bug #466: better fix for the race condition: this new patch add an initParallel()
function which must be called at the initialization time of any multi-threaded
application calling Eigen from multiple threads.

Eigen/Core

diff --git a/Eigen/Core b/Eigen/Core
index a8a15c0..0cf1016 100644
--- a/Eigen/Core
+++ b/Eigen/Core
@@ -329,9 +329,9 @@
 #include "src/Core/GeneralProduct.h"
 #include "src/Core/TriangularMatrix.h"
 #include "src/Core/SelfAdjointView.h"
+#include "src/Core/products/GeneralBlockPanelKernel.h"
 #include "src/Core/products/Parallelizer.h"
 #include "src/Core/products/CoeffBasedProduct.h"
-#include "src/Core/products/GeneralBlockPanelKernel.h"
 #include "src/Core/products/GeneralMatrixVector.h"
 #include "src/Core/products/GeneralMatrixMatrix.h"
 #include "src/Core/SolveTriangular.h"

Eigen/src/Core/products/GeneralBlockPanelKernel.h

diff --git a/Eigen/src/Core/products/GeneralBlockPanelKernel.h b/Eigen/src/Core/products/GeneralBlockPanelKernel.h
index 5728546..d631fa2 100644
--- a/Eigen/src/Core/products/GeneralBlockPanelKernel.h
+++ b/Eigen/src/Core/products/GeneralBlockPanelKernel.h
@@ -26,7 +26,7 @@
 #define EIGEN_GENERAL_BLOCK_PANEL_H
 
 namespace Eigen { 
-
+  
 namespace internal {
 
 template<typename _LhsScalar, typename _RhsScalar, bool _ConjLhs=false, bool _ConjRhs=false>
@@ -44,8 +44,7 @@
 {
   static std::ptrdiff_t m_l1CacheSize = 0;
   static std::ptrdiff_t m_l2CacheSize = 0;
-  #pragma omp threadprivate(m_l1CacheSize,m_l2CacheSize)
-  if(m_l1CacheSize==0)
+  if(m_l2CacheSize==0)
   {
     m_l1CacheSize = manage_caching_sizes_helper(queryL1CacheSize(),8 * 1024);
     m_l2CacheSize = manage_caching_sizes_helper(queryTopLevelCacheSize(),1*1024*1024);

Eigen/src/Core/products/Parallelizer.h

diff --git a/Eigen/src/Core/products/Parallelizer.h b/Eigen/src/Core/products/Parallelizer.h
index bb1d70d..7252161 100644
--- a/Eigen/src/Core/products/Parallelizer.h
+++ b/Eigen/src/Core/products/Parallelizer.h
@@ -57,12 +57,23 @@
   }
 }
 
+}
+
+/** Must be call first when calling Eigen from multiple threads */
+inline void initParallel()
+{
+  int nbt;
+  internal::manage_multi_threading(GetAction, &nbt);
+  std::ptrdiff_t l1, l2;
+  internal::manage_caching_sizes(GetAction, &l1, &l2);
+}
+
 /** \returns the max number of threads reserved for Eigen
   * \sa setNbThreads */
 inline int nbThreads()
 {
   int ret;
-  manage_multi_threading(GetAction, &ret);
+  internal::manage_multi_threading(GetAction, &ret);
   return ret;
 }
 
@@ -70,9 +81,11 @@
   * \sa nbThreads */
 inline void setNbThreads(int v)
 {
-  manage_multi_threading(SetAction, &v);
+  internal::manage_multi_threading(SetAction, &v);
 }
 
+namespace internal {
+
 template<typename Index> struct GemmParallelInfo
 {
   GemmParallelInfo() : sync(-1), users(0), rhs_start(0), rhs_length(0) {}
@@ -121,6 +134,7 @@
   if(threads==1)
     return func(0,rows, 0,cols);
 
+  Eigen::initParallel();
   func.initParallelSession();
 
   if(transpose)

doc/TopicMultithreading.dox

diff --git a/doc/TopicMultithreading.dox b/doc/TopicMultithreading.dox
new file mode 100644
index 0000000..f7d0826
--- /dev/null
+++ b/doc/TopicMultithreading.dox
@@ -0,0 +1,46 @@
+namespace Eigen {
+
+/** \page TopicMultiThreading Eigen and multi-threading
+
+\section TopicMultiThreading_MakingEigenMT Make Eigen run in parallel
+
+Some Eigen's algorithms can exploit the multiple cores present in your hardware. To this end, it is enough to enable OpenMP on your compiler, for instance:
+ * GCC: \c -fopenmp
+ * ICC: \c -openmp
+ * MSVC: check the respective option in the build properties.
+You can control the number of thread that will be used using either the OpenMP API or Eiegn's API using the following priority:
+\code
+ OMP_NUM_THREADS=n ./my_program
+ omp_set_num_threads(n);
+ Eigen::setNbThreads(n);
+\endcode
+Unless setNbThreads has been called, Eigen uses the number of threads specified by OpenMP. You can restore this bahavior by calling \code setNbThreads(0); \endcode
+You can query the number of threads that will be used with:
+\code
+n = Eigen::nbThreads(n);
+\endcode
+You can disable Eigen's multi threading at compile time by defining the EIGEN_DONT_PARALLELIZE preprocessor token.
+
+Currently, the following algorithms can make use of multi-threading:
+ * general matrix - matrix products
+ * PartialPivLU
+
+\section TopicMultiThreading_UsingEigenWithMT Using Eigen in a multi-threaded application
+
+In the case your own application is multithreaded, and multiple threads make calls to Eigen, then you have to initialize Eigen by calling the following routine \b before creating the threads:
+\code
+#include <Eigen/Core>
+
+int main(int argc, char** argv)
+{
+  Eigen::initParallel();
+  
+  ...
+}
+\endcode
+
+In the case your application is parallelized with OpenMP, you might want to disable Eigen's own parallization as detailed in the previous section.
+
+*/
+
+}