bump

It did not work with clang, and I'm unsure how it worked for gcc/msvc since QuaternionBase was introduced

.hgeol

@@ -1,3 +1,8 @@
-[patterns]
-**.* = native
-eigen_autoexp_part.dat = CRLF
+[patterns]
+scripts/*.in = LF
+debug/msvc/*.dat = CRLF
+unsupported/test/mpreal/*.* = CRLF
+** = native
+
+[repository]
+native = LF

CMakeLists.txt

@@ -101,6 +101,10 @@ if(EIGEN_DEFAULT_TO_ROW_MAJOR)
   add_definitions("-DEIGEN_DEFAULT_TO_ROW_MAJOR")
 endif()
 
+add_definitions("-DEIGEN_PERMANENTLY_DISABLE_STUPID_WARNINGS")
+
+set(EIGEN_TEST_MAX_SIZE "320" CACHE STRING "Maximal matrix/vector size, default is 320")
+
 if(CMAKE_COMPILER_IS_GNUCXX)
   set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wnon-virtual-dtor -Wno-long-long -ansi -Wundef -Wcast-align -Wchar-subscripts -Wall -W -Wpointer-arith -Wwrite-strings -Wformat-security -fexceptions -fno-check-new -fno-common -fstrict-aliasing")
   set(CMAKE_CXX_FLAGS_DEBUG "-g3")
@@ -205,6 +209,7 @@ endif(MSVC)
 
 option(EIGEN_TEST_NO_EXPLICIT_VECTORIZATION "Disable explicit vectorization in tests/examples" OFF)
 option(EIGEN_TEST_X87 "Force using X87 instructions. Implies no vectorization." OFF)
+option(EIGEN_TEST_32BIT "Force generating 32bit code." OFF)
 
 if(EIGEN_TEST_X87)
   set(EIGEN_TEST_NO_EXPLICIT_VECTORIZATION ON)
@@ -216,6 +221,15 @@ if(EIGEN_TEST_X87)
   endif()
 endif()
 
+if(EIGEN_TEST_32BIT)
+  if(CMAKE_COMPILER_IS_GNUCXX)
+    set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -m32")
+    message(STATUS "Forcing generation of 32-bit code in tests/examples")
+  else()
+    message(STATUS "EIGEN_TEST_32BIT ignored on your compiler")
+  endif()
+endif()
+
 if(EIGEN_TEST_NO_EXPLICIT_VECTORIZATION)
   add_definitions(-DEIGEN_DONT_VECTORIZE=1)
   message(STATUS "Disabling vectorization in tests/examples")
@@ -267,9 +281,21 @@ install(FILES
   )
 
 if(EIGEN_BUILD_PKGCONFIG)
+    SET(path_separator ":")
+    STRING(REPLACE ${path_separator} ";" pkg_config_libdir_search "$ENV{PKG_CONFIG_LIBDIR}")
+    message(STATUS "searching for 'pkgconfig' directory in PKG_CONFIG_LIBDIR ( $ENV{PKG_CONFIG_LIBDIR} ), ${CMAKE_INSTALL_PREFIX}/share, and ${CMAKE_INSTALL_PREFIX}/lib")
+    FIND_PATH(pkg_config_libdir pkgconfig ${pkg_config_libdir_search} ${CMAKE_INSTALL_PREFIX}/share ${CMAKE_INSTALL_PREFIX}/lib ${pkg_config_libdir_search})
+    if(pkg_config_libdir)
+        SET(pkg_config_install_dir ${pkg_config_libdir})
+        message(STATUS "found ${pkg_config_libdir}/pkgconfig" )
+    else(pkg_config_libdir)
+        SET(pkg_config_install_dir ${CMAKE_INSTALL_PREFIX}/share)
+        message(STATUS "pkgconfig not found; installing in ${pkg_config_install_dir}" )
+    endif(pkg_config_libdir)
+
     configure_file(eigen3.pc.in eigen3.pc)
     install(FILES ${CMAKE_CURRENT_BINARY_DIR}/eigen3.pc
-        DESTINATION share/pkgconfig
+        DESTINATION ${pkg_config_install_dir}/pkgconfig
         )
 endif(EIGEN_BUILD_PKGCONFIG)
 
@@ -277,10 +303,10 @@ add_subdirectory(Eigen)
 
 add_subdirectory(doc EXCLUDE_FROM_ALL)
 
-include(CTest)
+include(EigenConfigureTesting)
+# fixme, not sure this line is still needed:
 enable_testing() # must be called from the root CMakeLists, see man page
-include(EigenTesting)
-ei_init_testing()
+
 
 if(EIGEN_LEAVE_TEST_IN_ALL_TARGET)
   add_subdirectory(test) # can't do EXCLUDE_FROM_ALL here, breaks CTest
@@ -288,20 +314,18 @@ else()
   add_subdirectory(test EXCLUDE_FROM_ALL)
 endif()
 
+if(EIGEN_LEAVE_TEST_IN_ALL_TARGET)
+  add_subdirectory(blas)
+  add_subdirectory(lapack)
+else()
+  add_subdirectory(blas EXCLUDE_FROM_ALL)
+  add_subdirectory(lapack EXCLUDE_FROM_ALL)
+endif()
+
 add_subdirectory(unsupported)
 
 add_subdirectory(demos EXCLUDE_FROM_ALL)
 
-if(NOT MSVC)
-  if(EIGEN_LEAVE_TEST_IN_ALL_TARGET)
-    add_subdirectory(blas)
-    add_subdirectory(lapack)
-  else()
-    add_subdirectory(blas EXCLUDE_FROM_ALL)
-    add_subdirectory(lapack EXCLUDE_FROM_ALL)
-  endif()
-endif(NOT MSVC)
-
 # must be after test and unsupported, for configuring buildtests.in
 add_subdirectory(scripts EXCLUDE_FROM_ALL)
 

CTestConfig.cmake

@@ -11,7 +11,3 @@ set(CTEST_DROP_METHOD "http")
 set(CTEST_DROP_SITE "eigen.tuxfamily.org")
 set(CTEST_DROP_LOCATION "/CDash/submit.php?project=Eigen")
 set(CTEST_DROP_SITE_CDASH TRUE)
-
-## A tribute to Dynamic!
-set(CTEST_CUSTOM_MAXIMUM_NUMBER_OF_WARNINGS "33331")
-set(CTEST_CUSTOM_MAXIMUM_NUMBER_OF_ERRORS "33331")

CTestCustom.cmake.in

@@ -0,0 +1,4 @@
+
+## A tribute to Dynamic!
+set(CTEST_CUSTOM_MAXIMUM_NUMBER_OF_WARNINGS "33331")
+set(CTEST_CUSTOM_MAXIMUM_NUMBER_OF_ERRORS "33331")

Eigen/Cholesky

@@ -3,7 +3,7 @@
 
 #include "Core"
 
-#include "src/Core/util/DisableMSVCWarnings.h"
+#include "src/Core/util/DisableStupidWarnings.h"
 
 namespace Eigen {
 
@@ -27,7 +27,7 @@ namespace Eigen {
 
 } // namespace Eigen
 
-#include "src/Core/util/EnableMSVCWarnings.h"
+#include "src/Core/util/ReenableStupidWarnings.h"
 
 #endif // EIGEN_CHOLESKY_MODULE_H
 /* vim: set filetype=cpp et sw=2 ts=2 ai: */

Eigen/CholmodSupport

@@ -0,0 +1,34 @@
+#ifndef EIGEN_CHOLMODSUPPORT_MODULE_H
+#define EIGEN_CHOLMODSUPPORT_MODULE_H
+
+#include "SparseCore"
+
+#include "src/Core/util/DisableStupidWarnings.h"
+
+extern "C" {
+  #include <cholmod.h>
+}
+
+namespace Eigen {
+
+/** \ingroup Support_modules
+  * \defgroup CholmodSupport_Module CholmodSupport module
+  *
+  *
+  * \code
+  * #include <Eigen/CholmodSupport>
+  * \endcode
+  */
+
+#include "src/misc/Solve.h"
+#include "src/misc/SparseSolve.h"
+
+#include "src/CholmodSupport/CholmodSupport.h"
+
+
+} // namespace Eigen
+
+#include "src/Core/util/ReenableStupidWarnings.h"
+
+#endif // EIGEN_CHOLMODSUPPORT_MODULE_H
+

Eigen/Core

@@ -26,8 +26,8 @@
 #ifndef EIGEN_CORE_H
 #define EIGEN_CORE_H
 
-// first thing Eigen does: prevent MSVC from committing suicide
-#include "src/Core/util/DisableMSVCWarnings.h"
+// first thing Eigen does: stop the compiler from committing suicide
+#include "src/Core/util/DisableStupidWarnings.h"
 
 // then include this file where all our macros are defined. It's really important to do it first because
 // it's where we do all the alignment settings (platform detection and honoring the user's will if he
@@ -51,16 +51,16 @@
       #define EIGEN_SSE2_ON_MSVC_2008_OR_LATER
     #endif
   #endif
-#endif
-
-// Remember that usage of defined() in a #define is undefined by the standard
-#if (defined __SSE2__) && ( (!defined __GNUC__) || EIGEN_GNUC_AT_LEAST(4,2) )
-  #define EIGEN_SSE2_BUT_NOT_OLD_GCC
+#else
+  // Remember that usage of defined() in a #define is undefined by the standard
+  #if (defined __SSE2__) && ( (!defined __GNUC__) || EIGEN_GNUC_AT_LEAST(4,2) )
+    #define EIGEN_SSE2_ON_NON_MSVC_BUT_NOT_OLD_GCC
+  #endif
 #endif
 
 #ifndef EIGEN_DONT_VECTORIZE
 
-  #if defined (EIGEN_SSE2_BUT_NOT_OLD_GCC) || defined(EIGEN_SSE2_ON_MSVC_2008_OR_LATER)
+  #if defined (EIGEN_SSE2_ON_NON_MSVC_BUT_NOT_OLD_GCC) || defined(EIGEN_SSE2_ON_MSVC_2008_OR_LATER)
 
     // Defines symbols for compile-time detection of which instructions are
     // used.
@@ -143,6 +143,7 @@
 #ifdef EIGEN_HAS_ERRNO
 #include <cerrno>
 #endif
+#include <cstddef>
 #include <cstdlib>
 #include <cmath>
 #include <complex>
@@ -158,7 +159,7 @@
 
 // for outputting debug info
 #ifdef EIGEN_DEBUG_ASSIGN
-#include<iostream>
+#include <iostream>
 #endif
 
 // required for __cpuid, needs to be included after cmath
@@ -166,7 +167,7 @@
   #include <intrin.h>
 #endif
 
-#if (defined(_CPPUNWIND) || defined(__EXCEPTIONS)) && !defined(EIGEN_NO_EXCEPTIONS)
+#if defined(_CPPUNWIND) || defined(__EXCEPTIONS)
   #define EIGEN_EXCEPTIONS
 #endif
 
