Title: [117519] trunk/Source/_javascript_Core
- Revision
- 117519
- Author
- [email protected]
- Date
- 2012-05-17 17:55:43 -0700 (Thu, 17 May 2012)
Log Message
Not reviewed.
Rolled out r117495 because it caused som out of memory crashes.
* heap/Heap.cpp:
(JSC::Heap::collect):
Modified Paths
Diff
Modified: trunk/Source/_javascript_Core/ChangeLog (117518 => 117519)
--- trunk/Source/_javascript_Core/ChangeLog 2012-05-18 00:50:30 UTC (rev 117518)
+++ trunk/Source/_javascript_Core/ChangeLog 2012-05-18 00:55:43 UTC (rev 117519)
@@ -1,5 +1,14 @@
2012-05-17 Geoffrey Garen <[email protected]>
+ Not reviewed.
+
+ Rolled out r117495 because it caused som out of memory crashes.
+
+ * heap/Heap.cpp:
+ (JSC::Heap::collect):
+
+2012-05-17 Geoffrey Garen <[email protected]>
+
Refactored the Heap to move more MarkedSpace logic into MarkedSpace
https://bugs.webkit.org/show_bug.cgi?id=86790
Modified: trunk/Source/_javascript_Core/heap/Heap.cpp (117518 => 117519)
--- trunk/Source/_javascript_Core/heap/Heap.cpp 2012-05-18 00:50:30 UTC (rev 117518)
+++ trunk/Source/_javascript_Core/heap/Heap.cpp 2012-05-18 00:55:43 UTC (rev 117519)
@@ -706,14 +706,15 @@
m_bytesAbandoned = 0;
}
- // To avoid pathological GC churn in large / growing heaps, we set the
- // new allocation limit based on the current heap size. Note: This
- // number only governs "emergency" GC caused by rapid allocation,
- // so it's OK to be liberal here.
+ // To avoid pathological GC churn in large heaps, we set the new allocation
+ // limit to be the current size of the heap. This heuristic
+ // is a bit arbitrary. Using the current size of the heap after this
+ // collection gives us a 2X multiplier, which is a 1:1 (heap size :
+ // new bytes allocated) proportion, and seems to work well in benchmarks.
size_t newSize = size();
if (fullGC) {
m_sizeAfterLastCollect = newSize;
- m_bytesAllocatedLimit = max(2 * newSize, m_minBytesPerCycle);
+ m_bytesAllocatedLimit = max(newSize, m_minBytesPerCycle);
}
m_bytesAllocated = 0;
double lastGCEndTime = WTF::currentTime();
_______________________________________________
webkit-changes mailing list
[email protected]
http://lists.webkit.org/mailman/listinfo.cgi/webkit-changes
