012-02-11 06:50:28 -0800 (Sat, 11 Feb 2012)


Log Message
Use Element's hasName/hasID flags to avoid unnecessary work when looking up name/id attributes.
http://webkit.org/b/77845

Reviewed by Anders Carlsson.

Have Element::getIdAttribute() check the hasID() flag before looking up the attribute.
Add an Element::getNameAttribute() to do the same thing with hasName().
Update call sites to make use of these helpers whenever possible.

* accessibility/AccessibilityRenderObject.cpp:
(WebCore::AccessibilityRenderObject::accessibilityDescription):
* dom/DocumentOrderedMap.cpp:
(WebCore::keyMatchesId):
* dom/Element.h:
(Element):
(WebCore::Element::getIdAttribute):
(WebCore):
(WebCore::Element::getNameAttribute):
* dom/NameNodeList.cpp:
(WebCore::NameNodeList::nodeMatches):
* dom/StaticHashSetNodeList.cpp:
(WebCore::StaticHashSetNodeList::itemWithName):
* dom/StaticNodeList.cpp:
(WebCore::StaticNodeList::itemWithName):
* html/HTMLAnchorElement.cpp:
(WebCore::HTMLAnchorElement::name):
* html/HTMLAppletElement.cpp:
(WebCore::HTMLAppletElement::createRenderer):
* html/HTMLCollection.cpp:
(WebCore::HTMLCollection::checkForNameMatch):
(WebCore::HTMLCollection::updateNameCache):
* html/HTMLEmbedElement.cpp:
(WebCore::HTMLEmbedElement::updateWidget):
* html/HTMLFormCollection.cpp:
(WebCore::HTMLFormCollection::updateNameCache):
* html/HTMLFormControlElement.cpp:
(WebCore::HTMLFormControlElement::formControlName):
* html/HTMLFormElement.cpp:
(WebCore::HTMLFormElement::name):
* html/HTMLFrameElementBase.cpp:
(WebCore::HTMLFrameElementBase::setNameAndOpenURL):
* html/HTMLMetaElement.cpp:
(WebCore::HTMLMetaElement::name):
* html/HTMLNameCollection.cpp:
(WebCore::HTMLNameCollection::itemAfter):
* html/HTMLObjectElement.cpp:
(WebCore::HTMLObjectElement::updateWidget):
(WebCore::HTMLObjectElement::updateDocNamedItem):
(WebCore::HTMLObjectElement::containsJavaApplet):
(WebCore::HTMLObjectElement::formControlName):
* inspector/InspectorPageAgent.cpp:
(WebCore::InspectorPageAgent::buildObjectForFrame):
* page/Frame.cpp:
(WebCore::Frame::matchLabelsAgainstElement):
* rendering/svg/RenderSVGResourceContainer.cpp:
(WebCore::RenderSVGResourceContainer::RenderSVGResourceContainer):
* svg/SVGSVGElement.cpp:
(WebCore::SVGSVGElement::getElementById):

Modified Paths

trunk/Source/WebCore/ChangeLog
trunk/Source/WebCore/accessibility/AccessibilityRenderObject.cpp
trunk/Source/WebCore/dom/DocumentOrderedMap.cpp
trunk/Source/WebCore/dom/Element.h
trunk/Source/WebCore/dom/NameNodeList.cpp
trunk/Source/WebCore/dom/StaticHashSetNodeList.cpp
trunk/Source/WebCore/dom/StaticNodeList.cpp
trunk/Source/WebCore/html/HTMLAnchorElement.cpp
trunk/Source/WebCore/html/HTMLAppletElement.cpp
trunk/Source/WebCore/html/HTMLCollection.cpp
trunk/Source/WebCore/html/HTMLEmbedElement.cpp
trunk/Source/WebCore/html/HTMLFormCollection.cpp
trunk/Source/WebCore/html/HTMLFormControlElement.cpp
trunk/Source/WebCore/html/HTMLFormElement.cpp
trunk/Source/WebCore/html/HTMLFrameElementBase.cpp
trunk/Source/WebCore/html/HTMLMetaElement.cpp
trunk/Source/WebCore/html/HTMLNameCollection.cpp
trunk/Source/WebCore/html/HTMLObjectElement.cpp
trunk/Source/WebCore/inspector/InspectorPageAgent.cpp
trunk/Source/WebCore/page/Frame.cpp
trunk/Source/WebCore/rendering/svg/RenderSVGResourceContainer.cpp
trunk/Source/WebCore/svg/SVGSVGElement.cpp




Diff

Modified: trunk/Source/WebCore/ChangeLog (107476 => 107477)

--- trunk/Source/WebCore/ChangeLog	2012-02-11 08:56:19 UTC (rev 107476)
+++ trunk/Source/WebCore/ChangeLog	2012-02-11 14:50:28 UTC (rev 107477)
@@ -1,3 +1,64 @@
+2012-02-11  Andreas Kling  awesomekl...@apple.com
+
+Use Element's hasName/hasID flags to avoid unnecessary work when looking up name/id attributes.
+http://webkit.org/b/77845
+
+Reviewed by Anders Carlsson.
+
+Have Element::getIdAttribute() check the hasID() flag before looking up the attribute.
+Add an Element::getNameAttribute() to do the same thing with hasName().
+Update call sites to make use of these helpers whenever possible.
+
+* accessibility/AccessibilityRenderObject.cpp:
+(WebCore::AccessibilityRenderObject::accessibilityDescription):
+* dom/DocumentOrderedMap.cpp:
+(WebCore::keyMatchesId):
+* dom/Element.h:
+(Element):
+(WebCore::Element::getIdAttribute):
+(WebCore):
+(WebCore::Element::getNameAttribute):
+* dom/NameNodeList.cpp:
+(WebCore::NameNodeList::nodeMatches):
+* dom/StaticHashSetNodeList.cpp:
+(WebCore::StaticHashSetNodeList::itemWithName):
+* dom/StaticNodeList.cpp:
+(WebCore::StaticNodeList::itemWithName):
+* html/HTMLAnchorElement.cpp:
+(WebCore::HTMLAnchorElement::name):
+* html/HTMLAppletElement.cpp:
+(WebCore::HTMLAppletElement::createRenderer):
+* html/HTMLCollection.cpp:
+

[webkit-changes] [107478] trunk/Source/WebCore

2012-02-11 Thread kling
Title: [107478] trunk/Source/WebCore








Revision 107478
Author kl...@webkit.org
Date 2012-02-11 06:54:38 -0800 (Sat, 11 Feb 2012)


Log Message
HTMLTablePartElement: Add helper method to find parent table.
http://webkit.org/b/78413

Reviewed by Anders Carlsson.

Add HTMLTablePartElement::findParentTable() and use that in subclasses instead
of duplicating the code.

* html/HTMLTableCellElement.cpp:
(WebCore::HTMLTableCellElement::additionalAttributeStyle):
* html/HTMLTableColElement.cpp:
(WebCore::HTMLTableColElement::additionalAttributeStyle):
* html/HTMLTablePartElement.cpp:
(WebCore::HTMLTablePartElement::findParentTable):
(WebCore):
* html/HTMLTablePartElement.h:
(WebCore):
(HTMLTablePartElement):
* html/HTMLTableSectionElement.cpp:
(WebCore::HTMLTableSectionElement::additionalAttributeStyle):

Modified Paths

trunk/Source/WebCore/ChangeLog
trunk/Source/WebCore/html/HTMLTableCellElement.cpp
trunk/Source/WebCore/html/HTMLTableColElement.cpp
trunk/Source/WebCore/html/HTMLTablePartElement.cpp
trunk/Source/WebCore/html/HTMLTablePartElement.h
trunk/Source/WebCore/html/HTMLTableSectionElement.cpp




Diff

Modified: trunk/Source/WebCore/ChangeLog (107477 => 107478)

--- trunk/Source/WebCore/ChangeLog	2012-02-11 14:50:28 UTC (rev 107477)
+++ trunk/Source/WebCore/ChangeLog	2012-02-11 14:54:38 UTC (rev 107478)
@@ -1,5 +1,28 @@
 2012-02-11  Andreas Kling  awesomekl...@apple.com
 
+HTMLTablePartElement: Add helper method to find parent table.
+http://webkit.org/b/78413
+
+Reviewed by Anders Carlsson.
+
+Add HTMLTablePartElement::findParentTable() and use that in subclasses instead
+of duplicating the code.
+
+* html/HTMLTableCellElement.cpp:
+(WebCore::HTMLTableCellElement::additionalAttributeStyle):
+* html/HTMLTableColElement.cpp:
+(WebCore::HTMLTableColElement::additionalAttributeStyle):
+* html/HTMLTablePartElement.cpp:
+(WebCore::HTMLTablePartElement::findParentTable):
+(WebCore):
+* html/HTMLTablePartElement.h:
+(WebCore):
+(HTMLTablePartElement):
+* html/HTMLTableSectionElement.cpp:
+(WebCore::HTMLTableSectionElement::additionalAttributeStyle):
+
+2012-02-11  Andreas Kling  awesomekl...@apple.com
+
 Use Element's hasName/hasID flags to avoid unnecessary work when looking up name/id attributes.
 http://webkit.org/b/77845
 


Modified: trunk/Source/WebCore/html/HTMLTableCellElement.cpp (107477 => 107478)

--- trunk/Source/WebCore/html/HTMLTableCellElement.cpp	2012-02-11 14:50:28 UTC (rev 107477)
+++ trunk/Source/WebCore/html/HTMLTableCellElement.cpp	2012-02-11 14:54:38 UTC (rev 107478)
@@ -109,12 +109,9 @@
 
 StylePropertySet* HTMLTableCellElement::additionalAttributeStyle()
 {
-ContainerNode* p = parentNode();
-while (p  !p-hasTagName(tableTag))
-p = p-parentNode();
-if (!p)
-return 0;
-return static_castHTMLTableElement*(p)-additionalCellStyle();
+if (HTMLTableElement* table = findParentTable())
+return table-additionalCellStyle();
+return 0;
 }
 
 bool HTMLTableCellElement::isURLAttribute(Attribute *attr) const


Modified: trunk/Source/WebCore/html/HTMLTableColElement.cpp (107477 => 107478)

--- trunk/Source/WebCore/html/HTMLTableColElement.cpp	2012-02-11 14:50:28 UTC (rev 107477)
+++ trunk/Source/WebCore/html/HTMLTableColElement.cpp	2012-02-11 14:54:38 UTC (rev 107478)
@@ -72,12 +72,9 @@
 {
 if (!hasLocalName(colgroupTag))
 return 0;
-ContainerNode* p = parentNode();
-while (p  !p-hasTagName(tableTag))
-p = p-parentNode();
-if (!p)
-return 0;
-return static_castHTMLTableElement*(p)-additionalGroupStyle(false);
+if (HTMLTableElement* table = findParentTable())
+return table-additionalGroupStyle(false);
+return 0;
 }
 
 void HTMLTableColElement::setSpan(int n)


Modified: trunk/Source/WebCore/html/HTMLTablePartElement.cpp (107477 => 107478)

--- trunk/Source/WebCore/html/HTMLTablePartElement.cpp	2012-02-11 14:50:28 UTC (rev 107477)
+++ trunk/Source/WebCore/html/HTMLTablePartElement.cpp	2012-02-11 14:54:38 UTC (rev 107478)
@@ -31,6 +31,7 @@
 #include Document.h
 #include HTMLNames.h
 #include HTMLParserIdioms.h
+#include HTMLTableElement.h
 
 namespace WebCore {
 
@@ -86,4 +87,12 @@
 HTMLElement::parseAttribute(attr);
 }
 
+HTMLTableElement* HTMLTablePartElement::findParentTable() const
+{
+ContainerNode* parent = parentNode();
+while (parent  !parent-hasTagName(tableTag))
+parent = parent-parentNode();
+return static_castHTMLTableElement*(parent);
 }
+
+}


Modified: trunk/Source/WebCore/html/HTMLTablePartElement.h (107477 => 107478)

--- trunk/Source/WebCore/html/HTMLTablePartElement.h	2012-02-11 14:50:28 UTC (rev 107477)
+++ trunk/Source/WebCore/html/HTMLTablePartElement.h	2012-02-11 14:54:38 UTC (rev 107478)
@@ -30,6 +30,8 @@
 
 namespace WebCore {
 
+class HTMLTableElement;
+
 class HTMLTablePartElement : public 

[webkit-changes] [107479] trunk

2012-02-11 Thread kling
Title: [107479] trunk








Revision 107479
Author kl...@webkit.org
Date 2012-02-11 08:17:52 -0800 (Sat, 11 Feb 2012)


Log Message
Node.isEqualNode() compares attributes twice.
http://webkit.org/b/78414

Reviewed by Anders Carlsson.