@@ -174,16 +175,7 @@
   #include <new>
 #endif
 
-// this needs to be done after all possible windows C header includes and before any Eigen source includes
-// (system C++ includes are supposed to be able to deal with this already):
-// windows.h defines min and max macros which would make Eigen fail to compile.
-#if defined(min) || defined(max)
-#error The preprocessor symbols 'min' or 'max' are defined. If you are compiling on Windows, do #define NOMINMAX to prevent windows.h from defining these symbols.
-#endif
-
-// defined in bits/termios.h
-#undef B0
-
+/** \brief Namespace containing all symbols from the %Eigen library. */
 namespace Eigen {
 
 inline static const char *SimdInstructionSetsInUse(void) {
@@ -239,6 +231,8 @@ inline static const char *SimdInstructionSetsInUse(void) {
 // we use size_t frequently and we'll never remember to prepend it with std:: everytime just to
 // ensure QNX/QCC support
 using std::size_t;
+// gcc 4.6.0 wants std:: for ptrdiff_t 
+using std::ptrdiff_t;
 
 /** \defgroup Core_Module Core module
   * This is the main module of Eigen providing dense matrix and vector support
@@ -250,6 +244,10 @@ using std::size_t;
   * \endcode
   */
 
+/** \defgroup Support_modules Support modules [category]
+  * Category of modules which add support for external libraries.
+  */
+
 #include "src/Core/util/Constants.h"
 #include "src/Core/util/ForwardDeclarations.h"
 #include "src/Core/util/Meta.h"
@@ -321,7 +319,7 @@ using std::size_t;
 #include "src/Core/CommaInitializer.h"
 #include "src/Core/Flagged.h"
 #include "src/Core/ProductBase.h"
-#include "src/Core/Product.h"
+#include "src/Core/GeneralProduct.h"
 #include "src/Core/TriangularMatrix.h"
 #include "src/Core/SelfAdjointView.h"
 #include "src/Core/SolveTriangular.h"
@@ -350,11 +348,17 @@ using std::size_t;
 #include "src/Core/ArrayBase.h"
 #include "src/Core/ArrayWrapper.h"
 
+#ifdef EIGEN_ENABLE_EVALUATORS
+#include "src/Core/Product.h"
+#include "src/Core/CoreEvaluators.h"
+#include "src/Core/AssignEvaluator.h"
+#endif 
+
 } // namespace Eigen
 
 #include "src/Core/GlobalFunctions.h"
 
-#include "src/Core/util/EnableMSVCWarnings.h"
+#include "src/Core/util/ReenableStupidWarnings.h"
 
 #ifdef EIGEN2_SUPPORT
 #include "Eigen2Support"

Eigen/Eigen2Support

@@ -29,11 +29,12 @@
 #error Eigen2 support must be enabled by defining EIGEN2_SUPPORT before including any Eigen header
 #endif
 
-#include "src/Core/util/DisableMSVCWarnings.h"
+#include "src/Core/util/DisableStupidWarnings.h"
 
 namespace Eigen {
 
-/** \defgroup Eigen2Support_Module Eigen2 support module
+/** \ingroup Support_modules
+  * \defgroup Eigen2Support_Module Eigen2 support module
   * This module provides a couple of deprecated functions improving the compatibility with Eigen2.
   *
   * To use it, define EIGEN2_SUPPORT before including any Eigen header
@@ -58,11 +59,29 @@ namespace Eigen {
 
 } // namespace Eigen
 
-#include "src/Core/util/EnableMSVCWarnings.h"
+#include "src/Core/util/ReenableStupidWarnings.h"
 
 // Eigen2 used to include iostream
 #include<iostream>
 
+#define EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE_AND_SIZE(TypeSuffix, SizeSuffix) \
+using Eigen::Matrix##SizeSuffix##TypeSuffix; \
+using Eigen::Vector##SizeSuffix##TypeSuffix; \
+using Eigen::RowVector##SizeSuffix##TypeSuffix;
+
+#define EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE(TypeSuffix) \
+EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE_AND_SIZE(TypeSuffix, 2) \
+EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE_AND_SIZE(TypeSuffix, 3) \
+EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE_AND_SIZE(TypeSuffix, 4) \
+EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE_AND_SIZE(TypeSuffix, X) \
+
+#define EIGEN_USING_MATRIX_TYPEDEFS \
+EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE(i) \
+EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE(f) \
+EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE(d) \
+EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE(cf) \
+EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE(cd)
+
 #define USING_PART_OF_NAMESPACE_EIGEN \
 EIGEN_USING_MATRIX_TYPEDEFS \
 using Eigen::Matrix; \

Eigen/Eigenvalues

@@ -3,12 +3,13 @@
 
 #include "Core"
 
-#include "src/Core/util/DisableMSVCWarnings.h"
+#include "src/Core/util/DisableStupidWarnings.h"
 
 #include "Cholesky"
 #include "Jacobi"
 #include "Householder"
 #include "LU"
+#include "Geometry"
 
 namespace Eigen {
 
@@ -38,7 +39,7 @@ namespace Eigen {
 
 } // namespace Eigen
 
-#include "src/Core/util/EnableMSVCWarnings.h"
+#include "src/Core/util/ReenableStupidWarnings.h"
 
 #endif // EIGEN_EIGENVALUES_MODULE_H
 /* vim: set filetype=cpp et sw=2 ts=2 ai: */

Eigen/Geometry

@@ -3,7 +3,7 @@
 
 #include "Core"
 
-#include "src/Core/util/DisableMSVCWarnings.h"
+#include "src/Core/util/DisableStupidWarnings.h"
 
 #include "SVD"
 #include "LU"
@@ -60,7 +60,7 @@ namespace Eigen {
 
 } // namespace Eigen
 
-#include "src/Core/util/EnableMSVCWarnings.h"
+#include "src/Core/util/ReenableStupidWarnings.h"
 
 #endif // EIGEN_GEOMETRY_MODULE_H
 /* vim: set filetype=cpp et sw=2 ts=2 ai: */

Eigen/Householder

@@ -3,7 +3,7 @@
 
 #include "Core"
 
-#include "src/Core/util/DisableMSVCWarnings.h"
+#include "src/Core/util/DisableStupidWarnings.h"
 
 namespace Eigen {
 
@@ -21,7 +21,7 @@ namespace Eigen {
 
 } // namespace Eigen
 
-#include "src/Core/util/EnableMSVCWarnings.h"
+#include "src/Core/util/ReenableStupidWarnings.h"
 
 #endif // EIGEN_HOUSEHOLDER_MODULE_H
 /* vim: set filetype=cpp et sw=2 ts=2 ai: */

Eigen/IterativeLinearSolvers

@@ -0,0 +1,37 @@
+#ifndef EIGEN_ITERATIVELINEARSOLVERS_MODULE_H
+#define EIGEN_ITERATIVELINEARSOLVERS_MODULE_H
+
+#include "SparseCore"
+
+#include "src/Core/util/DisableStupidWarnings.h"
+
+namespace Eigen {
+
+/** \ingroup Sparse_modules
+  * \defgroup IterativeLinearSolvers_Module IterativeLinearSolvers module
+  *
+  * This module currently provides iterative methods to solve problems of the form \c A \c x = \c b, where \c A is a squared matrix, usually very large and sparse.
+  * Those solvers are accessible via the following classes:
+  *  - ConjugateGradient for selfadjoint (hermitian) matrices,
+  *  - BiCGSTAB for general square matrices.
+  *
+  * Such problems can also be solved using the direct sparse decomposition modules: SparseCholesky, CholmodSupport, UmfPackSupport, SuperLUSupport.
+  *
+  * \code
+  * #include <Eigen/IterativeLinearSolvers>
+  * \endcode
+  */
+
+#include "src/misc/Solve.h"
+#include "src/misc/SparseSolve.h"
+
+#include "src/IterativeLinearSolvers/IterativeSolverBase.h"
+#include "src/IterativeLinearSolvers/BasicPreconditioners.h"
+#include "src/IterativeLinearSolvers/ConjugateGradient.h"
+#include "src/IterativeLinearSolvers/BiCGSTAB.h"
+
+} // namespace Eigen
+
+#include "src/Core/util/ReenableStupidWarnings.h"
+
+#endif // EIGEN_ITERATIVELINEARSOLVERS_MODULE_H

Eigen/Jacobi

@@ -3,7 +3,7 @@
 
 #include "Core"
 
-#include "src/Core/util/DisableMSVCWarnings.h"
+#include "src/Core/util/DisableStupidWarnings.h"
 
 namespace Eigen {
 
@@ -23,7 +23,7 @@ namespace Eigen {
 
 } // namespace Eigen
 
-#include "src/Core/util/EnableMSVCWarnings.h"
+#include "src/Core/util/ReenableStupidWarnings.h"
 
 #endif // EIGEN_JACOBI_MODULE_H
 /* vim: set filetype=cpp et sw=2 ts=2 ai: */

Eigen/LU

@@ -3,7 +3,7 @@
 
 #include "Core"
 
-#include "src/Core/util/DisableMSVCWarnings.h"
+#include "src/Core/util/DisableStupidWarnings.h"
 
 namespace Eigen {
 
@@ -36,7 +36,7 @@ namespace Eigen {
 
 } // namespace Eigen
 
-#include "src/Core/util/EnableMSVCWarnings.h"
+#include "src/Core/util/ReenableStupidWarnings.h"
 
 #endif // EIGEN_LU_MODULE_H
 /* vim: set filetype=cpp et sw=2 ts=2 ai: */

Eigen/LeastSquares

@@ -5,9 +5,12 @@
 #error LeastSquares is only available in Eigen2 support mode (define EIGEN2_SUPPORT)
 #endif
 
+// exclude from normal eigen3-only documentation
+#ifdef EIGEN2_SUPPORT
+
 #include "Core"
 
-#include "src/Core/util/DisableMSVCWarnings.h"
+#include "src/Core/util/DisableStupidWarnings.h"
 
 #include "Eigenvalues"
 #include "Geometry"
@@ -26,6 +29,8 @@ namespace Eigen {
 
 } // namespace Eigen
 
-#include "src/Core/util/EnableMSVCWarnings.h"
+#include "src/Core/util/ReenableStupidWarnings.h"
+
+#endif // EIGEN2_SUPPORT
 
 #endif // EIGEN_REGRESSION_MODULE_H

Eigen/OrderingMethods

@@ -0,0 +1,27 @@
+#ifndef EIGEN_ORDERINGMETHODS_MODULE_H
+#define EIGEN_ORDERINGMETHODS_MODULE_H
+
+#include "SparseCore"
+
+#include "src/Core/util/DisableStupidWarnings.h"
+
+namespace Eigen {
+
+/** \ingroup Sparse_modules
+  * \defgroup OrderingMethods_Module OrderingMethods module
+  *
+  * This module is currently for internal use only.
+  *
+  *
+  * \code
+  * #include <Eigen/OrderingMethods>
+  * \endcode
+  */
+
+#include "src/OrderingMethods/Amd.h"
+
+} // namespace Eigen
+
+#include "src/Core/util/ReenableStupidWarnings.h"
+
+#endif // EIGEN_ORDERINGMETHODS_MODULE_H

