[SVG] Resources should be laid out in dependecy order.

Source/core/svg/SVGElement.cpp

Index: Source/core/svg/SVGElement.cpp
diff --git a/Source/core/svg/SVGElement.cpp b/Source/core/svg/SVGElement.cpp
index 2649482c4c99bffe4fc765f2c93b1665b6be2034..99a09ae4b3d689ac9511271b186660800587259f 100644
--- a/Source/core/svg/SVGElement.cpp
+++ b/Source/core/svg/SVGElement.cpp
@@ -463,6 +463,23 @@ void SVGElement::updateRelativeLengthsInformation(bool hasRelativeLengths, SVGEl
     }
 }
 
+void SVGElement::invalidateRelativeLengthClients(SubtreeLayoutScope* layoutScope)
+{
+    if (!inDocument())
+        return;
+
+    HashSet<SVGElement*>::iterator end = m_elementsWithRelativeLengths.end();
+    for (HashSet<SVGElement*>::iterator it = m_elementsWithRelativeLengths.begin(); it != end; ++it) {

[pdr., 2013/09/23 22:26:02]

This always scares me. Are we sure nothing in this loop can modify
m_elementsWithRelativeLengths? Can we assert it?

[f(malita), 2013/09/23 23:20:33]

Pretty sure:

* the only way to modify m_elementsWithRelativeLengths is via
updateRelativeLengthsInformation, which is only called on node
insertions/removal + svgAttributeChanged().
* the only call in the loop is setNeedsLayout, which should not trigger any of
the above.

I'll see if I can come up with a way to assert this.

+        if (*it == this)
+            continue;
+
+        if ((*it)->renderer() && (*it)->selfHasRelativeLengths())
+            (*it)->renderer()->setNeedsLayout(MarkContainingBlockChain, layoutScope);
+
+        (*it)->invalidateRelativeLengthClients(layoutScope);
+    }
+}
+
 SVGSVGElement* SVGElement::ownerSVGElement() const
 {
     ContainerNode* n = parentOrShadowHostNode();