MultiThreaded Demo:

- fixing various race conditions throughout (usage of static vars, etc)
 - addition of a few lightweight mutexes (which are compiled out by default)
 - slight code rearrangement in discreteDynamicsWorld to facilitate multithreading
 - PoolAllocator::allocate() can now be called when pool is full without
     crashing (null pointer returned)
 - PoolAllocator allocate and freeMemory, are OPTIONALLY threadsafe
     (default is un-threadsafe)
 - CollisionDispatcher no longer checks if the pool allocator is full
     before calling allocate(), instead it just calls allocate() and
     checks if the return is null -- this avoids a race condition
 - SequentialImpulseConstraintSolver OPTIONALLY uses different logic in
     getOrInitSolverBody() to avoid a race condition with kinematic bodies
 - addition of 2 classes which together allow simulation islands to be run
   in parallel:
    - btSimulationIslandManagerMt
    - btDiscreteDynamicsWorldMt
 - MultiThreadedDemo example in the example browser demonstrating use of
   OpenMP, Microsoft PPL, and Intel TBB
 - use multithreading for other demos
 - benchmark demo: add parallel raycasting

src/BulletDynamics/MLCPSolvers/btMLCPSolver.cpp

@@ -215,12 +215,12 @@ void btMLCPSolver::createMLCPFast(const btContactSolverInfo& infoGlobal)
 		jointNodeArray.reserve(2*m_allConstraintPtrArray.size());
 	}
 
-	static btMatrixXu J3;
+    btMatrixXu& J3 = m_scratchJ3;
 	{
 		BT_PROFILE("J3.resize");
 		J3.resize(2*m,8);
 	}
-	static btMatrixXu JinvM3;
+    btMatrixXu& JinvM3 = m_scratchJInvM3;
 	{
 		BT_PROFILE("JinvM3.resize/setZero");
 
@@ -230,7 +230,7 @@ void btMLCPSolver::createMLCPFast(const btContactSolverInfo& infoGlobal)
 	}
 	int cur=0;
 	int rowOffset = 0;
-	static btAlignedObjectArray<int> ofs;
+    btAlignedObjectArray<int>& ofs = m_scratchOfs;
 	{
 		BT_PROFILE("ofs resize");
 		ofs.resize(0);
@@ -489,7 +489,7 @@ void btMLCPSolver::createMLCP(const btContactSolverInfo& infoGlobal)
 		}
 	}
  
-	static btMatrixXu Minv;
+    btMatrixXu& Minv = m_scratchMInv;
 	Minv.resize(6*numBodies,6*numBodies);
 	Minv.setZero();
 	for (int i=0;i<numBodies;i++)
@@ -506,7 +506,7 @@ void btMLCPSolver::createMLCP(const btContactSolverInfo& infoGlobal)
 				setElem(Minv,i*6+3+r,i*6+3+c,orgBody? orgBody->getInvInertiaTensorWorld()[r][c] : 0);
 	}
  
-	static btMatrixXu J;
+    btMatrixXu& J = m_scratchJ;
 	J.resize(numConstraintRows,6*numBodies);
 	J.setZero();
  
@@ -541,10 +541,10 @@ void btMLCPSolver::createMLCP(const btContactSolverInfo& infoGlobal)
 		}
 	}
  
-	static btMatrixXu J_transpose;
+    btMatrixXu& J_transpose = m_scratchJTranspose;
 	J_transpose= J.transpose();
 
-	static btMatrixXu tmp;
+    btMatrixXu& tmp = m_scratchTmp;
 
 	{
 		{