Source/WebCore: 

A single pass across the attribute maps should be enough for anyone.

Added a test verifying correct behavior of Node.isEqualNode() when comparing
two elements, one of which has had attributes that were all removed,
resulting in an empty but non-null NamedNodeMap hanging off of the element.
Note that this change is not fixing a regression, I'm just adding the test
since I came close to introducing a bug here.

Test: fast/dom/isEqualNode-after-removeAttribute.html

* dom/Node.cpp:
(WebCore::Node::isEqualNode):

LayoutTests: 

* fast/dom/isEqualNode-after-removeAttribute-expected.txt: Added.
* fast/dom/isEqualNode-after-removeAttribute.html: Added.

Modified Paths

trunk/LayoutTests/ChangeLog
trunk/Source/WebCore/ChangeLog
trunk/Source/WebCore/dom/Node.cpp


Added Paths

trunk/LayoutTests/fast/dom/isEqualNode-after-removeAttribute-expected.txt
trunk/LayoutTests/fast/dom/isEqualNode-after-removeAttribute.html




Diff

Modified: trunk/LayoutTests/ChangeLog (107478 => 107479)

--- trunk/LayoutTests/ChangeLog	2012-02-11 14:54:38 UTC (rev 107478)
+++ trunk/LayoutTests/ChangeLog	2012-02-11 16:17:52 UTC (rev 107479)
@@ -1,3 +1,13 @@
+2012-02-11  Andreas Kling  awesomekl...@apple.com
+
+Node.isEqualNode() compares attributes twice.
+http://webkit.org/b/78414
+
+Reviewed by Anders Carlsson.
+
+* fast/dom/isEqualNode-after-removeAttribute-expected.txt: Added.
+* fast/dom/isEqualNode-after-removeAttribute.html: Added.
+
 2012-02-11  Csaba Osztrogonác  o...@webkit.org
 
 [Qt] Unreviewed weekend gardening.


Added: trunk/LayoutTests/fast/dom/isEqualNode-after-removeAttribute-expected.txt (0 => 107479)

--- trunk/LayoutTests/fast/dom/isEqualNode-after-removeAttribute-expected.txt	(rev 0)
+++ trunk/LayoutTests/fast/dom/isEqualNode-after-removeAttribute-expected.txt	2012-02-11 16:17:52 UTC (rev 107479)
@@ -0,0 +1,15 @@
+Tests that Node.isEqualNode behaves correctly after an element that originally had attributes has had them removed.
+
+On success, you will see a series of PASS messages, followed by TEST COMPLETE.
+
+
+PASS e1.isEqualNode(e2) is true
+PASS e1.isEqualNode(e2) is false
+PASS e1.isEqualNode(e2) is true
+PASS e1.isEqualNode(e2) is true
+PASS e1.isEqualNode(e2) is false
+PASS e1.isEqualNode(e2) is true
+PASS successfullyParsed is true
+
+TEST COMPLETE
+


Added: trunk/LayoutTests/fast/dom/isEqualNode-after-removeAttribute.html (0 => 107479)

--- trunk/LayoutTests/fast/dom/isEqualNode-after-removeAttribute.html	(rev 0)
+++ trunk/LayoutTests/fast/dom/isEqualNode-after-removeAttribute.html	2012-02-11 16:17:52 UTC (rev 107479)
@@ -0,0 +1,33 @@
+!DOCTYPE html
+html
+head
+meta charset=utf-8
+script src=""
+/head
+body
+script
+
+description(Tests that Node.isEqualNode behaves correctly after an element that originally had attributes has had them removed.);
+
+e1 = document.createElement('span');
+e2 = document.createElement('span');
+
+shouldBe(e1.isEqualNode(e2), true);
+e2.setAttribute(foo, bar);
+shouldBe(e1.isEqualNode(e2), false);
+e2.removeAttribute(foo);
+shouldBe(e1.isEqualNode(e2), true);
+
+e1 = document.createElement('span');
+e2 = document.createElement('span');
+
+shouldBe(e1.isEqualNode(e2), true);
+e1.setAttribute(foo, bar);
+shouldBe(e1.isEqualNode(e2), false);
+e1.removeAttribute(foo);
+shouldBe(e1.isEqualNode(e2), true);
+
+/script
+script src=""
+/body
+/html


Modified: trunk/Source/WebCore/ChangeLog (107478 => 107479)

--- trunk/Source/WebCore/ChangeLog	2012-02-11 14:54:38 UTC (rev 107478)
+++ trunk/Source/WebCore/ChangeLog	2012-02-11 16:17:52 UTC (rev 107479)
@@ -1,5 +1,25 @@
 2012-02-11  Andreas Kling  awesomekl...@apple.com
 
+Node.isEqualNode() compares attributes twice.
+http://webkit.org/b/78414
+
+Reviewed by Anders Carlsson.
+
+A single pass across the attribute maps should be enough for anyone.
+
+Added a test verifying correct behavior of Node.isEqualNode() when comparing
+two elements, one of which has had attributes that were all removed,
+resulting in an empty but non-null NamedNodeMap hanging off of the element.
+Note that this change is not fixing a regression, I'm just adding the test
+since I came close to introducing a bug here.
+
+Test: fast/dom/isEqualNode-after-removeAttribute.html
+
+* dom/Node.cpp:
+(WebCore::Node::isEqualNode):
+
+2012-02-11  Andreas Kling  awesomekl...@apple.com
+
 HTMLTablePartElement: Add helper method to find parent table.
 http://webkit.org/b/78413
 


Modified: trunk/Source/WebCore/dom/Node.cpp (107478 => 107479)

--- trunk/Source/WebCore/dom/Node.cpp	2012-02-11 14:54:38 UTC (rev 107478)
+++ 

[webkit-changes] [107480] trunk/Source/WebCore

2012-02-11 Thread senorblanco
Title: [107480] trunk/Source/WebCore








Revision 107480
Author senorbla...@chromium.org
Date 2012-02-11 09:46:49 -0800 (Sat, 11 Feb 2012)


Log Message
[chromium] Enable CSS filters on composited layers.
https://bugs.webkit.org/show_bug.cgi?id=77266

Reviewed by James Robinson.

Will be covered by existing tests in css3/filters (when enabled).

* WebCore.gypi:
Add CCRenderSurfaceFilters.* to the Chromium build.
* platform/graphics/chromium/GraphicsLayerChromium.cpp:
(WebCore::GraphicsLayerChromium::setFilters):
* platform/graphics/chromium/GraphicsLayerChromium.h:
Override setFilters() virtual from GraphicsLayer.
* platform/graphics/chromium/LayerChromium.cpp:
(WebCore::LayerChromium::setFilters):
Implement setFilters() to cache the filters here...
(WebCore::LayerChromium::pushPropertiesTo):
... and push them to the CCLayerImpl at commit time.
* platform/graphics/chromium/LayerChromium.h:
(WebCore::LayerChromium::filters):
Implement accessor.
* platform/graphics/chromium/RenderSurfaceChromium.h:
(WebCore::RenderSurfaceChromium::setFilters):
Implement stub version of setFilters(), to satisfy the templates.
* platform/graphics/chromium/cc/CCLayerImpl.cpp:
(WebCore::CCLayerImpl::setFilters):
Implement setter to receive filters at commit time.
* platform/graphics/chromium/cc/CCLayerImpl.h:
(WebCore::CCLayerImpl::filters):
Implement member var and accessor for filters.
* platform/graphics/chromium/cc/CCLayerTreeHostCommon.cpp:
(WebCore::calculateDrawTransformsAndVisibilityInternal):
Add another clause here to force creation of a RenderSurface when
filters are present, and to forward them from the layer to the
RenderSurface.
* platform/graphics/chromium/cc/CCRenderSurface.cpp:
(WebCore::CCRenderSurface::draw):
(WebCore::CCRenderSurface::drawLayer):
Check for filters at draw time, apply them, and forward the result
through the drawing traversal.
(WebCore::CCRenderSurface::drawSurface):
If filter bitmap is present, bind it instead of the normal
RenderSurface texture.
(WebCore::CCRenderSurface::applyFilters):
Apply filters to the render surface texture, and return the result.
This function is a no-op for the threaded compositor, due to use
of the SharedGraphicsContext3D.
* platform/graphics/chromium/cc/CCRenderSurface.h:
(WebCore::CCRenderSurface::setFilters):
(WebCore::CCRenderSurface::filters):
(CCRenderSurface):
Filters getters and setters.
* platform/graphics/chromium/cc/CCRenderSurfaceFilters.cpp: Added.
(WebCore::CCRenderSurfaceFilters::apply):
External interface for this (static) class.  All internal
implementation and helper functions are in the unnamed namespace.
* platform/graphics/chromium/cc/CCRenderSurfaceFilters.h: Added.

Modified Paths

trunk/Source/WebCore/ChangeLog
trunk/Source/WebCore/WebCore.gypi
trunk/Source/WebCore/platform/graphics/chromium/GraphicsLayerChromium.cpp
trunk/Source/WebCore/platform/graphics/chromium/GraphicsLayerChromium.h
trunk/Source/WebCore/platform/graphics/chromium/LayerChromium.cpp
trunk/Source/WebCore/platform/graphics/chromium/LayerChromium.h
trunk/Source/WebCore/platform/graphics/chromium/RenderSurfaceChromium.h
trunk/Source/WebCore/platform/graphics/chromium/cc/CCLayerImpl.cpp
trunk/Source/WebCore/platform/graphics/chromium/cc/CCLayerImpl.h
trunk/Source/WebCore/platform/graphics/chromium/cc/CCLayerTreeHostCommon.cpp
trunk/Source/WebCore/platform/graphics/chromium/cc/CCRenderSurface.cpp
trunk/Source/WebCore/platform/graphics/chromium/cc/CCRenderSurface.h


Added Paths

trunk/Source/WebCore/platform/graphics/chromium/cc/CCRenderSurfaceFilters.cpp
trunk/Source/WebCore/platform/graphics/chromium/cc/CCRenderSurfaceFilters.h




Diff

Modified: trunk/Source/WebCore/ChangeLog (107479 => 107480)

--- trunk/Source/WebCore/ChangeLog	2012-02-11 16:17:52 UTC (rev 107479)
+++ trunk/Source/WebCore/ChangeLog	2012-02-11 17:46:49 UTC (rev 107480)
@@ -1,3 +1,63 @@
+2012-02-08  Stephen White  senorbla...@chromium.org
+
+[chromium] Enable CSS filters on composited layers.
+https://bugs.webkit.org/show_bug.cgi?id=77266
+
+Reviewed by James Robinson.
+
+Will be covered by existing tests in css3/filters (when enabled).
+
+* WebCore.gypi:
+Add CCRenderSurfaceFilters.* to the Chromium build.
+* platform/graphics/chromium/GraphicsLayerChromium.cpp:
+(WebCore::GraphicsLayerChromium::setFilters):
+* platform/graphics/chromium/GraphicsLayerChromium.h:
+Override setFilters() virtual from GraphicsLayer.
+* platform/graphics/chromium/LayerChromium.cpp:
+(WebCore::LayerChromium::setFilters):
+Implement setFilters() to cache the filters here...
+(WebCore::LayerChromium::pushPropertiesTo):
+... and push them to the CCLayerImpl at commit time.
+* platform/graphics/chromium/LayerChromium.h:
+(WebCore::LayerChromium::filters):
+Implement accessor.
+* platform/graphics/chromium/RenderSurfaceChromium.h:
+

[webkit-changes] [107481] trunk/Source/WebCore

2012-02-11 Thread andersca
Title: [107481] trunk/Source/WebCore








Revision 107481
Author ander...@apple.com
Date 2012-02-11 10:18:08 -0800 (Sat, 11 Feb 2012)


Log Message
Try to fix the Windows build.

