diff --git a/Demos/DX11ClothDemo/CMakeLists.txt b/Demos/DX11ClothDemo/CMakeLists.txt
index e591c21a2..7485e959a 100644
--- a/Demos/DX11ClothDemo/CMakeLists.txt
+++ b/Demos/DX11ClothDemo/CMakeLists.txt
@@ -63,7 +63,11 @@ ENDIF(COMMAND cmake_policy)
 
 SET(DXSDK_DIR $ENV{DXSDK_DIR})
 SET(DX11_INCLUDE_PATH  "${DXSDK_DIR}Include" CACHE DOCSTRING "Microsoft directX SDK include path")
-SET(DX11_LIB_PATH "${DXSDK_DIR}Lib/x86" CACHE DOCSTRING "Microsoft DirectX SDK library path")
+IF (CMAKE_CL_64)
+	SET(DX11_LIB_PATH "${DXSDK_DIR}Lib/x64" CACHE DOCSTRING "Microsoft DirectX SDK library path")
+ELSE()
+	SET(DX11_LIB_PATH "${DXSDK_DIR}Lib/x86" CACHE DOCSTRING "Microsoft DirectX SDK library path")	
+ENDIF()
 
 SET(BulletClothExample_SRCS
 	DXUT/Core/DXUT.cpp
diff --git a/Extras/Serialize/BulletFileLoader/btBulletFile.cpp b/Extras/Serialize/BulletFileLoader/btBulletFile.cpp
index 154dc983c..0307b3163 100644
--- a/Extras/Serialize/BulletFileLoader/btBulletFile.cpp
+++ b/Extras/Serialize/BulletFileLoader/btBulletFile.cpp
@@ -46,7 +46,8 @@ using namespace bParse;
 btBulletFile::btBulletFile()
 :bFile("", "BULLET ")
 {
-	mMemoryDNA = new bDNA();
+	mMemoryDNA = new bDNA(); //this memory gets released in the bFile::~bFile destructor,@todo not consistent with the rule 'who allocates it, has to deallocate it"
+
 	m_DnaCopy = 0;
 
 
@@ -97,6 +98,15 @@ btBulletFile::~btBulletFile()
 {
 	if (m_DnaCopy)
 		btAlignedFree(m_DnaCopy);
+
+	
+	while (m_dataBlocks.size())
+	{
+		char* dataBlock = m_dataBlocks[m_dataBlocks.size()-1];
+		delete[] dataBlock;
+		m_dataBlocks.pop_back();
+	}
+
 }
 
 
@@ -211,7 +221,8 @@ void btBulletFile::parseData()
 
 void	btBulletFile::addDataBlock(char* dataBlock)
 {
-	//mMain->addDatablock(dataBlock);
+	m_dataBlocks.push_back(dataBlock);
+
 }
 
 
diff --git a/Extras/Serialize/BulletFileLoader/btBulletFile.h b/Extras/Serialize/BulletFileLoader/btBulletFile.h
index eda4134ce..2575cdb4e 100644
--- a/Extras/Serialize/BulletFileLoader/btBulletFile.h
+++ b/Extras/Serialize/BulletFileLoader/btBulletFile.h
@@ -50,6 +50,7 @@ namespace bParse {
 
 		btAlignedObjectArray<bStructHandle*>	m_triangleInfoMaps;
 		
+		btAlignedObjectArray<char*>				m_dataBlocks;
 		btBulletFile();
 
 		btBulletFile(const char* fileName);
diff --git a/Extras/Serialize/BulletWorldImporter/btBulletWorldImporter.h b/Extras/Serialize/BulletWorldImporter/btBulletWorldImporter.h
index 0d0579fcf..ff043a7c4 100644
--- a/Extras/Serialize/BulletWorldImporter/btBulletWorldImporter.h
+++ b/Extras/Serialize/BulletWorldImporter/btBulletWorldImporter.h
@@ -96,7 +96,7 @@ public:
 
 	///delete all memory collision shapes, rigid bodies, constraints etc. allocated during the load.
 	///make sure you don't use the dynamics world containing objects after you call this method
-	void deleteAllData();
+	virtual void deleteAllData();
 
 	bool	loadFile(const char* fileName);
 
diff --git a/src/BulletMultiThreaded/btParallelConstraintSolver.cpp b/src/BulletMultiThreaded/btParallelConstraintSolver.cpp
index 1aae42f43..8423f554d 100644
--- a/src/BulletMultiThreaded/btParallelConstraintSolver.cpp
+++ b/src/BulletMultiThreaded/btParallelConstraintSolver.cpp
@@ -745,9 +745,14 @@ void CustomSolveConstraintsParallel(
 		uint32_t tmpBytes = poolBytes - 2 * (sizeof(PfxParallelGroup) + sizeof(PfxParallelBatch)*(PFX_MAX_SOLVER_PHASES*PFX_MAX_SOLVER_BATCHES) + 128);
 		void *tmpBuff = pool.allocate(tmpBytes);
 		
-		CustomSplitConstraints(contactPairs,numContactPairs,*cgroup,cbatches,maxTasks,numRigidBodies,tmpBuff,tmpBytes);
-		CustomSplitConstraints(jointPairs,numJointPairs,*jgroup,jbatches,maxTasks,numRigidBodies,tmpBuff,tmpBytes);
+		{
+			BT_PROFILE("CustomSplitConstraints");
+			CustomSplitConstraints(contactPairs,numContactPairs,*cgroup,cbatches,maxTasks,numRigidBodies,tmpBuff,tmpBytes);
+			CustomSplitConstraints(jointPairs,numJointPairs,*jgroup,jbatches,maxTasks,numRigidBodies,tmpBuff,tmpBytes);
+		}
 
+		{
+			BT_PROFILE("PFX_CONSTRAINT_SOLVER_CMD_SOLVE_CONSTRAINTS");
 //#define SOLVE_SEQUENTIAL
 #ifdef SOLVE_SEQUENTIAL
 		CustomSolveConstraintsTask(
@@ -802,10 +807,12 @@ void CustomSolveConstraintsParallel(
 			threadSupport->waitForResponse(&arg0,&arg1);
 		}
 #endif
+		}
 		pool.clear();
 	}
 
 	{
+			BT_PROFILE("PFX_CONSTRAINT_SOLVER_CMD_POST_SOLVER");
 		int batch = ((int)numRigidBodies + maxTasks - 1) / maxTasks;
 		int rest = (int)numRigidBodies;
 		int start = 0;
@@ -954,6 +961,7 @@ struct	btParallelSolverMemoryCache
 	
 };
 
+
 btConstraintSolverIO* createSolverIO(int numThreads)
 {
 	return new btConstraintSolverIO[numThreads];
@@ -974,6 +982,7 @@ btParallelConstraintSolver::btParallelConstraintSolver(btThreadSupportInterface*
 btParallelConstraintSolver::~btParallelConstraintSolver()
 {
 	delete m_memoryCache;
+	delete m_solverIO;
 }