Eigen/QR

@@ -3,7 +3,7 @@
 
 #include "Core"
 
-#include "src/Core/util/DisableMSVCWarnings.h"
+#include "src/Core/util/DisableStupidWarnings.h"
 
 #include "Cholesky"
 #include "Jacobi"
@@ -35,7 +35,7 @@ namespace Eigen {
 
 } // namespace Eigen
 
-#include "src/Core/util/EnableMSVCWarnings.h"
+#include "src/Core/util/ReenableStupidWarnings.h"
 
 #ifdef EIGEN2_SUPPORT
 #include "Eigenvalues"

Eigen/QtAlignedMalloc

@@ -6,7 +6,7 @@
 
 #if (!EIGEN_MALLOC_ALREADY_ALIGNED)
 
-#include "src/Core/util/DisableMSVCWarnings.h"
+#include "src/Core/util/DisableStupidWarnings.h"
 
 void *qMalloc(size_t size)
 {
@@ -26,7 +26,7 @@ void *qRealloc(void *ptr, size_t size)
   return newPtr;
 }
 
-#include "src/Core/util/EnableMSVCWarnings.h"
+#include "src/Core/util/ReenableStupidWarnings.h"
 
 #endif
 

Eigen/SVD

@@ -5,7 +5,7 @@
 #include "Householder"
 #include "Jacobi"
 
-#include "src/Core/util/DisableMSVCWarnings.h"
+#include "src/Core/util/DisableStupidWarnings.h"
 
 namespace Eigen {
 
@@ -32,7 +32,7 @@ namespace Eigen {
 
 } // namespace Eigen
 
-#include "src/Core/util/EnableMSVCWarnings.h"
+#include "src/Core/util/ReenableStupidWarnings.h"
 
 #endif // EIGEN_SVD_MODULE_H
 /* vim: set filetype=cpp et sw=2 ts=2 ai: */

Eigen/Sparse

@@ -1,69 +1,27 @@
 #ifndef EIGEN_SPARSE_MODULE_H
 #define EIGEN_SPARSE_MODULE_H
 
-#include "Core"
-
-#include "src/Core/util/DisableMSVCWarnings.h"
-
-#include <vector>
-#include <map>
-#include <cstdlib>
-#include <cstring>
-#include <algorithm>
-
-#ifdef EIGEN2_SUPPORT
-#define EIGEN_YES_I_KNOW_SPARSE_MODULE_IS_NOT_STABLE_YET
-#endif
-
-#ifndef EIGEN_YES_I_KNOW_SPARSE_MODULE_IS_NOT_STABLE_YET
-#error The sparse module API is not stable yet. To use it anyway, please define the EIGEN_YES_I_KNOW_SPARSE_MODULE_IS_NOT_STABLE_YET preprocessor token.
-#endif
-
 namespace Eigen {
 
-/** \defgroup Sparse_Module Sparse module
+/** \defgroup Sparse_modules Sparse modules
   *
-  *
-  *
-  * See the \ref TutorialSparse "Sparse tutorial"
+  * Meta-module including all related modules:
+  * - SparseCore
+  * - OrderingMethods
+  * - SparseCholesky
+  * - IterativeLinearSolvers
   *
   * \code
   * #include <Eigen/Sparse>
   * \endcode
   */
 
-/** The type used to identify a general sparse storage. */
-struct Sparse {};
-
-#include "src/Sparse/SparseUtil.h"
-#include "src/Sparse/SparseMatrixBase.h"
-#include "src/Sparse/CompressedStorage.h"
-#include "src/Sparse/AmbiVector.h"
-#include "src/Sparse/SparseMatrix.h"
-#include "src/Sparse/DynamicSparseMatrix.h"
-#include "src/Sparse/MappedSparseMatrix.h"
-#include "src/Sparse/SparseVector.h"
-#include "src/Sparse/CoreIterators.h"
-#include "src/Sparse/SparseBlock.h"
-#include "src/Sparse/SparseTranspose.h"
-#include "src/Sparse/SparseCwiseUnaryOp.h"
-#include "src/Sparse/SparseCwiseBinaryOp.h"
-#include "src/Sparse/SparseDot.h"
-#include "src/Sparse/SparseAssign.h"
-#include "src/Sparse/SparseRedux.h"
-#include "src/Sparse/SparseFuzzy.h"
-#include "src/Sparse/SparseProduct.h"
-#include "src/Sparse/SparseSparseProduct.h"
-#include "src/Sparse/SparseDenseProduct.h"
-#include "src/Sparse/SparseDiagonalProduct.h"
-#include "src/Sparse/SparseTriangularView.h"
-#include "src/Sparse/SparseSelfAdjointView.h"
-#include "src/Sparse/TriangularSolver.h"
-#include "src/Sparse/SparseView.h"
-
 } // namespace Eigen
 
-#include "src/Core/util/EnableMSVCWarnings.h"
+#include "SparseCore"
+#include "OrderingMethods"
+#include "SparseCholesky"
+#include "IterativeLinearSolvers"
 
 #endif // EIGEN_SPARSE_MODULE_H
 

Eigen/SparseCholesky

@@ -0,0 +1,34 @@
+#ifndef EIGEN_SPARSECHOLESKY_MODULE_H
+#define EIGEN_SPARSECHOLESKY_MODULE_H
+
+#include "SparseCore"
+
+#include "src/Core/util/DisableStupidWarnings.h"
+
+namespace Eigen {
+
+/** \ingroup Sparse_modules
+  * \defgroup SparseCholesky_Module SparseCholesky module
+  *
+  * This module currently provides two variants of the direct sparse Cholesky decomposition for selfadjoint (hermitian) matrices.
+  * Those decompositions are accessible via the following classes:
+  *  - SimplicialLLt,
+  *  - SimplicialLDLt
+  *
+  * Such problems can also be solved using the ConjugateGradient solver from the IterativeLinearSolvers module.
+  *
+  * \code
+  * #include <Eigen/SparseCholesky>
+  * \endcode
+  */
+
+#include "src/misc/Solve.h"
+#include "src/misc/SparseSolve.h"
+
+#include "src/SparseCholesky/SimplicialCholesky.h"
+
+} // namespace Eigen
+
+#include "src/Core/util/ReenableStupidWarnings.h"
+
+#endif // EIGEN_SPARSECHOLESKY_MODULE_H

Eigen/SparseCore

@@ -0,0 +1,65 @@
+#ifndef EIGEN_SPARSECORE_MODULE_H
+#define EIGEN_SPARSECORE_MODULE_H
+
+#include "Core"
+
+#include "src/Core/util/DisableStupidWarnings.h"
+
+#include <vector>
+#include <map>
+#include <cstdlib>
+#include <cstring>
+#include <algorithm>
+
+namespace Eigen {
+
+/** \ingroup Sparse_modules
+  * \defgroup SparseCore_Module SparseCore module
+  *
+  * This module provides a sparse matrix representation, and basic associatd matrix manipulations
+  * and operations.
+  *
+  * See the \ref TutorialSparse "Sparse tutorial"
+  *
+  * \code
+  * #include <Eigen/SparseCore>
+  * \endcode
+  *
+  * This module depends on: Core.
+  */
+
+/** The type used to identify a general sparse storage. */
+struct Sparse {};
+
+#include "src/SparseCore/SparseUtil.h"
+#include "src/SparseCore/SparseMatrixBase.h"
+#include "src/SparseCore/CompressedStorage.h"
+#include "src/SparseCore/AmbiVector.h"
+#include "src/SparseCore/SparseMatrix.h"
+#include "src/SparseCore/MappedSparseMatrix.h"
+#include "src/SparseCore/SparseVector.h"
+#include "src/SparseCore/CoreIterators.h"
+#include "src/SparseCore/SparseBlock.h"
+#include "src/SparseCore/SparseTranspose.h"
+#include "src/SparseCore/SparseCwiseUnaryOp.h"
+#include "src/SparseCore/SparseCwiseBinaryOp.h"
+#include "src/SparseCore/SparseDot.h"
+#include "src/SparseCore/SparseAssign.h"
+#include "src/SparseCore/SparseRedux.h"
+#include "src/SparseCore/SparseFuzzy.h"
+#include "src/SparseCore/ConservativeSparseSparseProduct.h"
+#include "src/SparseCore/SparseSparseProductWithPruning.h"
+#include "src/SparseCore/SparseProduct.h"
+#include "src/SparseCore/SparseDenseProduct.h"
+#include "src/SparseCore/SparseDiagonalProduct.h"
+#include "src/SparseCore/SparseTriangularView.h"
+#include "src/SparseCore/SparseSelfAdjointView.h"
+#include "src/SparseCore/TriangularSolver.h"
+#include "src/SparseCore/SparseView.h"
+
+} // namespace Eigen
+
+#include "src/Core/util/ReenableStupidWarnings.h"
+
+#endif // EIGEN_SPARSECORE_MODULE_H
+

Eigen/SuperLUSupport

@@ -0,0 +1,53 @@
+#ifndef EIGEN_SUPERLUSUPPORT_MODULE_H
+#define EIGEN_SUPERLUSUPPORT_MODULE_H
+
+#include "SparseCore"
+
+#include "src/Core/util/DisableStupidWarnings.h"
+
+#ifdef EMPTY
+#define EIGEN_EMPTY_WAS_ALREADY_DEFINED
+#endif
+
+typedef int int_t;
+#include <slu_Cnames.h>
+#include <supermatrix.h>
+#include <slu_util.h>
+
+// slu_util.h defines a preprocessor token named EMPTY which is really polluting,
+// so we remove it in favor of a SUPERLU_EMPTY token.
+// If EMPTY was already, defined then we don't undef it.
+
+#if defined(EIGEN_EMPTY_WAS_ALREADY_DEFINED)
+# undef EIGEN_EMPTY_WAS_ALREADY_DEFINED
+#elif defined(EMPTY)
+# undef EMPTY
+#endif
+
+#define SUPERLU_EMPTY (-1)
+
+namespace Eigen { struct SluMatrix; }
+
+namespace Eigen {
+
+/** \ingroup Support_modules
+  * \defgroup SuperLUSupport_Module SuperLUSupport module
+  *
+  * \warning When including this module, you have to use SUPERLU_EMPTY instead of EMPTY which is no longer defined because it is too polluting.
+  *
+  * \code
+  * #include <Eigen/SuperLUSupport>
+  * \endcode
+  */
+
+#include "src/misc/Solve.h"
+#include "src/misc/SparseSolve.h"
+
+#include "src/SuperLUSupport/SuperLUSupport.h"
+
+
+} // namespace Eigen
+
+#include "src/Core/util/ReenableStupidWarnings.h"
+
+#endif // EIGEN_SUPERLUSUPPORT_MODULE_H