* platform/win/PopupMenuWin.cpp:
(WebCore::PopupMenuWin::scrollToRevealSelection):

Modified Paths

trunk/Source/WebCore/ChangeLog
trunk/Source/WebCore/platform/win/PopupMenuWin.cpp




Diff

Modified: trunk/Source/WebCore/ChangeLog (107480 => 107481)

--- trunk/Source/WebCore/ChangeLog	2012-02-11 17:46:49 UTC (rev 107480)
+++ trunk/Source/WebCore/ChangeLog	2012-02-11 18:18:08 UTC (rev 107481)
@@ -1,3 +1,10 @@
+2012-02-11  Anders Carlsson  ander...@apple.com
+
+Try to fix the Windows build.
+
+* platform/win/PopupMenuWin.cpp:
+(WebCore::PopupMenuWin::scrollToRevealSelection):
+
 2012-02-08  Stephen White  senorbla...@chromium.org
 
 [chromium] Enable CSS filters on composited layers.


Modified: trunk/Source/WebCore/platform/win/PopupMenuWin.cpp (107480 => 107481)

--- trunk/Source/WebCore/platform/win/PopupMenuWin.cpp	2012-02-11 17:46:49 UTC (rev 107480)
+++ trunk/Source/WebCore/platform/win/PopupMenuWin.cpp	2012-02-11 18:18:08 UTC (rev 107481)
@@ -536,12 +536,12 @@
 int index = focusedIndex();
 
 if (index  m_scrollOffset) {
-ScrollableArea::scrollToOffsetWithoutAnimation(0, index);
+ScrollableArea::scrollToOffsetWithoutAnimation(VerticalScrollbar, index);
 return true;
 }
 
 if (index = m_scrollOffset + visibleItems()) {
-ScrollableArea::scrollToOffsetWithoutAnimation(0, index - visibleItems() + 1);
+ScrollableArea::scrollToOffsetWithoutAnimation(VerticalScrollbar, index - visibleItems() + 1);
 return true;
 }
 






___
webkit-changes mailing list
webkit-changes@lists.webkit.org
http://lists.webkit.org/mailman/listinfo.cgi/webkit-changes


[webkit-changes] [107482] trunk/Source/WebCore

2012-02-11 Thread commit-queue
Title: [107482] trunk/Source/WebCore








Revision 107482
Author commit-qu...@webkit.org
Date 2012-02-11 10:36:37 -0800 (Sat, 11 Feb 2012)


Log Message
[GStreamer] html5test.com says that gstreamer ports do not support WebM for audio
https://bugs.webkit.org/show_bug.cgi?id=78244

Patch by Martin Robinson mrobin...@igalia.com on 2012-02-11
Reviewed by Eric Seidel.

Specifically advertise support for audio/webm when we support the vorbis
audio codec. This is necessary because gstreamer doesn't advertise it.

* platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
(WebCore::mimeTypeCache): Add an override for audio/webm.

Modified Paths

trunk/Source/WebCore/ChangeLog
trunk/Source/WebCore/platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp




Diff

Modified: trunk/Source/WebCore/ChangeLog (107481 => 107482)

--- trunk/Source/WebCore/ChangeLog	2012-02-11 18:18:08 UTC (rev 107481)
+++ trunk/Source/WebCore/ChangeLog	2012-02-11 18:36:37 UTC (rev 107482)
@@ -1,3 +1,16 @@
+2012-02-11  Martin Robinson  mrobin...@igalia.com
+
+[GStreamer] html5test.com says that gstreamer ports do not support WebM for audio
+https://bugs.webkit.org/show_bug.cgi?id=78244
+
+Reviewed by Eric Seidel.
+
+Specifically advertise support for audio/webm when we support the vorbis
+audio codec. This is necessary because gstreamer doesn't advertise it.
+
+* platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
+(WebCore::mimeTypeCache): Add an override for audio/webm.
+
 2012-02-11  Anders Carlsson  ander...@apple.com
 
 Try to fix the Windows build.


Modified: trunk/Source/WebCore/platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp (107481 => 107482)

--- trunk/Source/WebCore/platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp	2012-02-11 18:18:08 UTC (rev 107481)
+++ trunk/Source/WebCore/platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp	2012-02-11 18:36:37 UTC (rev 107482)
@@ -1494,8 +1494,10 @@
 const gchar* name = gst_structure_get_name(structure);
 bool cached = false;
 
-// These formats are supported by GStreamer, but not
-// correctly advertised.
+// There isn't a one-to-one correspondance of caps to supported mime types in
+// GStreamer, so we need to manually map between them. At some point in the future,
+// GStreamer may reduce the differences between caps and mime types and we can
+// remove mappings.
 if (g_str_equal(name, video/x-h264)) {
 cache.add(String(video/mp4));
 cached = true;
@@ -1523,6 +1525,7 @@
 if (g_str_equal(name, audio/x-vorbis)) {
 cache.add(String(audio/ogg));
 cache.add(String(audio/x-vorbis+ogg));
+cache.add(String(audio/webm));
 cached = true;
 }
 






___
webkit-changes mailing list
webkit-changes@lists.webkit.org
http://lists.webkit.org/mailman/listinfo.cgi/webkit-changes


[webkit-changes] [107483] trunk

2012-02-11 Thread commit-queue
Title: [107483] trunk








Revision 107483
Author commit-qu...@webkit.org
Date 2012-02-11 10:51:53 -0800 (Sat, 11 Feb 2012)


Log Message
HTML 5: Support click() method on HTMLElement.
https://bugs.webkit.org/show_bug.cgi?id=27880

Patch by Arko Saha a...@motorola.com on 2012-02-11
Reviewed by Timothy Hatcher.

Source/WebCore:

Test: fast/dom/click-method-on-html-element.html

* bindings/objc/PublicDOMInterfaces.h: Added click() method in DOMHTMLElement
with availability macro AVAILABLE_AFTER_WEBKIT_VERSION_5_1.
* html/HTMLButtonElement.idl: Moved click() method under LANGUAGE_OBJECTIVE_C.
* html/HTMLElement.idl: Added click() IDL method.
* html/HTMLInputElement.idl: Moved click() method under LANGUAGE_OBJECTIVE_C.

LayoutTests:

* fast/dom/click-method-on-html-element-expected.txt: Added.
* fast/dom/click-method-on-html-element.html: Added.

Modified Paths

trunk/LayoutTests/ChangeLog
trunk/Source/WebCore/ChangeLog
trunk/Source/WebCore/bindings/objc/PublicDOMInterfaces.h
trunk/Source/WebCore/html/HTMLButtonElement.idl
trunk/Source/WebCore/html/HTMLElement.idl
trunk/Source/WebCore/html/HTMLInputElement.idl


Added Paths

trunk/LayoutTests/fast/dom/click-method-on-html-element-expected.txt
trunk/LayoutTests/fast/dom/click-method-on-html-element.html




Diff

Modified: trunk/LayoutTests/ChangeLog (107482 => 107483)

--- trunk/LayoutTests/ChangeLog	2012-02-11 18:36:37 UTC (rev 107482)
+++ trunk/LayoutTests/ChangeLog	2012-02-11 18:51:53 UTC (rev 107483)
@@ -1,3 +1,13 @@
+2012-02-11   Arko Saha  a...@motorola.com
+
+HTML 5: Support click() method on HTMLElement.
+https://bugs.webkit.org/show_bug.cgi?id=27880
+
+Reviewed by Timothy Hatcher.
+
+* fast/dom/click-method-on-html-element-expected.txt: Added.
+* fast/dom/click-method-on-html-element.html: Added.
+
 2012-02-11  Andreas Kling  awesomekl...@apple.com
 
 Node.isEqualNode() compares attributes twice.


Added: trunk/LayoutTests/fast/dom/click-method-on-html-element-expected.txt (0 => 107483)

--- trunk/LayoutTests/fast/dom/click-method-on-html-element-expected.txt	(rev 0)
+++ trunk/LayoutTests/fast/dom/click-method-on-html-element-expected.txt	2012-02-11 18:51:53 UTC (rev 107483)
@@ -0,0 +1,134 @@
+The support of click() method on HTMLElement.
+
+Pass: Click event handler called for: a
+Pass: Click event handler called for: abbr
+Pass: Click event handler called for: acronym
+Pass: Click event handler called for: address
+Pass: Click event handler called for: applet
+Pass: Click event handler called for: area
+Pass: Click event handler called for: article
+Pass: Click event handler called for: aside
+Pass: Click event handler called for: audio
+Pass: Click event handler called for: b
+Pass: Click event handler called for: base
+Pass: Click event handler called for: basefont
+Pass: Click event handler called for: bdo
+Pass: Click event handler called for: bgsound
+Pass: Click event handler called for: big
+Pass: Click event handler called for: blockquote
+Pass: Click event handler called for: body
+Pass: Click event handler called for: br
+Pass: Click event handler called for: button
+Pass: Click event handler called for: canvas
+Pass: Click event handler called for: caption
+Pass: Click event handler called for: center
+Pass: Click event handler called for: cite
+Pass: Click event handler called for: code
+Pass: Click event handler called for: col
+Pass: Click event handler called for: colgroup
+Pass: Click event handler called for: command
+Pass: Click event handler called for: datalist
+Pass: Click event handler called for: dd
+Pass: Click event handler called for: del
+Pass: Click event handler called for: details
+Pass: Click event handler called for: dfn
+Pass: Click event handler called for: dir
+Pass: Click event handler called for: div
+Pass: Click event handler called for: dl
+Pass: Click event handler called for: dt
+Pass: Click event handler called for: em
+Pass: Click event handler called for: embed
+Pass: Click event handler called for: fieldset
+Pass: Click event handler called for: figcaption
+Pass: Click event handler called for: figure
+Pass: Click event handler called for: font
+Pass: Click event handler called for: footer
+Pass: Click event handler called for: form
+Pass: Click event handler called for: frame
+Pass: Click event handler called for: frameset
+Pass: Click event handler called for: h1
+Pass: Click event handler called for: h2
+Pass: Click event handler called for: h3
+Pass: Click event handler called for: h4
+Pass: Click event handler called for: h5
+Pass: Click event handler called for: h6
+Pass: Click event handler called for: head
+Pass: Click event handler called for: header
+Pass: Click event handler called for: hgroup
+Pass: Click event handler called for: hr
+Pass: Click event handler called for: html
+Pass: Click event handler called for: i
+Pass: Click event handler called for: iframe
+Pass: Click event handler called for: img
+Pass: Click 

[webkit-changes] [107486] trunk/Source/WebCore

2012-02-11 Thread andersca
Title: [107486] trunk/Source/WebCore








Revision 107486
Author ander...@apple.com
Date 2012-02-11 13:37:44 -0800 (Sat, 11 Feb 2012)


Log Message
Pass wheel events to a scroll elasticity controller on the scrolling thread
https://bugs.webkit.org/show_bug.cgi?id=78421

Reviewed by Sam Weinig.

