Title: [111264] trunk/Source/_javascript_Core
Revision
111264
Author
[email protected]
Date
2012-03-19 16:01:57 -0700 (Mon, 19 Mar 2012)

Log Message

GC should not attempt to clear LLInt instruction inline caches for code blocks that are in
the process of being generated
https://bugs.webkit.org/show_bug.cgi?id=81565

Reviewed by Oliver Hunt.

* bytecode/CodeBlock.cpp:
(JSC::CodeBlock::finalizeUnconditionally):

Modified Paths

Diff

Modified: trunk/Source/_javascript_Core/ChangeLog (111263 => 111264)


--- trunk/Source/_javascript_Core/ChangeLog	2012-03-19 22:53:57 UTC (rev 111263)
+++ trunk/Source/_javascript_Core/ChangeLog	2012-03-19 23:01:57 UTC (rev 111264)
@@ -1,3 +1,14 @@
+2012-03-19  Filip Pizlo  <[email protected]>
+
+        GC should not attempt to clear LLInt instruction inline caches for code blocks that are in
+        the process of being generated
+        https://bugs.webkit.org/show_bug.cgi?id=81565
+
+        Reviewed by Oliver Hunt.
+
+        * bytecode/CodeBlock.cpp:
+        (JSC::CodeBlock::finalizeUnconditionally):
+
 2012-03-19  Eric Seidel  <[email protected]>
 
         Fix WTF header include discipline in Chromium WebKit

Modified: trunk/Source/_javascript_Core/bytecode/CodeBlock.cpp (111263 => 111264)


--- trunk/Source/_javascript_Core/bytecode/CodeBlock.cpp	2012-03-19 22:53:57 UTC (rev 111263)
+++ trunk/Source/_javascript_Core/bytecode/CodeBlock.cpp	2012-03-19 23:01:57 UTC (rev 111264)
@@ -1765,7 +1765,7 @@
     Interpreter* interpreter = m_globalData->interpreter;
     // interpreter->classicEnabled() returns true if the old C++ interpreter is enabled. If that's enabled
     // then we're not using LLInt.
-    if (!interpreter->classicEnabled()) {
+    if (!interpreter->classicEnabled() && !!numberOfInstructions()) {
         for (size_t size = m_propertyAccessInstructions.size(), i = 0; i < size; ++i) {
             Instruction* curInstruction = &instructions()[m_propertyAccessInstructions[i]];
             switch (interpreter->getOpcodeID(curInstruction[0].u.opcode)) {
_______________________________________________
webkit-changes mailing list
[email protected]
http://lists.webkit.org/mailman/listinfo.cgi/webkit-changes

Reply via email to