Eigen/UmfPackSupport

@@ -0,0 +1,34 @@
+#ifndef EIGEN_UMFPACKSUPPORT_MODULE_H
+#define EIGEN_UMFPACKSUPPORT_MODULE_H
+
+#include "SparseCore"
+
+#include "src/Core/util/DisableStupidWarnings.h"
+
+extern "C" {
+#include <umfpack.h>
+}
+
+namespace Eigen {
+
+/** \ingroup Support_modules
+  * \defgroup UmfPackSupport_Module UmfPackSupport module
+  *
+  *
+  *
+  *
+  * \code
+  * #include <Eigen/UmfPackSupport>
+  * \endcode
+  */
+
+#include "src/misc/Solve.h"
+#include "src/misc/SparseSolve.h"
+
+#include "src/UmfPackSupport/UmfPackSupport.h"
+
+} // namespace Eigen
+
+#include "src/Core/util/ReenableStupidWarnings.h"
+
+#endif // EIGEN_UMFPACKSUPPORT_MODULE_H

Eigen/src/Cholesky/LDLT.h

@@ -31,7 +31,7 @@ namespace internal {
 template<typename MatrixType, int UpLo> struct LDLT_Traits;
 }
 
-/** \ingroup cholesky_Module
+/** \ingroup Cholesky_Module
   *
   * \class LDLT
   *
@@ -158,10 +158,19 @@ template<typename _MatrixType, int _UpLo> class LDLT
     }
 
     /** \returns a solution x of \f$ A x = b \f$ using the current decomposition of A.
+      *
+      * This function also supports in-place solves using the syntax <tt>x = decompositionObject.solve(x)</tt> .
       *
       * \note_about_checking_solutions
       *
-      * \sa solveInPlace(), MatrixBase::ldlt()
+      * More precisely, this method solves \f$ A x = b \f$ using the decomposition \f$ A = P^T L D L^* P \f$
+      * by solving the systems \f$ P^T y_1 = b \f$, \f$ L y_2 = y_1 \f$, \f$ D y_3 = y_2 \f$, 
+      * \f$ L^* y_4 = y_3 \f$ and \f$ P x = y_4 \f$ in succession. If the matrix \f$ A \f$ is singular, then
+      * \f$ D \f$ will also be singular (all the other matrices are invertible). In that case, the
+      * least-square solution of \f$ D y_3 = y_2 \f$ is computed. This does not mean that this function
+      * computes the least-square solution of \f$ A x = b \f$ is \f$ A \f$ is singular.
+      *
+      * \sa MatrixBase::ldlt()
       */
     template<typename Rhs>
     inline const internal::solve_retval<LDLT, Rhs>
@@ -376,7 +385,21 @@ struct solve_retval<LDLT<_MatrixType,_UpLo>, Rhs>
     dec().matrixL().solveInPlace(dst);
 
     // dst = D^-1 (L^-1 P b)
-    dst = dec().vectorD().asDiagonal().inverse() * dst;
+    // more precisely, use pseudo-inverse of D (see bug 241)
+    using std::abs;
+    using std::max;
+    typedef typename LDLTType::MatrixType MatrixType;
+    typedef typename LDLTType::Scalar Scalar;
+    typedef typename LDLTType::RealScalar RealScalar;
+    const Diagonal<const MatrixType> vectorD = dec().vectorD();
+    RealScalar tolerance = (max)(vectorD.array().abs().maxCoeff() * NumTraits<Scalar>::epsilon(),
+				 RealScalar(1) / NumTraits<RealScalar>::highest()); // motivated by LAPACK's xGELSS
+    for (Index i = 0; i < vectorD.size(); ++i) {
+      if(abs(vectorD(i)) > tolerance)
+	dst.row(i) /= vectorD(i);
+      else
+	dst.row(i).setZero();
+    }
 
     // dst = L^-T (D^-1 L^-1 P b)
     dec().matrixU().solveInPlace(dst);

Eigen/src/Cholesky/LLT.h

@@ -29,7 +29,7 @@ namespace internal{
 template<typename MatrixType, int UpLo> struct LLT_Traits;
 }
 
-/** \ingroup cholesky_Module
+/** \ingroup Cholesky_Module
   *
   * \class LLT
   *
@@ -49,6 +49,9 @@ template<typename MatrixType, int UpLo> struct LLT_Traits;
   * use LDLT instead for the semidefinite case. Also, do not use a Cholesky decomposition to determine whether a system of equations
   * has a solution.
   *
+  * Example: \include LLT_example.cpp
+  * Output: \verbinclude LLT_example.out
+  *    
   * \sa MatrixBase::llt(), class LDLT
   */
  /* HEY THIS DOX IS DISABLED BECAUSE THERE's A BUG EITHER HERE OR IN LDLT ABOUT THAT (OR BOTH)
@@ -178,6 +181,9 @@ template<typename _MatrixType, int _UpLo> class LLT
     inline Index rows() const { return m_matrix.rows(); }
     inline Index cols() const { return m_matrix.cols(); }
 
+    template<typename VectorType>
+    void rankUpdate(const VectorType& vec);
+
   protected:
     /** \internal
       * Used to compute and store L
@@ -233,7 +239,7 @@ template<> struct llt_inplace<Lower>
 
     Index blockSize = size/8;
     blockSize = (blockSize/16)*16;
-    blockSize = std::min(std::max(blockSize,Index(8)), Index(128));
+    blockSize = (std::min)((std::max)(blockSize,Index(8)), Index(128));
 
     for (Index k=0; k<size; k+=blockSize)
     {
@@ -241,7 +247,7 @@ template<> struct llt_inplace<Lower>
       //       A00 |  -  |  -
       // lu  = A10 | A11 |  -
       //       A20 | A21 | A22
-      Index bs = std::min(blockSize, size-k);
+      Index bs = (std::min)(blockSize, size-k);
       Index rs = size - k - bs;
       Block<MatrixType,Dynamic,Dynamic> A11(m,k,   k,   bs,bs);
       Block<MatrixType,Dynamic,Dynamic> A21(m,k+bs,k,   rs,bs);
@@ -254,6 +260,35 @@ template<> struct llt_inplace<Lower>
     }
     return -1;
   }
+
+  template<typename MatrixType, typename VectorType>
+  static void rankUpdate(MatrixType& mat, const VectorType& vec)
+  {
+    typedef typename MatrixType::ColXpr ColXpr;
+    typedef typename internal::remove_all<ColXpr>::type ColXprCleaned;
+    typedef typename ColXprCleaned::SegmentReturnType ColXprSegment;
+    typedef typename MatrixType::Scalar Scalar;
+    typedef Matrix<Scalar,Dynamic,1> TempVectorType;
+    typedef typename TempVectorType::SegmentReturnType TempVecSegment;
+
+    int n = mat.cols();
+    eigen_assert(mat.rows()==n && vec.size()==n);
+    TempVectorType temp(vec);
+
+    for(int i=0; i<n; ++i)
+    {
+      JacobiRotation<Scalar> g;
+      g.makeGivens(mat(i,i), -temp(i), &mat(i,i));
+
+      int rs = n-i-1;
+      if(rs>0)
+      {
+        ColXprSegment x(mat.col(i).tail(rs));
+        TempVecSegment y(temp.tail(rs));
+        apply_rotation_in_the_plane(x, y, g);
+      }
+    }
+  }
 };
 
 template<> struct llt_inplace<Upper>
@@ -270,6 +305,12 @@ template<> struct llt_inplace<Upper>
     Transpose<MatrixType> matt(mat);
     return llt_inplace<Lower>::blocked(matt);
   }
+  template<typename MatrixType, typename VectorType>
+  static void rankUpdate(MatrixType& mat, const VectorType& vec)
+  {
+    Transpose<MatrixType> matt(mat);
+    return llt_inplace<Lower>::rankUpdate(matt, vec.conjugate());
+  }
 };
 
 template<typename MatrixType> struct LLT_Traits<MatrixType,Lower>
@@ -295,9 +336,11 @@ template<typename MatrixType> struct LLT_Traits<MatrixType,Upper>
 } // end namespace internal
 
 /** Computes / recomputes the Cholesky decomposition A = LL^* = U^*U of \a matrix
-  *
   *
   * \returns a reference to *this
+  *
+  * Example: \include TutorialLinAlgComputeTwice.cpp
+  * Output: \verbinclude TutorialLinAlgComputeTwice.out
   */
 template<typename MatrixType, int _UpLo>
 LLT<MatrixType,_UpLo>& LLT<MatrixType,_UpLo>::compute(const MatrixType& a)
@@ -314,6 +357,20 @@ LLT<MatrixType,_UpLo>& LLT<MatrixType,_UpLo>::compute(const MatrixType& a)
   return *this;
 }
 
+/** Performs a rank one update of the current decomposition.
+  * If A = LL^* before the rank one update,
+  * then after it we have LL^* = A + vv^* where \a v must be a vector
+  * of same dimension.
+  *
+  */
+template<typename MatrixType, int _UpLo>
+template<typename VectorType>
+void LLT<MatrixType,_UpLo>::rankUpdate(const VectorType& v)
+{
+  EIGEN_STATIC_ASSERT_VECTOR_ONLY(VectorType);
+  internal::llt_inplace<UpLo>::rankUpdate(m_matrix,v);
+}
+    
 namespace internal {
 template<typename _MatrixType, int UpLo, typename Rhs>
 struct solve_retval<LLT<_MatrixType, UpLo>, Rhs>
@@ -384,3 +441,4 @@ SelfAdjointView<MatrixType, UpLo>::llt() const
 }
 
 #endif // EIGEN_LLT_H
+

Eigen/src/CholmodSupport/CMakeLists.txt

@@ -0,0 +1,6 @@
+FILE(GLOB Eigen_CholmodSupport_SRCS "*.h")
+
+INSTALL(FILES 
+  ${Eigen_CholmodSupport_SRCS}
+  DESTINATION ${INCLUDE_INSTALL_DIR}/Eigen/src/CholmodSupport COMPONENT Devel
+  )

Eigen/src/CholmodSupport/CholmodSupport.h

@@ -69,9 +69,9 @@ cholmod_sparse viewAsCholmod(SparseMatrix<_Scalar,_Options,_Index>& mat)
   res.nzmax   = mat.nonZeros();
   res.nrow    = mat.rows();;
   res.ncol    = mat.cols();
-  res.p       = mat._outerIndexPtr();
-  res.i       = mat._innerIndexPtr();
-  res.x       = mat._valuePtr();
+  res.p       = mat.outerIndexPtr();
+  res.i       = mat.innerIndexPtr();
+  res.x       = mat.valuePtr();
   res.sorted  = 1;
   res.packed  = 1;
   res.dtype   = 0;
@@ -149,7 +149,9 @@ enum CholmodMode {
   CholmodAuto, CholmodSimplicialLLt, CholmodSupernodalLLt, CholmodLDLt
 };
 