Add a ScrollElasticityController to ScrollingTreeNodeMac and pass wheel events to it.
Fix ScrollingTreeNodeMac::scrollBy to clamp by the minimum and maximum scroll positions.

* page/scrolling/mac/ScrollingTreeNodeMac.h:
(ScrollingTreeNodeMac):
* page/scrolling/mac/ScrollingTreeNodeMac.mm:
(WebCore::ScrollingTreeNodeMac::ScrollingTreeNodeMac):
(WebCore::ScrollingTreeNodeMac::handleWheelEvent):
(WebCore::ScrollingTreeNodeMac::immediateScrollBy):
(WebCore::ScrollingTreeNodeMac::setScrollLayerPosition):
(WebCore::ScrollingTreeNodeMac::minimumScrollPosition):
(WebCore):
(WebCore::ScrollingTreeNodeMac::maximumScrollPosition):
(WebCore::ScrollingTreeNodeMac::scrollBy):

Modified Paths

trunk/Source/WebCore/ChangeLog
trunk/Source/WebCore/page/scrolling/mac/ScrollingTreeNodeMac.h
trunk/Source/WebCore/page/scrolling/mac/ScrollingTreeNodeMac.mm




Diff

Modified: trunk/Source/WebCore/ChangeLog (107485 => 107486)

--- trunk/Source/WebCore/ChangeLog	2012-02-11 21:07:25 UTC (rev 107485)
+++ trunk/Source/WebCore/ChangeLog	2012-02-11 21:37:44 UTC (rev 107486)
@@ -1,3 +1,25 @@
+2012-02-11  Anders Carlsson  ander...@apple.com
+
+Pass wheel events to a scroll elasticity controller on the scrolling thread
+https://bugs.webkit.org/show_bug.cgi?id=78421
+
+Reviewed by Sam Weinig.
+
+Add a ScrollElasticityController to ScrollingTreeNodeMac and pass wheel events to it.
+Fix ScrollingTreeNodeMac::scrollBy to clamp by the minimum and maximum scroll positions.
+
+* page/scrolling/mac/ScrollingTreeNodeMac.h:
+(ScrollingTreeNodeMac):
+* page/scrolling/mac/ScrollingTreeNodeMac.mm:
+(WebCore::ScrollingTreeNodeMac::ScrollingTreeNodeMac):
+(WebCore::ScrollingTreeNodeMac::handleWheelEvent):
+(WebCore::ScrollingTreeNodeMac::immediateScrollBy):
+(WebCore::ScrollingTreeNodeMac::setScrollLayerPosition):
+(WebCore::ScrollingTreeNodeMac::minimumScrollPosition):
+(WebCore):
+(WebCore::ScrollingTreeNodeMac::maximumScrollPosition):
+(WebCore::ScrollingTreeNodeMac::scrollBy):
+
 2012-02-11  Andreas Kling  awesomekl...@apple.com
 
 Attribute styles should be created lazily.


Modified: trunk/Source/WebCore/page/scrolling/mac/ScrollingTreeNodeMac.h (107485 => 107486)

--- trunk/Source/WebCore/page/scrolling/mac/ScrollingTreeNodeMac.h	2012-02-11 21:07:25 UTC (rev 107485)
+++ trunk/Source/WebCore/page/scrolling/mac/ScrollingTreeNodeMac.h	2012-02-11 21:37:44 UTC (rev 107486)
@@ -63,8 +63,11 @@
 IntPoint scrollPosition() const;
 void setScrollLayerPosition(const IntPoint);
 
+IntPoint minimumScrollPosition() const;
+IntPoint maximumScrollPosition() const;
 void scrollBy(const IntSize);
 
+ScrollElasticityController m_scrollElasticityController;
 RetainPtrCALayer m_scrollLayer;
 };
 


Modified: trunk/Source/WebCore/page/scrolling/mac/ScrollingTreeNodeMac.mm (107485 => 107486)

--- trunk/Source/WebCore/page/scrolling/mac/ScrollingTreeNodeMac.mm	2012-02-11 21:07:25 UTC (rev 107485)
+++ trunk/Source/WebCore/page/scrolling/mac/ScrollingTreeNodeMac.mm	2012-02-11 21:37:44 UTC (rev 107486)
@@ -41,6 +41,7 @@
 
 ScrollingTreeNodeMac::ScrollingTreeNodeMac(ScrollingTree* scrollingTree)
 : ScrollingTreeNode(scrollingTree)
+, m_scrollElasticityController(this)
 {
 }
 
@@ -54,8 +55,7 @@
 
 void ScrollingTreeNodeMac::handleWheelEvent(const PlatformWheelEvent wheelEvent)
 {
-// FXIME: This needs to handle rubberbanding.
-scrollBy(IntSize(-wheelEvent.deltaX(), -wheelEvent.deltaY()));
+m_scrollElasticityController.handleWheelEvent(wheelEvent);
 }
 
 void ScrollingTreeNodeMac::setScrollPosition(const IntPoint scrollPosition)
@@ -117,9 +117,9 @@
 return IntPoint();
 }
 
-void ScrollingTreeNodeMac::immediateScrollBy(const FloatSize)
+void ScrollingTreeNodeMac::immediateScrollBy(const FloatSize offset)
 {
-// FIXME: Implement.
+scrollBy(roundedIntSize(offset));
 }
 
 void ScrollingTreeNodeMac::immediateScrollByWithoutContentEdgeConstraints(const FloatSize)
@@ -146,13 +146,33 @@
 void ScrollingTreeNodeMac::setScrollLayerPosition(const IntPoint position)
 {
 ASSERT(!shouldUpdateScrollLayerPositionOnMainThread());
-
 m_scrollLayer.get().position = CGPointMake(-position.x(), -position.y());
 }
 
+IntPoint ScrollingTreeNodeMac::minimumScrollPosition() const
+{
+// FIXME: This should take the scroll origin into account.
+return IntPoint(0, 0);
+}
+
+IntPoint ScrollingTreeNodeMac::maximumScrollPosition() const
+{
+// FIXME: This should take the scroll origin into account.
+IntPoint position(contentsSize().width() - viewportRect().width(),
+  

[webkit-changes] [107487] trunk/Source/WebCore

2012-02-11 Thread antti
Title: [107487] trunk/Source/WebCore








Revision 107487
Author an...@apple.com
Date 2012-02-11 13:44:47 -0800 (Sat, 11 Feb 2012)


Log Message
Add size assert for Length
https://bugs.webkit.org/show_bug.cgi?id=78420

Rubber-stamped by Andreas Kling.

Length type is memory critical and must not grow.

* platform/Length.cpp:
(SameSizeAsLength):
(WebCore):

Modified Paths

trunk/Source/WebCore/ChangeLog
trunk/Source/WebCore/platform/Length.cpp




Diff

Modified: trunk/Source/WebCore/ChangeLog (107486 => 107487)

--- trunk/Source/WebCore/ChangeLog	2012-02-11 21:37:44 UTC (rev 107486)
+++ trunk/Source/WebCore/ChangeLog	2012-02-11 21:44:47 UTC (rev 107487)
@@ -1,3 +1,16 @@
+2012-02-11  Antti Koivisto  an...@apple.com
+
+Add size assert for Length
+https://bugs.webkit.org/show_bug.cgi?id=78420
+
+Rubber-stamped by Andreas Kling.
+
+Length type is memory critical and must not grow.
+
+* platform/Length.cpp:
+(SameSizeAsLength):
+(WebCore):
+
 2012-02-11  Anders Carlsson  ander...@apple.com
 
 Pass wheel events to a scroll elasticity controller on the scrolling thread


Modified: trunk/Source/WebCore/platform/Length.cpp (107486 => 107487)

--- trunk/Source/WebCore/platform/Length.cpp	2012-02-11 21:37:44 UTC (rev 107486)
+++ trunk/Source/WebCore/platform/Length.cpp	2012-02-11 21:44:47 UTC (rev 107487)
@@ -149,4 +149,10 @@
 return r.release();
 }
 
+class SameSizeAsLength {
+int32_t value;
+int32_t metaData;
+};
+COMPILE_ASSERT(sizeof(Length) == sizeof(SameSizeAsLength), length_should_stay_small);
+
 } // namespace WebCore






___
webkit-changes mailing list
webkit-changes@lists.webkit.org
http://lists.webkit.org/mailman/listinfo.cgi/webkit-changes


[webkit-changes] [107488] trunk/Source/WebCore

2012-02-11 Thread andersca
Title: [107488] trunk/Source/WebCore








Revision 107488
Author ander...@apple.com
Date 2012-02-11 15:02:59 -0800 (Sat, 11 Feb 2012)


Log Message
Implement more ScrollElasticityControllerClient member functions
https://bugs.webkit.org/show_bug.cgi?id=78425
rdar://problem/10710727

Reviewed by Sam Weinig.

* page/scrolling/ScrollingTreeNode.h:
(ScrollingTreeNode):
(WebCore::ScrollingTreeNode::horizontalScrollElasticity):
(WebCore::ScrollingTreeNode::verticalScrollElasticity):
(WebCore::ScrollingTreeNode::hasEnabledHorizontalScrollbar):
(WebCore::ScrollingTreeNode::hasEnabledVerticalScrollbar):
Add new getters.

* page/scrolling/mac/ScrollingTreeNodeMac.h:
* page/scrolling/mac/ScrollingTreeNodeMac.mm:
(WebCore::ScrollingTreeNodeMac::allowsHorizontalStretching):
(WebCore::ScrollingTreeNodeMac::allowsVerticalStretching):
(WebCore::ScrollingTreeNodeMac::stretchAmount):
(WebCore::ScrollingTreeNodeMac::pinnedInDirection):
(WebCore::ScrollingTreeNodeMac::canScrollHorizontally):
(WebCore::ScrollingTreeNodeMac::canScrollVertically):
(WebCore::ScrollingTreeNodeMac::absoluteScrollPosition):
(WebCore::ScrollingTreeNodeMac::immediateScrollByWithoutContentEdgeConstraints):
(WebCore::ScrollingTreeNodeMac::startSnapRubberbandTimer):
(WebCore::ScrollingTreeNodeMac::stopSnapRubberbandTimer):
(WebCore::ScrollingTreeNodeMac::scrollByWithoutContentEdgeConstraints):
Implement ScrollElasticityControllerClient member functions.

Modified Paths

trunk/Source/WebCore/ChangeLog
trunk/Source/WebCore/page/scrolling/ScrollingTreeNode.h
trunk/Source/WebCore/page/scrolling/mac/ScrollingTreeNodeMac.h
trunk/Source/WebCore/page/scrolling/mac/ScrollingTreeNodeMac.mm




Diff

Modified: trunk/Source/WebCore/ChangeLog (107487 => 107488)

--- trunk/Source/WebCore/ChangeLog	2012-02-11 21:44:47 UTC (rev 107487)
+++ trunk/Source/WebCore/ChangeLog	2012-02-11 23:02:59 UTC (rev 107488)
@@ -1,3 +1,34 @@
+2012-02-11  Anders Carlsson  ander...@apple.com
+
+Implement more ScrollElasticityControllerClient member functions
+https://bugs.webkit.org/show_bug.cgi?id=78425
+rdar://problem/10710727
+
+Reviewed by Sam Weinig.
+
+* page/scrolling/ScrollingTreeNode.h:
+(ScrollingTreeNode):
+(WebCore::ScrollingTreeNode::horizontalScrollElasticity):
+(WebCore::ScrollingTreeNode::verticalScrollElasticity):
+(WebCore::ScrollingTreeNode::hasEnabledHorizontalScrollbar):
+(WebCore::ScrollingTreeNode::hasEnabledVerticalScrollbar):
+Add new getters.
+
+* page/scrolling/mac/ScrollingTreeNodeMac.h:
+* page/scrolling/mac/ScrollingTreeNodeMac.mm:
+(WebCore::ScrollingTreeNodeMac::allowsHorizontalStretching):
+(WebCore::ScrollingTreeNodeMac::allowsVerticalStretching):
+(WebCore::ScrollingTreeNodeMac::stretchAmount):
+(WebCore::ScrollingTreeNodeMac::pinnedInDirection):
+(WebCore::ScrollingTreeNodeMac::canScrollHorizontally):
+(WebCore::ScrollingTreeNodeMac::canScrollVertically):
+(WebCore::ScrollingTreeNodeMac::absoluteScrollPosition):
+(WebCore::ScrollingTreeNodeMac::immediateScrollByWithoutContentEdgeConstraints):
+(WebCore::ScrollingTreeNodeMac::startSnapRubberbandTimer):
+(WebCore::ScrollingTreeNodeMac::stopSnapRubberbandTimer):
+(WebCore::ScrollingTreeNodeMac::scrollByWithoutContentEdgeConstraints):
+Implement ScrollElasticityControllerClient member functions.
+
 2012-02-11  Antti Koivisto  an...@apple.com
 
 Add size assert for Length


Modified: trunk/Source/WebCore/page/scrolling/ScrollingTreeNode.h (107487 => 107488)

--- trunk/Source/WebCore/page/scrolling/ScrollingTreeNode.h	2012-02-11 21:44:47 UTC (rev 107487)
+++ trunk/Source/WebCore/page/scrolling/ScrollingTreeNode.h	2012-02-11 23:02:59 UTC (rev 107488)
@@ -51,10 +51,18 @@
 explicit ScrollingTreeNode(ScrollingTree*);
 
 ScrollingTree* scrollingTree() const { return m_scrollingTree; }
+
 const IntRect viewportRect() const { return m_viewportRect; }
 const IntSize contentsSize() const { return m_contentsSize; }
+
 bool shouldUpdateScrollLayerPositionOnMainThread() const { return m_shouldUpdateScrollLayerPositionOnMainThread; }
 
+ScrollElasticity horizontalScrollElasticity() const { return m_horizontalScrollElasticity; }
+ScrollElasticity verticalScrollElasticity() const { return m_verticalScrollElasticity; }
+
+bool hasEnabledHorizontalScrollbar() const { return m_hasEnabledHorizontalScrollbar; }
+bool hasEnabledVerticalScrollbar() const { return m_hasEnabledVerticalScrollbar; }
+
 private:
 ScrollingTree* m_scrollingTree;
 


Modified: trunk/Source/WebCore/page/scrolling/mac/ScrollingTreeNodeMac.h (107487 => 107488)

--- trunk/Source/WebCore/page/scrolling/mac/ScrollingTreeNodeMac.h	2012-02-11 21:44:47 UTC (rev 107487)
+++ trunk/Source/WebCore/page/scrolling/mac/ScrollingTreeNodeMac.h	2012-02-11 23:02:59 UTC (rev 107488)
@@ -65,9 +65,13 @@
 
 IntPoint 

[webkit-changes] [107489] trunk/Source/JavaScriptCore

2012-02-11 Thread weinig
Title: [107489] trunk/Source/_javascript_Core








Revision 107489
Author wei...@apple.com
Date 2012-02-11 15:10:24 -0800 (Sat, 11 Feb 2012)


Log Message
Prepare _javascript_Core to build with libc++
rdar://problem/10426673
https://bugs.webkit.org/show_bug.cgi?id=78424

Reviewed by Anders Carlsson.

* wtf/NullPtr.cpp:
* wtf/NullPtr.h:
libc++ provides std::nullptr emulation, so we don't have to.

Modified Paths

trunk/Source/_javascript_Core/ChangeLog
trunk/Source/_javascript_Core/wtf/NullPtr.cpp
trunk/Source/_javascript_Core/wtf/NullPtr.h




Diff

Modified: trunk/Source/_javascript_Core/ChangeLog (107488 => 107489)

--- trunk/Source/_javascript_Core/ChangeLog	2012-02-11 23:02:59 UTC (rev 107488)
+++ trunk/Source/_javascript_Core/ChangeLog	2012-02-11 23:10:24 UTC (rev 107489)
@@ -1,3 +1,15 @@
+2012-02-11  Sam Weinig  s...@webkit.org
+
+Prepare _javascript_Core to build with libc++
+rdar://problem/10426673
+https://bugs.webkit.org/show_bug.cgi?id=78424
+
+Reviewed by Anders Carlsson.
+
+* wtf/NullPtr.cpp:
+* wtf/NullPtr.h:
+libc++ provides std::nullptr emulation, so we don't have to.
+
 2012-02-07  Filip Pizlo  fpi...@apple.com
 
 DFG should have polymorphic put_by_id caching


Modified: trunk/Source/_javascript_Core/wtf/NullPtr.cpp (107488 => 107489)

--- trunk/Source/_javascript_Core/wtf/NullPtr.cpp	2012-02-11 23:02:59 UTC (rev 107488)
+++ trunk/Source/_javascript_Core/wtf/NullPtr.cpp	2012-02-11 23:10:24 UTC (rev 107489)
@@ -27,7 +27,7 @@
 #include config.h
 #include NullPtr.h
 
-#if !COMPILER_SUPPORTS(CXX_NULLPTR)
+#if !(COMPILER_SUPPORTS(CXX_NULLPTR) || defined(_LIBCPP_VERSION))
 
 std::nullptr_t nullptr;
 


Modified: trunk/Source/_javascript_Core/wtf/NullPtr.h (107488 => 107489)

--- trunk/Source/_javascript_Core/wtf/NullPtr.h	2012-02-11 23:02:59 UTC (rev 107488)
+++ trunk/Source/_javascript_Core/wtf/NullPtr.h	2012-02-11 23:10:24 UTC (rev 107489)
@@ -31,8 +31,10 @@
 // nullptr_t type and nullptr object. They are defined in the same namespaces they
 // would be in compiler and library that had the support.
 
-#if COMPILER_SUPPORTS(CXX_NULLPTR)
+#include ciso646
 
+#if COMPILER_SUPPORTS(CXX_NULLPTR) || defined(_LIBCPP_VERSION)
+
 #include cstddef
 
 #else






___
webkit-changes mailing list
webkit-changes@lists.webkit.org
http://lists.webkit.org/mailman/listinfo.cgi/webkit-changes


[webkit-changes] [107490] trunk/Source/WebCore

2012-02-11 Thread andersca
Title: [107490] trunk/Source/WebCore








Revision 107490
Author ander...@apple.com
Date 2012-02-11 15:36:34 -0800 (Sat, 11 Feb 2012)


Log Message
Overlay scrollbars don't appear when scrolling on the scrolling thread
https://bugs.webkit.org/show_bug.cgi?id=78427

Reviewed by Sam Weinig.

Add a ScrollAnimator::notifyContentAreaScrolled and call it from ScrollAnimator::notifyContentAreaScrolled.
It is then overridden in ScrollAnimatorMac to tickle AppKit so that overlay scrollbars will be shown.

* platform/ScrollAnimator.h:
(WebCore::ScrollAnimator::notifyContentAreaScrolled):
(ScrollAnimator):
* platform/ScrollableArea.cpp:
(WebCore::ScrollableArea::notifyScrollPositionChanged):
* platform/mac/ScrollAnimatorMac.h:
(ScrollAnimatorMac):
* platform/mac/ScrollAnimatorMac.mm:
(WebCore::ScrollAnimatorMac::notifyPositionChanged):
(WebCore::ScrollAnimatorMac::notifyContentAreaScrolled):
(WebCore):

Modified Paths

trunk/Source/WebCore/ChangeLog
trunk/Source/WebCore/platform/ScrollAnimator.h
trunk/Source/WebCore/platform/ScrollableArea.cpp
trunk/Source/WebCore/platform/mac/ScrollAnimatorMac.h
trunk/Source/WebCore/platform/mac/ScrollAnimatorMac.mm




Diff

Modified: trunk/Source/WebCore/ChangeLog (107489 => 107490)

--- trunk/Source/WebCore/ChangeLog	2012-02-11 23:10:24 UTC (rev 107489)
+++ trunk/Source/WebCore/ChangeLog	2012-02-11 23:36:34 UTC (rev 107490)
@@ -1,5 +1,27 @@
 2012-02-11  Anders Carlsson  ander...@apple.com
 
+Overlay scrollbars don't appear when scrolling on the scrolling thread
+https://bugs.webkit.org/show_bug.cgi?id=78427
+
+Reviewed by Sam Weinig.
+
+Add a ScrollAnimator::notifyContentAreaScrolled and call it from ScrollAnimator::notifyContentAreaScrolled.
+It is then overridden in ScrollAnimatorMac to tickle AppKit so that overlay scrollbars will be shown.
+
+* platform/ScrollAnimator.h:
+(WebCore::ScrollAnimator::notifyContentAreaScrolled):
+(ScrollAnimator):
+* platform/ScrollableArea.cpp:
+(WebCore::ScrollableArea::notifyScrollPositionChanged):
+* platform/mac/ScrollAnimatorMac.h:
+(ScrollAnimatorMac):
+* platform/mac/ScrollAnimatorMac.mm:
+(WebCore::ScrollAnimatorMac::notifyPositionChanged):
+(WebCore::ScrollAnimatorMac::notifyContentAreaScrolled):
+(WebCore):
+
+2012-02-11  Anders Carlsson  ander...@apple.com
+
 Implement more ScrollElasticityControllerClient member functions
 https://bugs.webkit.org/show_bug.cgi?id=78425
 rdar://problem/10710727


Modified: trunk/Source/WebCore/platform/ScrollAnimator.h (107489 => 107490)

--- trunk/Source/WebCore/platform/ScrollAnimator.h	2012-02-11 23:10:24 UTC (rev 107489)
+++ trunk/Source/WebCore/platform/ScrollAnimator.h	2012-02-11 23:36:34 UTC (rev 107490)
@@ -98,6 +98,8 @@
 
 virtual bool shouldScrollbarParticipateInHitTesting(Scrollbar*) { return true; }
 
+virtual void notifyContentAreaScrolled() { }
+
 protected:
 ScrollAnimator(ScrollableArea*);
 


Modified: trunk/Source/WebCore/platform/ScrollableArea.cpp (107489 => 107490)

--- trunk/Source/WebCore/platform/ScrollableArea.cpp	2012-02-11 23:10:24 UTC (rev 107489)
+++ trunk/Source/WebCore/platform/ScrollableArea.cpp	2012-02-11 23:36:34 UTC (rev 107490)
@@ -149,6 +149,8 @@
 if (verticalScrollbar-isOverlayScrollbar()  !hasLayerForVerticalScrollbar())
 verticalScrollbar-invalidate();
 }
+
+scrollAnimator()-notifyContentAreaScrolled();
 }
 
 void ScrollableArea::zoomAnimatorTransformChanged(float, float, float, ZoomAnimationState)


Modified: trunk/Source/WebCore/platform/mac/ScrollAnimatorMac.h (107489 => 107490)

--- trunk/Source/WebCore/platform/mac/ScrollAnimatorMac.h	2012-02-11 23:10:24 UTC (rev 107489)
+++ trunk/Source/WebCore/platform/mac/ScrollAnimatorMac.h	2012-02-11 23:36:34 UTC (rev 107490)
@@ -111,6 +111,8 @@
 
 virtual bool shouldScrollbarParticipateInHitTesting(Scrollbar*);
 
+virtual void notifyContentAreaScrolled() OVERRIDE;
+
 FloatPoint adjustScrollPositionIfNecessary(const FloatPoint) const;
 
 void immediateScrollTo(const FloatPoint);


Modified: trunk/Source/WebCore/platform/mac/ScrollAnimatorMac.mm (107489 => 107490)

--- trunk/Source/WebCore/platform/mac/ScrollAnimatorMac.mm	2012-02-11 23:10:24 UTC (rev 107489)
+++ trunk/Source/WebCore/platform/mac/ScrollAnimatorMac.mm	2012-02-11 23:36:34 UTC (rev 107490)
@@ -653,13 +653,7 @@
 
 void ScrollAnimatorMac::notifyPositionChanged()
 {
-if (isScrollbarOverlayAPIAvailable()) {
-// This function is called when a page is going into the page cache, but the page 
-// isn't really scrolling in that case. We should only pass the message on to the
-// ScrollbarPainterController when we're really scrolling on an active page.
-if (scrollableArea()-isOnActivePage())
-[m_scrollbarPainterController.get() contentAreaScrolled];
-}
+notifyContentAreaScrolled();
 

[webkit-changes] [107491] trunk/Source/WebCore

2012-02-11 Thread weinig
Title: [107491] trunk/Source/WebCore








Revision 107491
Author wei...@apple.com
Date 2012-02-11 15:58:18 -0800 (Sat, 11 Feb 2012)


Log Message
Fix the windows build.