-/** \brief A Cholesky factorization and solver based on Cholmod
+/** \ingroup CholmodSupport_Module
+  * \class CholmodDecomposition
+  * \brief A Cholesky factorization and solver based on Cholmod
   *
   * This class allows to solve for A.X = B sparse linear problems via a LL^T or LDL^T Cholesky factorization
   * using the Cholmod library. The sparse matrix A must be selfajoint and positive definite. The vectors or matrices
@@ -159,6 +161,7 @@ enum CholmodMode {
   * \tparam _UpLo the triangular part that will be used for the computations. It can be Lower
   *               or Upper. Default is Lower.
   *
+  * \sa \ref TutorialSparseDirectSolvers
   */
 template<typename _MatrixType, int _UpLo = Lower>
 class CholmodDecomposition

Eigen/src/Core/Array.h

@@ -37,6 +37,9 @@
   * API for the %Matrix class provides easy access to linear-algebra
   * operations.
   *
+  * This class can be extended with the help of the plugin mechanism described on the page
+  * \ref TopicCustomizingEigen by defining the preprocessor symbol \c EIGEN_ARRAY_PLUGIN.
+  *
   * \sa \ref TutorialArrayClass, \ref TopicClassHierarchy
   */
 namespace internal {
@@ -65,10 +68,8 @@ class Array
     friend struct internal::conservative_resize_like_impl;
 
     using Base::m_storage;
+
   public:
-    enum { NeedsToAlign = (!(Options&DontAlign))
-                          && SizeAtCompileTime!=Dynamic && ((static_cast<int>(sizeof(Scalar))*SizeAtCompileTime)%16)==0 };
-    EIGEN_MAKE_ALIGNED_OPERATOR_NEW_IF(NeedsToAlign)
 
     using Base::base;
     using Base::coeff;

Eigen/src/Core/ArrayBase.h

@@ -42,7 +42,10 @@ template<typename ExpressionType> class MatrixWrapper;
   *
   * This class is the base that is inherited by all array expression types.
   *
-  * \param Derived is the derived type, e.g., an array or an expression type.
+  * \tparam Derived is the derived type, e.g., an array or an expression type.
+  *
+  * This class can be extended with the help of the plugin mechanism described on the page
+  * \ref TopicCustomizingEigen by defining the preprocessor symbol \c EIGEN_ARRAYBASE_PLUGIN.
   *
   * \sa class MatrixBase, \ref TopicClassHierarchy
   */

Eigen/src/Core/ArrayWrapper.h

@@ -53,6 +53,12 @@ class ArrayWrapper : public ArrayBase<ArrayWrapper<ExpressionType> >
     EIGEN_DENSE_PUBLIC_INTERFACE(ArrayWrapper)
     EIGEN_INHERIT_ASSIGNMENT_OPERATORS(ArrayWrapper)
 
+    typedef typename internal::conditional<
+                       internal::is_lvalue<ExpressionType>::value,
+                       Scalar,
+                       const Scalar
+                     >::type ScalarWithConstIfNotLvalue;
+
     typedef typename internal::nested<ExpressionType>::type NestedExpressionType;
 
     inline ArrayWrapper(const ExpressionType& matrix) : m_expression(matrix) {}
@@ -62,6 +68,9 @@ class ArrayWrapper : public ArrayBase<ArrayWrapper<ExpressionType> >
     inline Index outerStride() const { return m_expression.outerStride(); }
     inline Index innerStride() const { return m_expression.innerStride(); }
 
+    inline ScalarWithConstIfNotLvalue* data() { return m_expression.data(); }
+    inline const Scalar* data() const { return m_expression.data(); }
+
     inline const CoeffReturnType coeff(Index row, Index col) const
     {
       return m_expression.coeff(row, col);
@@ -119,6 +128,12 @@ class ArrayWrapper : public ArrayBase<ArrayWrapper<ExpressionType> >
     template<typename Dest>
     inline void evalTo(Dest& dst) const { dst = m_expression; }
 
+    const typename internal::remove_all<NestedExpressionType>::type& 
+    nestedExpression() const 
+    {
+      return m_expression;
+    }
+
   protected:
     const NestedExpressionType m_expression;
 };
@@ -151,6 +166,12 @@ class MatrixWrapper : public MatrixBase<MatrixWrapper<ExpressionType> >
     EIGEN_DENSE_PUBLIC_INTERFACE(MatrixWrapper)
     EIGEN_INHERIT_ASSIGNMENT_OPERATORS(MatrixWrapper)
 
+    typedef typename internal::conditional<
+                       internal::is_lvalue<ExpressionType>::value,
+                       Scalar,
+                       const Scalar
+                     >::type ScalarWithConstIfNotLvalue;
+
     typedef typename internal::nested<ExpressionType>::type NestedExpressionType;
 
     inline MatrixWrapper(const ExpressionType& matrix) : m_expression(matrix) {}
@@ -160,6 +181,9 @@ class MatrixWrapper : public MatrixBase<MatrixWrapper<ExpressionType> >
     inline Index outerStride() const { return m_expression.outerStride(); }
     inline Index innerStride() const { return m_expression.innerStride(); }
 
+    inline ScalarWithConstIfNotLvalue* data() { return m_expression.data(); }
+    inline const Scalar* data() const { return m_expression.data(); }
+
     inline const CoeffReturnType coeff(Index row, Index col) const
     {
       return m_expression.coeff(row, col);
@@ -214,6 +238,12 @@ class MatrixWrapper : public MatrixBase<MatrixWrapper<ExpressionType> >
       m_expression.const_cast_derived().template writePacket<LoadMode>(index, x);
     }
 
+    const typename internal::remove_all<NestedExpressionType>::type& 
+    nestedExpression() const 
+    {
+      return m_expression;
+    }
+
   protected:
     const NestedExpressionType m_expression;
 };

Eigen/src/Core/Assign.h

@@ -41,7 +41,7 @@ public:
     DstIsAligned = Derived::Flags & AlignedBit,
     DstHasDirectAccess = Derived::Flags & DirectAccessBit,
     SrcIsAligned = OtherDerived::Flags & AlignedBit,
-    JointAlignment = DstIsAligned && SrcIsAligned ? Aligned : Unaligned
+    JointAlignment = bool(DstIsAligned) && bool(SrcIsAligned) ? Aligned : Unaligned
   };
 
 private:
@@ -106,9 +106,9 @@ public:
                                              : int(NoUnrolling)
                   )
               : int(Traversal) == int(LinearVectorizedTraversal)
-                ? ( int(MayUnrollCompletely) && int(DstIsAligned) ? int(CompleteUnrolling) : int(NoUnrolling) )
+                ? ( bool(MayUnrollCompletely) && bool(DstIsAligned) ? int(CompleteUnrolling) : int(NoUnrolling) )
               : int(Traversal) == int(LinearTraversal)
-                ? ( int(MayUnrollCompletely) ? int(CompleteUnrolling) : int(NoUnrolling) )
+                ? ( bool(MayUnrollCompletely) ? int(CompleteUnrolling) : int(NoUnrolling) )
               : int(NoUnrolling)
   };
 
@@ -251,21 +251,22 @@ struct assign_innervec_InnerUnrolling<Derived1, Derived2, Stop, Stop>
 
 template<typename Derived1, typename Derived2,
          int Traversal = assign_traits<Derived1, Derived2>::Traversal,
-         int Unrolling = assign_traits<Derived1, Derived2>::Unrolling>
+         int Unrolling = assign_traits<Derived1, Derived2>::Unrolling,
+         int Version = Specialized>
 struct assign_impl;
 
 /************************
 *** Default traversal ***
 ************************/
 
-template<typename Derived1, typename Derived2, int Unrolling>
-struct assign_impl<Derived1, Derived2, InvalidTraversal, Unrolling>
+template<typename Derived1, typename Derived2, int Unrolling, int Version>
+struct assign_impl<Derived1, Derived2, InvalidTraversal, Unrolling, Version>
 {
   inline static void run(Derived1 &, const Derived2 &) { }
 };
 
-template<typename Derived1, typename Derived2>
-struct assign_impl<Derived1, Derived2, DefaultTraversal, NoUnrolling>
+template<typename Derived1, typename Derived2, int Version>
+struct assign_impl<Derived1, Derived2, DefaultTraversal, NoUnrolling, Version>
 {
   typedef typename Derived1::Index Index;
   inline static void run(Derived1 &dst, const Derived2 &src)
@@ -278,8 +279,8 @@ struct assign_impl<Derived1, Derived2, DefaultTraversal, NoUnrolling>
   }
 };
 
-template<typename Derived1, typename Derived2>
-struct assign_impl<Derived1, Derived2, DefaultTraversal, CompleteUnrolling>
+template<typename Derived1, typename Derived2, int Version>
+struct assign_impl<Derived1, Derived2, DefaultTraversal, CompleteUnrolling, Version>
 {
   EIGEN_STRONG_INLINE static void run(Derived1 &dst, const Derived2 &src)
   {
@@ -288,8 +289,8 @@ struct assign_impl<Derived1, Derived2, DefaultTraversal, CompleteUnrolling>
   }
 };
 
-template<typename Derived1, typename Derived2>
-struct assign_impl<Derived1, Derived2, DefaultTraversal, InnerUnrolling>
+template<typename Derived1, typename Derived2, int Version>
+struct assign_impl<Derived1, Derived2, DefaultTraversal, InnerUnrolling, Version>
 {
   typedef typename Derived1::Index Index;
   EIGEN_STRONG_INLINE static void run(Derived1 &dst, const Derived2 &src)
@@ -305,8 +306,8 @@ struct assign_impl<Derived1, Derived2, DefaultTraversal, InnerUnrolling>
 *** Linear traversal ***
 ***********************/
 
-template<typename Derived1, typename Derived2>
-struct assign_impl<Derived1, Derived2, LinearTraversal, NoUnrolling>
+template<typename Derived1, typename Derived2, int Version>
+struct assign_impl<Derived1, Derived2, LinearTraversal, NoUnrolling, Version>
 {
   typedef typename Derived1::Index Index;
   inline static void run(Derived1 &dst, const Derived2 &src)
@@ -317,8 +318,8 @@ struct assign_impl<Derived1, Derived2, LinearTraversal, NoUnrolling>
   }
 };
 
-template<typename Derived1, typename Derived2>
-struct assign_impl<Derived1, Derived2, LinearTraversal, CompleteUnrolling>
+template<typename Derived1, typename Derived2, int Version>
+struct assign_impl<Derived1, Derived2, LinearTraversal, CompleteUnrolling, Version>
 {
   EIGEN_STRONG_INLINE static void run(Derived1 &dst, const Derived2 &src)
   {
@@ -331,8 +332,8 @@ struct assign_impl<Derived1, Derived2, LinearTraversal, CompleteUnrolling>
 *** Inner vectorization ***
 **************************/
 