Since Windows uses an all-in-one file to compile, the isRespectedPresentationAttribute()
functions all need unique names.

* html/HTMLBodyElement.cpp:
* html/HTMLHRElement.cpp:
* html/HTMLIFrameElement.cpp:
* html/HTMLImageElement.cpp:
* html/HTMLInputElement.cpp:
* html/HTMLMarqueeElement.cpp:
* html/HTMLPlugInElement.cpp:
* html/HTMLTableCellElement.cpp:
* html/HTMLTablePartElement.cpp:
* mathml/MathMLElement.cpp:

Modified Paths

trunk/Source/WebCore/ChangeLog
trunk/Source/WebCore/html/HTMLBodyElement.cpp
trunk/Source/WebCore/html/HTMLHRElement.cpp
trunk/Source/WebCore/html/HTMLIFrameElement.cpp
trunk/Source/WebCore/html/HTMLImageElement.cpp
trunk/Source/WebCore/html/HTMLInputElement.cpp
trunk/Source/WebCore/html/HTMLMarqueeElement.cpp
trunk/Source/WebCore/html/HTMLPlugInElement.cpp
trunk/Source/WebCore/html/HTMLTableCellElement.cpp
trunk/Source/WebCore/html/HTMLTablePartElement.cpp
trunk/Source/WebCore/mathml/MathMLElement.cpp




Diff

Modified: trunk/Source/WebCore/ChangeLog (107490 => 107491)

--- trunk/Source/WebCore/ChangeLog	2012-02-11 23:36:34 UTC (rev 107490)
+++ trunk/Source/WebCore/ChangeLog	2012-02-11 23:58:18 UTC (rev 107491)
@@ -1,3 +1,21 @@
+2012-02-11  Sam Weinig  s...@webkit.org
+
+Fix the windows build.
+
+Since Windows uses an all-in-one file to compile, the isRespectedPresentationAttribute()
+functions all need unique names.
+
+* html/HTMLBodyElement.cpp:
+* html/HTMLHRElement.cpp:
+* html/HTMLIFrameElement.cpp:
+* html/HTMLImageElement.cpp:
+* html/HTMLInputElement.cpp:
+* html/HTMLMarqueeElement.cpp:
+* html/HTMLPlugInElement.cpp:
+* html/HTMLTableCellElement.cpp:
+* html/HTMLTablePartElement.cpp:
+* mathml/MathMLElement.cpp:
+
 2012-02-11  Anders Carlsson  ander...@apple.com
 
 Overlay scrollbars don't appear when scrolling on the scrolling thread


Modified: trunk/Source/WebCore/html/HTMLBodyElement.cpp (107490 => 107491)

--- trunk/Source/WebCore/html/HTMLBodyElement.cpp	2012-02-11 23:36:34 UTC (rev 107490)
+++ trunk/Source/WebCore/html/HTMLBodyElement.cpp	2012-02-11 23:58:18 UTC (rev 107491)
@@ -61,7 +61,7 @@
 {
 }
 
-static inline bool isRespectedPresentationAttribute(Attribute* attr)
+static inline bool isRespectedPresentationAttributeForHTMLBodyElement(Attribute* attr)
 {
 return attr-name() == backgroundAttr || attr-name() == marginwidthAttr || attr-name() == leftmarginAttr || attr-name() == marginheightAttr || attr-name() == topmarginAttr || attr-name() == bgcolorAttr || attr-name() == textAttr || attr-name() == bgpropertiesAttr;
 
@@ -87,14 +87,14 @@
 if (equalIgnoringCase(attr-value(), fixed))
style-setProperty(CSSPropertyBackgroundAttachment, CSSValueFixed);
 } else {
-ASSERT(!isRespectedPresentationAttribute(attr));
+ASSERT(!isRespectedPresentationAttributeForHTMLBodyElement(attr));
 HTMLElement::collectStyleForAttribute(attr, style);
 }
 }
 
 void HTMLBodyElement::parseAttribute(Attribute* attr)
 {
-if (isRespectedPresentationAttribute(attr))
+if (isRespectedPresentationAttributeForHTMLBodyElement(attr))
 setNeedsAttributeStyleUpdate();
 else if (attr-name() == vlinkAttr || attr-name() == alinkAttr || attr-name() == linkAttr) {
 if (attr-isNull()) {


Modified: trunk/Source/WebCore/html/HTMLHRElement.cpp (107490 => 107491)

--- trunk/Source/WebCore/html/HTMLHRElement.cpp	2012-02-11 23:36:34 UTC (rev 107490)
+++ trunk/Source/WebCore/html/HTMLHRElement.cpp	2012-02-11 23:58:18 UTC (rev 107491)
@@ -48,7 +48,7 @@
 return adoptRef(new HTMLHRElement(tagName, document));
 }
 
-static inline bool isRespectedPresentationAttribute(Attribute* attr)
+static inline bool isRespectedPresentationAttributeForHTMLHRElement(Attribute* attr)
 {
 return attr-name() == alignAttr || attr-name() == widthAttr || attr-name() == colorAttr || attr-name() == noshadeAttr || attr-name() == sizeAttr;
 }
@@ -95,14 +95,14 @@
 else
 addHTMLLengthToStyle(style, CSSPropertyHeight, String::number(size - 2)); // FIXME: Pass as integer.
 } else {
-ASSERT(!isRespectedPresentationAttribute(attr));
+ASSERT(!isRespectedPresentationAttributeForHTMLHRElement(attr));
 HTMLElement::collectStyleForAttribute(attr, style);
 }
 }
 
 void HTMLHRElement::parseAttribute(Attribute* attr)
 {
-if (isRespectedPresentationAttribute(attr))
+if (isRespectedPresentationAttributeForHTMLHRElement(attr))
 setNeedsAttributeStyleUpdate();
 else
 HTMLElement::parseAttribute(attr);


Modified: trunk/Source/WebCore/html/HTMLIFrameElement.cpp (107490 => 107491)

--- trunk/Source/WebCore/html/HTMLIFrameElement.cpp	2012-02-11 23:36:34 UTC (rev 107490)
+++ 

[webkit-changes] [107493] trunk/Source/JavaScriptCore

2012-02-11 Thread mhahnenberg
Title: [107493] trunk/Source/_javascript_Core








Revision 107493
Author mhahnenb...@apple.com
Date 2012-02-11 17:18:05 -0800 (Sat, 11 Feb 2012)


Log Message
Reduce the reentrancy limit of the interpreter for the iOS simulator
https://bugs.webkit.org/show_bug.cgi?id=78400

Reviewed by Gavin Barraclough.

* interpreter/Interpreter.h: Lowered the maximum reentrancy limit for large thread stacks.
(JSC):

Modified Paths

trunk/Source/_javascript_Core/ChangeLog
trunk/Source/_javascript_Core/interpreter/Interpreter.h




Diff

Modified: trunk/Source/_javascript_Core/ChangeLog (107492 => 107493)

--- trunk/Source/_javascript_Core/ChangeLog	2012-02-12 00:08:44 UTC (rev 107492)
+++ trunk/Source/_javascript_Core/ChangeLog	2012-02-12 01:18:05 UTC (rev 107493)
@@ -1,3 +1,13 @@
+2012-02-11  Mark Hahnenberg  mhahnenb...@apple.com
+
+Reduce the reentrancy limit of the interpreter for the iOS simulator
+https://bugs.webkit.org/show_bug.cgi?id=78400
+
+Reviewed by Gavin Barraclough.
+
+* interpreter/Interpreter.h: Lowered the maximum reentrancy limit for large thread stacks.
+(JSC):
+
 2012-02-11  Filip Pizlo  fpi...@apple.com
 
 [DFG] Misuse of WeakJSConstants in silentFillGPR code.


Modified: trunk/Source/_javascript_Core/interpreter/Interpreter.h (107492 => 107493)

--- trunk/Source/_javascript_Core/interpreter/Interpreter.h	2012-02-12 00:08:44 UTC (rev 107492)
+++ trunk/Source/_javascript_Core/interpreter/Interpreter.h	2012-02-12 01:18:05 UTC (rev 107493)
@@ -90,10 +90,10 @@
 }
 };
 
-#if PLATFORM(IOS)
 // We use a smaller reentrancy limit on iPhone because of the high amount of
 // stack space required on the web thread.
-enum { MaxLargeThreadReentryDepth = 93, MaxSmallThreadReentryDepth = 16 };
+#if PLATFORM(IOS)
+enum { MaxLargeThreadReentryDepth = 64, MaxSmallThreadReentryDepth = 16 };
 #else
 enum { MaxLargeThreadReentryDepth = 256, MaxSmallThreadReentryDepth = 16 };
 #endif // PLATFORM(IOS)






___
webkit-changes mailing list
webkit-changes@lists.webkit.org
http://lists.webkit.org/mailman/listinfo.cgi/webkit-changes


[webkit-changes] [107494] trunk/Source/JavaScriptCore

2012-02-11 Thread fpizlo
Title: [107494] trunk/Source/_javascript_Core








Revision 107494
Author fpi...@apple.com
Date 2012-02-11 17:24:49 -0800 (Sat, 11 Feb 2012)


Log Message
Old 32_64 JIT should assert that its use of map() is consistent with the DFG
OSR exit's expectations
https://bugs.webkit.org/show_bug.cgi?id=78419
rdar://problem/10817121

Reviewed by Oliver Hunt.

* jit/JITInlineMethods.h:
(JSC::JIT::map):

Modified Paths

trunk/Source/_javascript_Core/ChangeLog
trunk/Source/_javascript_Core/jit/JITInlineMethods.h




Diff

Modified: trunk/Source/_javascript_Core/ChangeLog (107493 => 107494)

--- trunk/Source/_javascript_Core/ChangeLog	2012-02-12 01:18:05 UTC (rev 107493)
+++ trunk/Source/_javascript_Core/ChangeLog	2012-02-12 01:24:49 UTC (rev 107494)
@@ -1,3 +1,15 @@
+2012-02-11  Filip Pizlo  fpi...@apple.com
+
+Old 32_64 JIT should assert that its use of map() is consistent with the DFG
+OSR exit's expectations
+https://bugs.webkit.org/show_bug.cgi?id=78419
+rdar://problem/10817121
+
+Reviewed by Oliver Hunt.
+
+* jit/JITInlineMethods.h:
+(JSC::JIT::map):
+
 2012-02-11  Mark Hahnenberg  mhahnenb...@apple.com
 
 Reduce the reentrancy limit of the interpreter for the iOS simulator


Modified: trunk/Source/_javascript_Core/jit/JITInlineMethods.h (107493 => 107494)

--- trunk/Source/_javascript_Core/jit/JITInlineMethods.h	2012-02-12 01:18:05 UTC (rev 107493)
+++ trunk/Source/_javascript_Core/jit/JITInlineMethods.h	2012-02-12 01:24:49 UTC (rev 107494)
@@ -680,6 +680,9 @@
 m_mappedVirtualRegisterIndex = virtualRegisterIndex;
 m_mappedTag = tag;
 m_mappedPayload = payload;
+
+ASSERT(!canBeOptimized() || m_mappedPayload == regT0);
+ASSERT(!canBeOptimized() || m_mappedTag == regT1);
 }
 
 inline void JIT::unmap(RegisterID registerID)






___
webkit-changes mailing list
webkit-changes@lists.webkit.org
http://lists.webkit.org/mailman/listinfo.cgi/webkit-changes


[webkit-changes] [107495] trunk/Source/JavaScriptCore

2012-02-11 Thread mhahnenberg
Title: [107495] trunk/Source/_javascript_Core








Revision 107495
Author mhahnenb...@apple.com
Date 2012-02-11 17:27:01 -0800 (Sat, 11 Feb 2012)


Log Message
Qt debug build fix

* heap/MarkedBlock.cpp:
(JSC::MarkedBlock::callDestructor): Platforms that don't use clang will allocate 
JSFinalObjects in the destuctor subspace, so we should remove this assert so it 
doesn't cause crashes.

Modified Paths

trunk/Source/_javascript_Core/ChangeLog
trunk/Source/_javascript_Core/heap/MarkedBlock.cpp




Diff

Modified: trunk/Source/_javascript_Core/ChangeLog (107494 => 107495)

--- trunk/Source/_javascript_Core/ChangeLog	2012-02-12 01:24:49 UTC (rev 107494)
+++ trunk/Source/_javascript_Core/ChangeLog	2012-02-12 01:27:01 UTC (rev 107495)
@@ -1,3 +1,12 @@
+2012-02-11  Mark Hahnenberg  mhahnenb...@apple.com
+
+Qt debug build fix
+
+* heap/MarkedBlock.cpp:
+(JSC::MarkedBlock::callDestructor): Platforms that don't use clang will allocate 
+JSFinalObjects in the destuctor subspace, so we should remove this assert so it 
+doesn't cause crashes.
+
 2012-02-11  Filip Pizlo  fpi...@apple.com
 
 Old 32_64 JIT should assert that its use of map() is consistent with the DFG


Modified: trunk/Source/_javascript_Core/heap/MarkedBlock.cpp (107494 => 107495)

--- trunk/Source/_javascript_Core/heap/MarkedBlock.cpp	2012-02-12 01:24:49 UTC (rev 107494)
+++ trunk/Source/_javascript_Core/heap/MarkedBlock.cpp	2012-02-12 01:27:01 UTC (rev 107495)
@@ -71,7 +71,6 @@
 #if ENABLE(SIMPLE_HEAP_PROFILING)
 m_heap-m_destroyedTypeCounts.countVPtr(vptr);
 #endif
-ASSERT(cell-classInfo() != JSFinalObject::s_info);
 cell-methodTable()-destroy(cell);
 
 cell-zap();






___
webkit-changes mailing list
webkit-changes@lists.webkit.org
http://lists.webkit.org/mailman/listinfo.cgi/webkit-changes


[webkit-changes] [107496] trunk

2012-02-11 Thread fpizlo
Title: [107496] trunk








Revision 107496
Author fpi...@apple.com
Date 2012-02-11 18:26:06 -0800 (Sat, 11 Feb 2012)


Log Message
DFG CFA assumes that a WeakJSConstant's structure is known
https://bugs.webkit.org/show_bug.cgi?id=78428
rdar://problem/10849492 rdar://problem/10849621

Source/_javascript_Core: 

Reviewed by Gavin Barraclough.

* dfg/DFGAbstractState.cpp:
(JSC::DFG::AbstractState::execute):

LayoutTests: 

Rubber stamped by Gavin Barraclough.

* fast/js/dfg-put-by-id-prototype-check-expected.txt: Added.
* fast/js/dfg-put-by-id-prototype-check.html: Added.
* fast/js/script-tests/dfg-put-by-id-prototype-check.js: Added.
(foo):
(Foo):

Modified Paths

trunk/LayoutTests/ChangeLog
trunk/Source/_javascript_Core/ChangeLog
trunk/Source/_javascript_Core/dfg/DFGAbstractState.cpp


Added Paths

trunk/LayoutTests/fast/js/dfg-put-by-id-prototype-check-expected.txt
trunk/LayoutTests/fast/js/dfg-put-by-id-prototype-check.html
trunk/LayoutTests/fast/js/script-tests/dfg-put-by-id-prototype-check.js




Diff

Modified: trunk/LayoutTests/ChangeLog (107495 => 107496)

--- trunk/LayoutTests/ChangeLog	2012-02-12 01:27:01 UTC (rev 107495)
+++ trunk/LayoutTests/ChangeLog	2012-02-12 02:26:06 UTC (rev 107496)
@@ -1,5 +1,19 @@
 2012-02-11  Filip Pizlo  fpi...@apple.com
 
+DFG CFA assumes that a WeakJSConstant's structure is known
+https://bugs.webkit.org/show_bug.cgi?id=78428
+rdar://problem/10849492 rdar://problem/10849621
+
+Rubber stamped by Gavin Barraclough.
+
+* fast/js/dfg-put-by-id-prototype-check-expected.txt: Added.
+* fast/js/dfg-put-by-id-prototype-check.html: Added.
+* fast/js/script-tests/dfg-put-by-id-prototype-check.js: Added.
+(foo):
+(Foo):
+
+2012-02-11  Filip Pizlo  fpi...@apple.com
+
 [DFG] Misuse of WeakJSConstants in silentFillGPR code.
 https://bugs.webkit.org/show_bug.cgi?id=78423
 rdar://problem/10849353 rdar://problem/10804043


Added: trunk/LayoutTests/fast/js/dfg-put-by-id-prototype-check-expected.txt (0 => 107496)

--- trunk/LayoutTests/fast/js/dfg-put-by-id-prototype-check-expected.txt	(rev 0)
+++ trunk/LayoutTests/fast/js/dfg-put-by-id-prototype-check-expected.txt	2012-02-12 02:26:06 UTC (rev 107496)
@@ -0,0 +1,1109 @@
+Tests that the DFG CFA is not overzealous in removing prototype structure checks for put_by_id transitions.
+
+On success, you will see a series of PASS messages, followed by TEST COMPLETE.
+
+
+PASS o.f is 0
+PASS o.f is 1
+PASS o.f is 2
+PASS o.f is 3
+PASS o.f is 4
+PASS o.f is 5
+PASS o.f is 6
+PASS o.f is 7
+PASS o.f is 8
+PASS o.f is 9
+PASS o.f is 10
+PASS o.f is 11
+PASS o.f is 12
+PASS o.f is 13
+PASS o.f is 14
+PASS o.f is 15
+PASS o.f is 16
+PASS o.f is 17
+PASS o.f is 18
+PASS o.f is 19
+PASS o.f is 20
+PASS o.f is 21
+PASS o.f is 22
+PASS o.f is 23
+PASS o.f is 24
+PASS o.f is 25
+PASS o.f is 26
+PASS o.f is 27
+PASS o.f is 28
+PASS o.f is 29
+PASS o.f is 30
+PASS o.f is 31
+PASS o.f is 32
+PASS o.f is 33
+PASS o.f is 34
+PASS o.f is 35
+PASS o.f is 36
+PASS o.f is 37
+PASS o.f is 38
+PASS o.f is 39
+PASS o.f is 40
+PASS o.f is 41
+PASS o.f is 42
+PASS o.f is 43
+PASS o.f is 44
+PASS o.f is 45
+PASS o.f is 46
+PASS o.f is 47
+PASS o.f is 48
+PASS o.f is 49
+PASS o.f is 50
+PASS o.f is 51
+PASS o.f is 52
+PASS o.f is 53
+PASS o.f is 54
+PASS o.f is 55
+PASS o.f is 56
+PASS o.f is 57
+PASS o.f is 58
+PASS o.f is 59
+PASS o.f is 60
+PASS o.f is 61
+PASS o.f is 62
+PASS o.f is 63
+PASS o.f is 64
+PASS o.f is 65
+PASS o.f is 66
+PASS o.f is 67
+PASS o.f is 68
+PASS o.f is 69
+PASS o.f is 70
+PASS o.f is 71
+PASS o.f is 72
+PASS o.f is 73
+PASS o.f is 74
+PASS o.f is 75
+PASS o.f is 76
+PASS o.f is 77
+PASS o.f is 78
+PASS o.f is 79
+PASS o.f is 80
+PASS o.f is 81
+PASS o.f is 82
+PASS o.f is 83
+PASS o.f is 84
+PASS o.f is 85
+PASS o.f is 86
+PASS o.f is 87
+PASS o.f is 88
+PASS o.f is 89
+PASS o.f is 90
+PASS o.f is 91
+PASS o.f is 92
+PASS o.f is 93
+PASS o.f is 94
+PASS o.f is 95
+PASS o.f is 96
+PASS o.f is 97
+PASS o.f is 98
+PASS o.f is 99
+PASS o.f is 100
+PASS o.f is 101
+PASS o.f is 102
+PASS o.f is 103
+PASS o.f is 104
+PASS o.f is 105
+PASS o.f is 106
+PASS o.f is 107
+PASS o.f is 108
+PASS o.f is 109
+PASS o.f is 110
+PASS o.f is 111
+PASS o.f is 112
+PASS o.f is 113
+PASS o.f is 114
+PASS o.f is 115
+PASS o.f is 116
+PASS o.f is 117
+PASS o.f is 118
+PASS o.f is 119
+PASS o.f is 120
+PASS o.f is 121
+PASS o.f is 122
+PASS o.f is 123
+PASS o.f is 124
+PASS o.f is 125
+PASS o.f is 126
+PASS o.f is 127
+PASS o.f is 128
+PASS o.f is 129
+PASS o.f is 130
+PASS o.f is 131
+PASS o.f is 132
+PASS o.f is 133
+PASS o.f is 134
+PASS o.f is 135
+PASS o.f is 136
+PASS o.f is 137
+PASS o.f is 138
+PASS o.f is 139
+PASS o.f is 140
+PASS o.f is 141
+PASS o.f is 142
+PASS o.f is 143
+PASS o.f is 144
+PASS o.f is 145
+PASS o.f is 146
+PASS o.f is 147
+PASS o.f is 148
+PASS o.f is 149
+PASS o.f is 150
+PASS o.f is 151
+PASS o.f is 152
+PASS o.f is 

[webkit-changes] [107497] trunk/Source/WebCore

2012-02-11 Thread benjamin
Title: [107497] trunk/Source/WebCore








Revision 107497
Author benja...@webkit.org
Date 2012-02-11 18:33:54 -0800 (Sat, 11 Feb 2012)


Log Message
Get rid of WebCore::URLString
https://bugs.webkit.org/show_bug.cgi?id=78429

Reviewed by Adam Barth.

URLString is unused, remove the class.

* GNUmakefile.list.am:
* WebCore.gypi:
* WebCore.xcodeproj/project.pbxproj:
* platform/KURL.cpp:
* platform/KURL.h:
(KURL):
* platform/URLString.h: Removed.

Modified Paths

trunk/Source/WebCore/ChangeLog
trunk/Source/WebCore/GNUmakefile.list.am
trunk/Source/WebCore/WebCore.gypi
trunk/Source/WebCore/WebCore.xcodeproj/project.pbxproj
trunk/Source/WebCore/platform/KURL.cpp
trunk/Source/WebCore/platform/KURL.h


Removed Paths

trunk/Source/WebCore/platform/URLString.h




Diff

Modified: trunk/Source/WebCore/ChangeLog (107496 => 107497)

--- trunk/Source/WebCore/ChangeLog	2012-02-12 02:26:06 UTC (rev 107496)
+++ trunk/Source/WebCore/ChangeLog	2012-02-12 02:33:54 UTC (rev 107497)
@@ -1,3 +1,20 @@
+2012-02-11  Benjamin Poulain  benja...@webkit.org
+
+Get rid of WebCore::URLString
+https://bugs.webkit.org/show_bug.cgi?id=78429
+
+Reviewed by Adam Barth.
+
+URLString is unused, remove the class.
+
+* GNUmakefile.list.am:
+* WebCore.gypi:
+* WebCore.xcodeproj/project.pbxproj:
+* platform/KURL.cpp:
+* platform/KURL.h:
+(KURL):
+* platform/URLString.h: Removed.
+
 2012-02-11  Sam Weinig  s...@webkit.org
 
 Fix the windows build.