-template<typename Derived1, typename Derived2>
-struct assign_impl<Derived1, Derived2, InnerVectorizedTraversal, NoUnrolling>
+template<typename Derived1, typename Derived2, int Version>
+struct assign_impl<Derived1, Derived2, InnerVectorizedTraversal, NoUnrolling, Version>
 {
   typedef typename Derived1::Index Index;
   inline static void run(Derived1 &dst, const Derived2 &src)
@@ -346,8 +347,8 @@ struct assign_impl<Derived1, Derived2, InnerVectorizedTraversal, NoUnrolling>
   }
 };
 
-template<typename Derived1, typename Derived2>
-struct assign_impl<Derived1, Derived2, InnerVectorizedTraversal, CompleteUnrolling>
+template<typename Derived1, typename Derived2, int Version>
+struct assign_impl<Derived1, Derived2, InnerVectorizedTraversal, CompleteUnrolling, Version>
 {
   EIGEN_STRONG_INLINE static void run(Derived1 &dst, const Derived2 &src)
   {
@@ -356,8 +357,8 @@ struct assign_impl<Derived1, Derived2, InnerVectorizedTraversal, CompleteUnrolli
   }
 };
 
-template<typename Derived1, typename Derived2>
-struct assign_impl<Derived1, Derived2, InnerVectorizedTraversal, InnerUnrolling>
+template<typename Derived1, typename Derived2, int Version>
+struct assign_impl<Derived1, Derived2, InnerVectorizedTraversal, InnerUnrolling, Version>
 {
   typedef typename Derived1::Index Index;
   EIGEN_STRONG_INLINE static void run(Derived1 &dst, const Derived2 &src)
@@ -398,8 +399,8 @@ struct unaligned_assign_impl<false>
   }
 };
 
-template<typename Derived1, typename Derived2>
-struct assign_impl<Derived1, Derived2, LinearVectorizedTraversal, NoUnrolling>
+template<typename Derived1, typename Derived2, int Version>
+struct assign_impl<Derived1, Derived2, LinearVectorizedTraversal, NoUnrolling, Version>
 {
   typedef typename Derived1::Index Index;
   EIGEN_STRONG_INLINE static void run(Derived1 &dst, const Derived2 &src)
@@ -426,8 +427,8 @@ struct assign_impl<Derived1, Derived2, LinearVectorizedTraversal, NoUnrolling>
   }
 };
 
-template<typename Derived1, typename Derived2>
-struct assign_impl<Derived1, Derived2, LinearVectorizedTraversal, CompleteUnrolling>
+template<typename Derived1, typename Derived2, int Version>
+struct assign_impl<Derived1, Derived2, LinearVectorizedTraversal, CompleteUnrolling, Version>
 {
   typedef typename Derived1::Index Index;
   EIGEN_STRONG_INLINE static void run(Derived1 &dst, const Derived2 &src)
@@ -445,8 +446,8 @@ struct assign_impl<Derived1, Derived2, LinearVectorizedTraversal, CompleteUnroll
 *** Slice vectorization ***
 ***************************/
 
-template<typename Derived1, typename Derived2>
-struct assign_impl<Derived1, Derived2, SliceVectorizedTraversal, NoUnrolling>
+template<typename Derived1, typename Derived2, int Version>
+struct assign_impl<Derived1, Derived2, SliceVectorizedTraversal, NoUnrolling, Version>
 {
   typedef typename Derived1::Index Index;
   inline static void run(Derived1 &dst, const Derived2 &src)
@@ -474,7 +475,7 @@ struct assign_impl<Derived1, Derived2, SliceVectorizedTraversal, NoUnrolling>
 
       // do the vectorizable part of the assignment
       for(Index inner = alignedStart; inner<alignedEnd; inner+=packetSize)
-        dst.template copyPacketByOuterInner<Derived2, Aligned, Unaligned>(outer, inner, src);
+        dst.template copyPacketByOuterInner<Derived2, dstAlignment, Unaligned>(outer, inner, src);
 
       // do the non-vectorizable part of the assignment
       for(Index inner = alignedEnd; inner<innerSize ; ++inner)

Eigen/src/Core/AssignEvaluator.h

@@ -0,0 +1,682 @@
+// This file is part of Eigen, a lightweight C++ template library
+// for linear algebra.
+//
+// Copyright (C) 2011 Benoit Jacob <jacob.benoit.1@gmail.com>
+// Copyright (C) 2011 Gael Guennebaud <gael.guennebaud@inria.fr>
+// Copyright (C) 2011 Jitse Niesen <jitse@maths.leeds.ac.uk>
+//
+// Eigen is free software; you can redistribute it and/or
+// modify it under the terms of the GNU Lesser General Public
+// License as published by the Free Software Foundation; either
+// version 3 of the License, or (at your option) any later version.
+//
+// Alternatively, you can redistribute it and/or
+// modify it under the terms of the GNU General Public License as
+// published by the Free Software Foundation; either version 2 of
+// the License, or (at your option) any later version.
+//
+// Eigen is distributed in the hope that it will be useful, but WITHOUT ANY
+// WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
+// FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License or the
+// GNU General Public License for more details.
+//
+// You should have received a copy of the GNU Lesser General Public
+// License and a copy of the GNU General Public License along with
+// Eigen. If not, see <http://www.gnu.org/licenses/>.
+
+#ifndef EIGEN_ASSIGN_EVALUATOR_H
+#define EIGEN_ASSIGN_EVALUATOR_H
+
+// This implementation is based on Assign.h
+
+namespace internal {
+  
+/***************************************************************************
+* Part 1 : the logic deciding a strategy for traversal and unrolling       *
+***************************************************************************/
+
+// copy_using_evaluator_traits is based on assign_traits
+// (actually, it's identical)
+
+template <typename Derived, typename OtherDerived>
+struct copy_using_evaluator_traits
+{
+public:
+  enum {
+    DstIsAligned = Derived::Flags & AlignedBit,
+    DstHasDirectAccess = Derived::Flags & DirectAccessBit,
+    SrcIsAligned = OtherDerived::Flags & AlignedBit,
+    JointAlignment = bool(DstIsAligned) && bool(SrcIsAligned) ? Aligned : Unaligned
+  };
+
+private:
+  enum {
+    InnerSize = int(Derived::IsVectorAtCompileTime) ? int(Derived::SizeAtCompileTime)
+              : int(Derived::Flags)&RowMajorBit ? int(Derived::ColsAtCompileTime)
+              : int(Derived::RowsAtCompileTime),
+    InnerMaxSize = int(Derived::IsVectorAtCompileTime) ? int(Derived::MaxSizeAtCompileTime)
+              : int(Derived::Flags)&RowMajorBit ? int(Derived::MaxColsAtCompileTime)
+              : int(Derived::MaxRowsAtCompileTime),
+    MaxSizeAtCompileTime = Derived::SizeAtCompileTime,
+    PacketSize = packet_traits<typename Derived::Scalar>::size
+  };
+
+  enum {
+    StorageOrdersAgree = (int(Derived::IsRowMajor) == int(OtherDerived::IsRowMajor)),
+    MightVectorize = StorageOrdersAgree
+                  && (int(Derived::Flags) & int(OtherDerived::Flags) & ActualPacketAccessBit),
+    MayInnerVectorize  = MightVectorize && int(InnerSize)!=Dynamic && int(InnerSize)%int(PacketSize)==0
+                       && int(DstIsAligned) && int(SrcIsAligned),
+    MayLinearize = StorageOrdersAgree && (int(Derived::Flags) & int(OtherDerived::Flags) & LinearAccessBit),
+    MayLinearVectorize = MightVectorize && MayLinearize && DstHasDirectAccess
+                       && (DstIsAligned || MaxSizeAtCompileTime == Dynamic),
+      /* If the destination isn't aligned, we have to do runtime checks and we don't unroll,
+         so it's only good for large enough sizes. */
+    MaySliceVectorize  = MightVectorize && DstHasDirectAccess
+                       && (int(InnerMaxSize)==Dynamic || int(InnerMaxSize)>=3*PacketSize)
+      /* slice vectorization can be slow, so we only want it if the slices are big, which is
+         indicated by InnerMaxSize rather than InnerSize, think of the case of a dynamic block
+         in a fixed-size matrix */
+  };
+
+public:
+  enum {
+    Traversal = int(MayInnerVectorize)  ? int(InnerVectorizedTraversal)
+              : int(MayLinearVectorize) ? int(LinearVectorizedTraversal)
+              : int(MaySliceVectorize)  ? int(SliceVectorizedTraversal)
+              : int(MayLinearize)       ? int(LinearTraversal)
+                                        : int(DefaultTraversal),
+    Vectorized = int(Traversal) == InnerVectorizedTraversal
+              || int(Traversal) == LinearVectorizedTraversal
+              || int(Traversal) == SliceVectorizedTraversal
+  };
+
+private:
+  enum {
+    UnrollingLimit      = EIGEN_UNROLLING_LIMIT * (Vectorized ? int(PacketSize) : 1),
+    MayUnrollCompletely = int(Derived::SizeAtCompileTime) != Dynamic
+                       && int(OtherDerived::CoeffReadCost) != Dynamic
+                       && int(Derived::SizeAtCompileTime) * int(OtherDerived::CoeffReadCost) <= int(UnrollingLimit),
+    MayUnrollInner      = int(InnerSize) != Dynamic
+                       && int(OtherDerived::CoeffReadCost) != Dynamic
+                       && int(InnerSize) * int(OtherDerived::CoeffReadCost) <= int(UnrollingLimit)
+  };
+
+public:
+  enum {
+    Unrolling = (int(Traversal) == int(InnerVectorizedTraversal) || int(Traversal) == int(DefaultTraversal))
+                ? (
+ 		    int(MayUnrollCompletely) ? int(CompleteUnrolling)
+                  : int(MayUnrollInner)      ? int(InnerUnrolling)
+                                             : int(NoUnrolling)
+                  )
+              : int(Traversal) == int(LinearVectorizedTraversal)
+                ? ( bool(MayUnrollCompletely) && bool(DstIsAligned) ? int(CompleteUnrolling) 
+                                                                    : int(NoUnrolling) )
+              : int(Traversal) == int(LinearTraversal)
+                ? ( bool(MayUnrollCompletely) ? int(CompleteUnrolling) 
+                                              : int(NoUnrolling) )
+              : int(NoUnrolling)
+  };
+
+#ifdef EIGEN_DEBUG_ASSIGN
+  static void debug()
+  {
+    EIGEN_DEBUG_VAR(DstIsAligned)
+    EIGEN_DEBUG_VAR(SrcIsAligned)
+    EIGEN_DEBUG_VAR(JointAlignment)
+    EIGEN_DEBUG_VAR(InnerSize)
+    EIGEN_DEBUG_VAR(InnerMaxSize)
+    EIGEN_DEBUG_VAR(PacketSize)
+    EIGEN_DEBUG_VAR(StorageOrdersAgree)
+    EIGEN_DEBUG_VAR(MightVectorize)
+    EIGEN_DEBUG_VAR(MayLinearize)
+    EIGEN_DEBUG_VAR(MayInnerVectorize)
+    EIGEN_DEBUG_VAR(MayLinearVectorize)
+    EIGEN_DEBUG_VAR(MaySliceVectorize)
+    EIGEN_DEBUG_VAR(Traversal)
+    EIGEN_DEBUG_VAR(UnrollingLimit)
+    EIGEN_DEBUG_VAR(MayUnrollCompletely)
+    EIGEN_DEBUG_VAR(MayUnrollInner)
+    EIGEN_DEBUG_VAR(Unrolling)
+  }
+#endif
+};
+
+/***************************************************************************
+* Part 2 : meta-unrollers
+***************************************************************************/
+
+/************************
+*** Default traversal ***
+************************/
+
+template<typename DstEvaluatorType, typename SrcEvaluatorType, int Index, int Stop>
+struct copy_using_evaluator_DefaultTraversal_CompleteUnrolling
+{
+  typedef typename DstEvaluatorType::XprType DstXprType;
+  
+  enum {
+    outer = Index / DstXprType::InnerSizeAtCompileTime,
+    inner = Index % DstXprType::InnerSizeAtCompileTime
+  };
+
+  EIGEN_STRONG_INLINE static void run(DstEvaluatorType &dstEvaluator, 
+				      SrcEvaluatorType &srcEvaluator)
+  {
+    dstEvaluator.copyCoeffByOuterInner(outer, inner, srcEvaluator);
+    copy_using_evaluator_DefaultTraversal_CompleteUnrolling
+      <DstEvaluatorType, SrcEvaluatorType, Index+1, Stop>
+      ::run(dstEvaluator, srcEvaluator);
+  }
+};
+
+template<typename DstEvaluatorType, typename SrcEvaluatorType, int Stop>
+struct copy_using_evaluator_DefaultTraversal_CompleteUnrolling<DstEvaluatorType, SrcEvaluatorType, Stop, Stop>
+{
+  EIGEN_STRONG_INLINE static void run(DstEvaluatorType&, SrcEvaluatorType&) { }
+};
+
+template<typename DstEvaluatorType, typename SrcEvaluatorType, int Index, int Stop>
+struct copy_using_evaluator_DefaultTraversal_InnerUnrolling
+{
+  EIGEN_STRONG_INLINE static void run(DstEvaluatorType &dstEvaluator, 
+				      SrcEvaluatorType &srcEvaluator, 
+				      int outer)
+  {
+    dstEvaluator.copyCoeffByOuterInner(outer, Index, srcEvaluator);
+    copy_using_evaluator_DefaultTraversal_InnerUnrolling
+      <DstEvaluatorType, SrcEvaluatorType, Index+1, Stop>
+      ::run(dstEvaluator, srcEvaluator, outer);
+  }
+};
+
+template<typename DstEvaluatorType, typename SrcEvaluatorType, int Stop>
+struct copy_using_evaluator_DefaultTraversal_InnerUnrolling<DstEvaluatorType, SrcEvaluatorType, Stop, Stop>
+{
+  EIGEN_STRONG_INLINE static void run(DstEvaluatorType&, SrcEvaluatorType&, int) { }
+};
+
+/***********************
+*** Linear traversal ***
+***********************/
+
+template<typename DstEvaluatorType, typename SrcEvaluatorType, int Index, int Stop>
+struct copy_using_evaluator_LinearTraversal_CompleteUnrolling
+{
+  EIGEN_STRONG_INLINE static void run(DstEvaluatorType &dstEvaluator, 
+				      SrcEvaluatorType &srcEvaluator)
+  {
+    dstEvaluator.copyCoeff(Index, srcEvaluator);
+    copy_using_evaluator_LinearTraversal_CompleteUnrolling
+      <DstEvaluatorType, SrcEvaluatorType, Index+1, Stop>
+      ::run(dstEvaluator, srcEvaluator);
+  }
+};
+
+template<typename DstEvaluatorType, typename SrcEvaluatorType, int Stop>
+struct copy_using_evaluator_LinearTraversal_CompleteUnrolling<DstEvaluatorType, SrcEvaluatorType, Stop, Stop>
+{
+  EIGEN_STRONG_INLINE static void run(DstEvaluatorType&, SrcEvaluatorType&) { }
+};
+
+/**************************
+*** Inner vectorization ***
+**************************/
+
+template<typename DstEvaluatorType, typename SrcEvaluatorType, int Index, int Stop>
+struct copy_using_evaluator_innervec_CompleteUnrolling
+{
+  typedef typename DstEvaluatorType::XprType DstXprType;
+  typedef typename SrcEvaluatorType::XprType SrcXprType;
+
+  enum {
+    outer = Index / DstXprType::InnerSizeAtCompileTime,
+    inner = Index % DstXprType::InnerSizeAtCompileTime,
+    JointAlignment = copy_using_evaluator_traits<DstXprType,SrcXprType>::JointAlignment
+  };
+
+  EIGEN_STRONG_INLINE static void run(DstEvaluatorType &dstEvaluator, 
+				      SrcEvaluatorType &srcEvaluator)
+  {
+    dstEvaluator.template copyPacketByOuterInner<Aligned, JointAlignment>(outer, inner, srcEvaluator);
+    enum { NextIndex = Index + packet_traits<typename DstXprType::Scalar>::size };
+    copy_using_evaluator_innervec_CompleteUnrolling
+      <DstEvaluatorType, SrcEvaluatorType, NextIndex, Stop>
+      ::run(dstEvaluator, srcEvaluator);
+  }
+};
+
+template<typename DstEvaluatorType, typename SrcEvaluatorType, int Stop>
+struct copy_using_evaluator_innervec_CompleteUnrolling<DstEvaluatorType, SrcEvaluatorType, Stop, Stop>
+{
+  EIGEN_STRONG_INLINE static void run(DstEvaluatorType&, SrcEvaluatorType&) { }
+};
+
+template<typename DstEvaluatorType, typename SrcEvaluatorType, int Index, int Stop>
+struct copy_using_evaluator_innervec_InnerUnrolling
+{
+  EIGEN_STRONG_INLINE static void run(DstEvaluatorType &dstEvaluator, 
+				      SrcEvaluatorType &srcEvaluator, 
+				      int outer)
+  {
+    dstEvaluator.template copyPacketByOuterInner<Aligned, Aligned>(outer, Index, srcEvaluator);
+    typedef typename DstEvaluatorType::XprType DstXprType;
+    enum { NextIndex = Index + packet_traits<typename DstXprType::Scalar>::size };
+    copy_using_evaluator_innervec_InnerUnrolling
+      <DstEvaluatorType, SrcEvaluatorType, NextIndex, Stop>
+      ::run(dstEvaluator, srcEvaluator, outer);
+  }
+};
+
+template<typename DstEvaluatorType, typename SrcEvaluatorType, int Stop>
+struct copy_using_evaluator_innervec_InnerUnrolling<DstEvaluatorType, SrcEvaluatorType, Stop, Stop>
+{
+  EIGEN_STRONG_INLINE static void run(DstEvaluatorType&, SrcEvaluatorType&, int) { }
+};
+
+/***************************************************************************
+* Part 3 : implementation of all cases