Modified: trunk/Source/WebCore/GNUmakefile.list.am (107496 => 107497)

--- trunk/Source/WebCore/GNUmakefile.list.am	2012-02-12 02:26:06 UTC (rev 107496)
+++ trunk/Source/WebCore/GNUmakefile.list.am	2012-02-12 02:33:54 UTC (rev 107497)
@@ -3351,7 +3351,6 @@
 	Source/WebCore/platform/Timer.cpp \
 	Source/WebCore/platform/Timer.h \
 	Source/WebCore/platform/TreeShared.h \
-	Source/WebCore/platform/URLString.h \
 	Source/WebCore/platform/UUID.cpp \
 	Source/WebCore/platform/UUID.h \
 	Source/WebCore/platform/Widget.cpp \


Modified: trunk/Source/WebCore/WebCore.gypi (107496 => 107497)

--- trunk/Source/WebCore/WebCore.gypi	2012-02-12 02:26:06 UTC (rev 107496)
+++ trunk/Source/WebCore/WebCore.gypi	2012-02-12 02:33:54 UTC (rev 107497)
@@ -303,7 +303,6 @@
 'platform/ThreadGlobalData.h',
 'platform/Timer.h',
 'platform/TreeShared.h',
-'platform/URLString.h',
 'platform/VisitedLinkStrategy.h',
 'platform/Widget.h',
 'platform/WindowsKeyboardCodes.h',


Modified: trunk/Source/WebCore/WebCore.xcodeproj/project.pbxproj (107496 => 107497)

--- trunk/Source/WebCore/WebCore.xcodeproj/project.pbxproj	2012-02-12 02:26:06 UTC (rev 107496)
+++ trunk/Source/WebCore/WebCore.xcodeproj/project.pbxproj	2012-02-12 02:33:54 UTC (rev 107497)
@@ -3312,7 +3312,6 @@
 		9705997A107D975200A50A7C /* PolicyChecker.h in Headers */ = {isa = PBXBuildFile; fileRef = 97059976107D975200A50A7C /* PolicyChecker.h */; settings = {ATTRIBUTES = (Private, ); }; };
 		970B728A144FFAC600F00A37 /* EventInterfaces.h in Headers */ = {isa = PBXBuildFile; fileRef = 970B7289144FFAC600F00A37 /* EventInterfaces.h */; settings = {ATTRIBUTES = (Private, ); }; };
 		970B72A6145008EB00F00A37 /* EventHeaders.h in Headers */ = {isa = PBXBuildFile; fileRef = 970B72A5145008EB00F00A37 /* EventHeaders.h */; };
-		971491DA12FD65E8001BFEB1 /* URLString.h in Headers */ = {isa = PBXBuildFile; fileRef = 971491D912FD65E8001BFEB1 /* URLString.h */; settings = {ATTRIBUTES = (Private, ); }; };
 		97205AAF123928CA00B17380 /* FTPDirectoryDocument.cpp in Sources */ = {isa = PBXBuildFile; fileRef = 97205AAD123928CA00B17380 /* FTPDirectoryDocument.cpp */; };
 		97205AB0123928CA00B17380 /* FTPDirectoryDocument.h in Headers */ = {isa = PBXBuildFile; fileRef = 97205AAE123928CA00B17380 /* FTPDirectoryDocument.h */; };
 		97205AB51239291000B17380 /* ImageDocument.cpp in Sources */ = {isa = PBXBuildFile; fileRef = 97205AB11239291000B17380 /* ImageDocument.cpp */; };
@@ -10157,7 +10156,6 @@
 		97059976107D975200A50A7C /* PolicyChecker.h */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.c.h; path = PolicyChecker.h; sourceTree = group; };
 		970B7289144FFAC600F00A37 /* EventInterfaces.h */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.c.h; path = EventInterfaces.h; sourceTree = group; };
 		970B72A5145008EB00F00A37 /* EventHeaders.h */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.c.h; path = EventHeaders.h; sourceTree = group; };
-		971491D912FD65E8001BFEB1 /* URLString.h */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.c.h; path = URLString.h; sourceTree = group; };
 		97205AAD123928CA00B17380 /* FTPDirectoryDocument.cpp */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.cpp.cpp; path = FTPDirectoryDocument.cpp; sourceTree = group; };
 		97205AAE123928CA00B17380 /* 

[webkit-changes] [107500] trunk/Source/JavaScriptCore

2012-02-11 Thread fpizlo
Title: [107500] trunk/Source/_javascript_Core








Revision 107500
Author fpi...@apple.com
Date 2012-02-11 19:38:08 -0800 (Sat, 11 Feb 2012)


Log Message
Build fix for Qt.

* wtf/DataLog.h:

Modified Paths

trunk/Source/_javascript_Core/ChangeLog
trunk/Source/_javascript_Core/wtf/DataLog.h




Diff

Modified: trunk/Source/_javascript_Core/ChangeLog (107499 => 107500)

--- trunk/Source/_javascript_Core/ChangeLog	2012-02-12 03:21:32 UTC (rev 107499)
+++ trunk/Source/_javascript_Core/ChangeLog	2012-02-12 03:38:08 UTC (rev 107500)
@@ -1,5 +1,11 @@
 2012-02-11  Filip Pizlo  fpi...@apple.com
 
+Build fix for Qt.
+
+* wtf/DataLog.h:
+
+2012-02-11  Filip Pizlo  fpi...@apple.com
+
 It should be possible to send all JSC debug logging to a file
 https://bugs.webkit.org/show_bug.cgi?id=78418
 


Modified: trunk/Source/_javascript_Core/wtf/DataLog.h (107499 => 107500)

--- trunk/Source/_javascript_Core/wtf/DataLog.h	2012-02-12 03:21:32 UTC (rev 107499)
+++ trunk/Source/_javascript_Core/wtf/DataLog.h	2012-02-12 03:38:08 UTC (rev 107500)
@@ -26,6 +26,7 @@
 #ifndef DataLog_h
 #define DataLog_h
 
+#include stdarg.h
 #include stdio.h
 #include wtf/Platform.h
 #include wtf/StdLibExtras.h






___
webkit-changes mailing list
webkit-changes@lists.webkit.org
http://lists.webkit.org/mailman/listinfo.cgi/webkit-changes


[webkit-changes] [107501] trunk/Source/JavaScriptCore

2012-02-11 Thread fpizlo
Title: [107501] trunk/Source/_javascript_Core








Revision 107501
Author fpi...@apple.com
Date 2012-02-11 19:54:41 -0800 (Sat, 11 Feb 2012)


Log Message
Old JIT put_by_id profiling counts every put_by_id_transition as taking slow path
https://bugs.webkit.org/show_bug.cgi?id=78430
rdar://problem/10849469 rdar://problem/10849684

Reviewed by Gavin Barraclough.

The old JIT's put_by_id transition caching involves repatching the slow call to
a generated stub. That means that the call is counted as slow case. So, this
patch inserts code to decrement the slow case count if the stub succeeds.

Looks like a ~1% speed-up on V8.

* jit/JITPropertyAccess.cpp:
(JSC::JIT::privateCompilePutByIdTransition):
* jit/JITPropertyAccess32_64.cpp:
(JSC::JIT::privateCompilePutByIdTransition):

Modified Paths

trunk/Source/_javascript_Core/ChangeLog
trunk/Source/_javascript_Core/jit/JITPropertyAccess.cpp
trunk/Source/_javascript_Core/jit/JITPropertyAccess32_64.cpp




Diff

Modified: trunk/Source/_javascript_Core/ChangeLog (107500 => 107501)

--- trunk/Source/_javascript_Core/ChangeLog	2012-02-12 03:38:08 UTC (rev 107500)
+++ trunk/Source/_javascript_Core/ChangeLog	2012-02-12 03:54:41 UTC (rev 107501)
@@ -1,5 +1,24 @@
 2012-02-11  Filip Pizlo  fpi...@apple.com
 
+Old JIT put_by_id profiling counts every put_by_id_transition as taking slow path
+https://bugs.webkit.org/show_bug.cgi?id=78430
+rdar://problem/10849469 rdar://problem/10849684
+
+Reviewed by Gavin Barraclough.
+
+The old JIT's put_by_id transition caching involves repatching the slow call to
+a generated stub. That means that the call is counted as slow case. So, this
+patch inserts code to decrement the slow case count if the stub succeeds.
+
+Looks like a ~1% speed-up on V8.
+
+* jit/JITPropertyAccess.cpp:
+(JSC::JIT::privateCompilePutByIdTransition):
+* jit/JITPropertyAccess32_64.cpp:
+(JSC::JIT::privateCompilePutByIdTransition):
+
+2012-02-11  Filip Pizlo  fpi...@apple.com
+
 Build fix for Qt.
 
 * wtf/DataLog.h:


Modified: trunk/Source/_javascript_Core/jit/JITPropertyAccess.cpp (107500 => 107501)

--- trunk/Source/_javascript_Core/jit/JITPropertyAccess.cpp	2012-02-12 03:38:08 UTC (rev 107500)
+++ trunk/Source/_javascript_Core/jit/JITPropertyAccess.cpp	2012-02-12 03:54:41 UTC (rev 107501)
@@ -526,8 +526,16 @@
 }
 }
 
-Call callTarget;
-
+// If we succeed in all of our checks, and the code was optimizable, then make sure we
+// decrement the rare case counter.
+#if ENABLE(VALUE_PROFILER)
+if (m_codeBlock-canCompileWithDFG()) {
+sub32(
+TrustedImm32(1),
+AbsoluteAddress(m_codeBlock-rareCaseProfileForBytecodeOffset(stubInfo-bytecodeIndex)-m_counter));
+}
+#endif
+
 // emit a call only if storage realloc is needed
 bool willNeedStorageRealloc = oldStructure-propertyStorageCapacity() != newStructure-propertyStorageCapacity();
 if (willNeedStorageRealloc) {


Modified: trunk/Source/_javascript_Core/jit/JITPropertyAccess32_64.cpp (107500 => 107501)

--- trunk/Source/_javascript_Core/jit/JITPropertyAccess32_64.cpp	2012-02-12 03:38:08 UTC (rev 107500)
+++ trunk/Source/_javascript_Core/jit/JITPropertyAccess32_64.cpp	2012-02-12 03:54:41 UTC (rev 107501)
@@ -493,6 +493,16 @@
 testPrototype((*it)-storedPrototype(), failureCases);
 }
 
+// If we succeed in all of our checks, and the code was optimizable, then make sure we
+// decrement the rare case counter.
+#if ENABLE(VALUE_PROFILER)
+if (m_codeBlock-canCompileWithDFG()) {
+sub32(
+TrustedImm32(1),
+AbsoluteAddress(m_codeBlock-rareCaseProfileForBytecodeOffset(stubInfo-bytecodeIndex)-m_counter));
+}
+#endif
+
 // Reallocate property storage if needed.
 Call callTarget;
 bool willNeedStorageRealloc = oldStructure-propertyStorageCapacity() != newStructure-propertyStorageCapacity();






___
webkit-changes mailing list
webkit-changes@lists.webkit.org
http://lists.webkit.org/mailman/listinfo.cgi/webkit-changes