+***************************************************************************/
+
+// copy_using_evaluator_impl is based on assign_impl
+
+template<typename DstXprType, typename SrcXprType,
+         int Traversal = copy_using_evaluator_traits<DstXprType, SrcXprType>::Traversal,
+         int Unrolling = copy_using_evaluator_traits<DstXprType, SrcXprType>::Unrolling>
+struct copy_using_evaluator_impl;
+
+/************************
+*** Default traversal ***
+************************/
+
+template<typename DstXprType, typename SrcXprType>
+struct copy_using_evaluator_impl<DstXprType, SrcXprType, DefaultTraversal, NoUnrolling>
+{
+  static void run(DstXprType& dst, const SrcXprType& src)
+  {
+    typedef typename evaluator<DstXprType>::type DstEvaluatorType;
+    typedef typename evaluator<SrcXprType>::type SrcEvaluatorType;
+    typedef typename DstXprType::Index Index;
+
+    DstEvaluatorType dstEvaluator(dst);
+    SrcEvaluatorType srcEvaluator(src);
+
+    for(Index outer = 0; outer < dst.outerSize(); ++outer) {
+      for(Index inner = 0; inner < dst.innerSize(); ++inner) {
+	dstEvaluator.copyCoeffByOuterInner(outer, inner, srcEvaluator);
+      }
+    }
+  }
+};
+
+template<typename DstXprType, typename SrcXprType>
+struct copy_using_evaluator_impl<DstXprType, SrcXprType, DefaultTraversal, CompleteUnrolling>
+{
+  EIGEN_STRONG_INLINE static void run(DstXprType &dst, const SrcXprType &src)
+  {
+    typedef typename evaluator<DstXprType>::type DstEvaluatorType;
+    typedef typename evaluator<SrcXprType>::type SrcEvaluatorType;
+
+    DstEvaluatorType dstEvaluator(dst);
+    SrcEvaluatorType srcEvaluator(src);
+
+    copy_using_evaluator_DefaultTraversal_CompleteUnrolling
+      <DstEvaluatorType, SrcEvaluatorType, 0, DstXprType::SizeAtCompileTime>
+      ::run(dstEvaluator, srcEvaluator);
+  }
+};
+
+template<typename DstXprType, typename SrcXprType>
+struct copy_using_evaluator_impl<DstXprType, SrcXprType, DefaultTraversal, InnerUnrolling>
+{
+  typedef typename DstXprType::Index Index;
+  EIGEN_STRONG_INLINE static void run(DstXprType &dst, const SrcXprType &src)
+  {
+    typedef typename evaluator<DstXprType>::type DstEvaluatorType;
+    typedef typename evaluator<SrcXprType>::type SrcEvaluatorType;
+
+    DstEvaluatorType dstEvaluator(dst);
+    SrcEvaluatorType srcEvaluator(src);
+
+    const Index outerSize = dst.outerSize();
+    for(Index outer = 0; outer < outerSize; ++outer)
+      copy_using_evaluator_DefaultTraversal_InnerUnrolling
+	<DstEvaluatorType, SrcEvaluatorType, 0, DstXprType::InnerSizeAtCompileTime>
+        ::run(dstEvaluator, srcEvaluator, outer);
+  }
+};
+
+/***************************
+*** Linear vectorization ***
+***************************/
+
+template <bool IsAligned = false>
+struct unaligned_copy_using_evaluator_impl
+{
+  // if IsAligned = true, then do nothing
+  template <typename SrcEvaluatorType, typename DstEvaluatorType>
+  static EIGEN_STRONG_INLINE void run(const SrcEvaluatorType&, DstEvaluatorType&, 
+				      typename SrcEvaluatorType::Index, typename SrcEvaluatorType::Index) {}
+};
+
+template <>
+struct unaligned_copy_using_evaluator_impl<false>
+{
+  // MSVC must not inline this functions. If it does, it fails to optimize the
+  // packet access path.
+#ifdef _MSC_VER
+  template <typename DstEvaluatorType, typename SrcEvaluatorType>
+  static EIGEN_DONT_INLINE void run(DstEvaluatorType &dstEvaluator, 
+				    const SrcEvaluatorType &srcEvaluator, 
+				    typename DstEvaluatorType::Index start, 
+				    typename DstEvaluatorType::Index end)
+#else
+  template <typename DstEvaluatorType, typename SrcEvaluatorType>
+  static EIGEN_STRONG_INLINE void run(DstEvaluatorType &dstEvaluator, 
+				      const SrcEvaluatorType &srcEvaluator, 
+				      typename DstEvaluatorType::Index start, 
+				      typename DstEvaluatorType::Index end)
+#endif
+  {
+    for (typename DstEvaluatorType::Index index = start; index < end; ++index)
+      dstEvaluator.copyCoeff(index, srcEvaluator);
+  }
+};
+
+template<typename DstXprType, typename SrcXprType>
+struct copy_using_evaluator_impl<DstXprType, SrcXprType, LinearVectorizedTraversal, NoUnrolling>
+{
+  EIGEN_STRONG_INLINE static void run(DstXprType &dst, const SrcXprType &src)
+  {
+    typedef typename evaluator<DstXprType>::type DstEvaluatorType;
+    typedef typename evaluator<SrcXprType>::type SrcEvaluatorType;
+    typedef typename DstXprType::Index Index;
+
+    DstEvaluatorType dstEvaluator(dst);
+    SrcEvaluatorType srcEvaluator(src);
+
+    const Index size = dst.size();
+    typedef packet_traits<typename DstXprType::Scalar> PacketTraits;
+    enum {
+      packetSize = PacketTraits::size,
+      dstIsAligned = int(copy_using_evaluator_traits<DstXprType,SrcXprType>::DstIsAligned),
+      dstAlignment = PacketTraits::AlignedOnScalar ? Aligned : dstIsAligned,
+      srcAlignment = copy_using_evaluator_traits<DstXprType,SrcXprType>::JointAlignment
+    };
+    const Index alignedStart = dstIsAligned ? 0 : first_aligned(&dstEvaluator.coeffRef(0), size);
+    const Index alignedEnd = alignedStart + ((size-alignedStart)/packetSize)*packetSize;
+
+    unaligned_copy_using_evaluator_impl<dstIsAligned!=0>::run(dstEvaluator, srcEvaluator, 0, alignedStart);
+
+    for(Index index = alignedStart; index < alignedEnd; index += packetSize)
+    {
+      dstEvaluator.template copyPacket<dstAlignment, srcAlignment>(index, srcEvaluator);
+    }
+
+    unaligned_copy_using_evaluator_impl<>::run(dstEvaluator, srcEvaluator, alignedEnd, size);
+  }
+};
+
+template<typename DstXprType, typename SrcXprType>
+struct copy_using_evaluator_impl<DstXprType, SrcXprType, LinearVectorizedTraversal, CompleteUnrolling>
+{
+  typedef typename DstXprType::Index Index;
+  EIGEN_STRONG_INLINE static void run(DstXprType &dst, const SrcXprType &src)
+  {
+    typedef typename evaluator<DstXprType>::type DstEvaluatorType;
+    typedef typename evaluator<SrcXprType>::type SrcEvaluatorType;
+
+    DstEvaluatorType dstEvaluator(dst);
+    SrcEvaluatorType srcEvaluator(src);
+
+    enum { size = DstXprType::SizeAtCompileTime,
+           packetSize = packet_traits<typename DstXprType::Scalar>::size,
+           alignedSize = (size/packetSize)*packetSize };
+
+    copy_using_evaluator_innervec_CompleteUnrolling
+      <DstEvaluatorType, SrcEvaluatorType, 0, alignedSize>
+      ::run(dstEvaluator, srcEvaluator);
+    copy_using_evaluator_DefaultTraversal_CompleteUnrolling
+      <DstEvaluatorType, SrcEvaluatorType, alignedSize, size>
+      ::run(dstEvaluator, srcEvaluator);
+  }
+};
+
+/**************************
+*** Inner vectorization ***
+**************************/
+
+template<typename DstXprType, typename SrcXprType>
+struct copy_using_evaluator_impl<DstXprType, SrcXprType, InnerVectorizedTraversal, NoUnrolling>
+{
+  inline static void run(DstXprType &dst, const SrcXprType &src)
+  {
+    typedef typename evaluator<DstXprType>::type DstEvaluatorType;
+    typedef typename evaluator<SrcXprType>::type SrcEvaluatorType;
+    typedef typename DstXprType::Index Index;
+
+    DstEvaluatorType dstEvaluator(dst);
+    SrcEvaluatorType srcEvaluator(src);
+
+    const Index innerSize = dst.innerSize();
+    const Index outerSize = dst.outerSize();
+    const Index packetSize = packet_traits<typename DstXprType::Scalar>::size;
+    for(Index outer = 0; outer < outerSize; ++outer)
+      for(Index inner = 0; inner < innerSize; inner+=packetSize) {
+	dstEvaluator.template copyPacketByOuterInner<Aligned, Aligned>(outer, inner, srcEvaluator);
+      }
+  }
+};
+
+template<typename DstXprType, typename SrcXprType>
+struct copy_using_evaluator_impl<DstXprType, SrcXprType, InnerVectorizedTraversal, CompleteUnrolling>
+{
+  EIGEN_STRONG_INLINE static void run(DstXprType &dst, const SrcXprType &src)
+  {
+    typedef typename evaluator<DstXprType>::type DstEvaluatorType;
+    typedef typename evaluator<SrcXprType>::type SrcEvaluatorType;
+
+    DstEvaluatorType dstEvaluator(dst);
+    SrcEvaluatorType srcEvaluator(src);
+
+    copy_using_evaluator_innervec_CompleteUnrolling
+      <DstEvaluatorType, SrcEvaluatorType, 0, DstXprType::SizeAtCompileTime>
+      ::run(dstEvaluator, srcEvaluator);
+  }
+};
+
+template<typename DstXprType, typename SrcXprType>
+struct copy_using_evaluator_impl<DstXprType, SrcXprType, InnerVectorizedTraversal, InnerUnrolling>
+{
+  typedef typename DstXprType::Index Index;
+  EIGEN_STRONG_INLINE static void run(DstXprType &dst, const SrcXprType &src)
+  {
+    typedef typename evaluator<DstXprType>::type DstEvaluatorType;
+    typedef typename evaluator<SrcXprType>::type SrcEvaluatorType;
+
+    DstEvaluatorType dstEvaluator(dst);
+    SrcEvaluatorType srcEvaluator(src);
+
+    const Index outerSize = dst.outerSize();
+    for(Index outer = 0; outer < outerSize; ++outer)
+      copy_using_evaluator_innervec_InnerUnrolling
+	<DstEvaluatorType, SrcEvaluatorType, 0, DstXprType::InnerSizeAtCompileTime>
+        ::run(dstEvaluator, srcEvaluator, outer);
+  }
+};
+
+/***********************
+*** Linear traversal ***
+***********************/
+
+template<typename DstXprType, typename SrcXprType>
+struct copy_using_evaluator_impl<DstXprType, SrcXprType, LinearTraversal, NoUnrolling>
+{
+  inline static void run(DstXprType &dst, const SrcXprType &src)
+  {
+    typedef typename evaluator<DstXprType>::type DstEvaluatorType;
+    typedef typename evaluator<SrcXprType>::type SrcEvaluatorType;
+    typedef typename DstXprType::Index Index;
+
+    DstEvaluatorType dstEvaluator(dst);
+    SrcEvaluatorType srcEvaluator(src);
+
+    const Index size = dst.size();
+    for(Index i = 0; i < size; ++i)
+      dstEvaluator.copyCoeff(i, srcEvaluator);
+  }
+};
+
+template<typename DstXprType, typename SrcXprType>
+struct copy_using_evaluator_impl<DstXprType, SrcXprType, LinearTraversal, CompleteUnrolling>
+{
+  EIGEN_STRONG_INLINE static void run(DstXprType &dst, const SrcXprType &src)
+  {
+    typedef typename evaluator<DstXprType>::type DstEvaluatorType;
+    typedef typename evaluator<SrcXprType>::type SrcEvaluatorType;
+
+    DstEvaluatorType dstEvaluator(dst);
+    SrcEvaluatorType srcEvaluator(src);
+
+    copy_using_evaluator_LinearTraversal_CompleteUnrolling
+      <DstEvaluatorType, SrcEvaluatorType, 0, DstXprType::SizeAtCompileTime>
+      ::run(dstEvaluator, srcEvaluator);
+  }
+};
+
+/**************************
+*** Slice vectorization ***
+***************************/
+
+template<typename DstXprType, typename SrcXprType>
+struct copy_using_evaluator_impl<DstXprType, SrcXprType, SliceVectorizedTraversal, NoUnrolling>
+{
+  inline static void run(DstXprType &dst, const SrcXprType &src)
+  {
+    typedef typename evaluator<DstXprType>::type DstEvaluatorType;
+    typedef typename evaluator<SrcXprType>::type SrcEvaluatorType;
+    typedef typename DstXprType::Index Index;
+
+    DstEvaluatorType dstEvaluator(dst);
+    SrcEvaluatorType srcEvaluator(src);
+
+    typedef packet_traits<typename DstXprType::Scalar> PacketTraits;
+    enum {
+      packetSize = PacketTraits::size,
+      alignable = PacketTraits::AlignedOnScalar,
+      dstAlignment = alignable ? Aligned : int(copy_using_evaluator_traits<DstXprType,SrcXprType>::DstIsAligned) ,
+      srcAlignment = copy_using_evaluator_traits<DstXprType,SrcXprType>::JointAlignment
+    };
+    const Index packetAlignedMask = packetSize - 1;
+    const Index innerSize = dst.innerSize();
+    const Index outerSize = dst.outerSize();
+    const Index alignedStep = alignable ? (packetSize - dst.outerStride() % packetSize) & packetAlignedMask : 0;
+    Index alignedStart = ((!alignable) || copy_using_evaluator_traits<DstXprType,SrcXprType>::DstIsAligned) ? 0
+                       : first_aligned(&dstEvaluator.coeffRef(0,0), innerSize);
+
+    for(Index outer = 0; outer < outerSize; ++outer)
+    {
+      const Index alignedEnd = alignedStart + ((innerSize-alignedStart) & ~packetAlignedMask);
+      // do the non-vectorizable part of the assignment
+      for(Index inner = 0; inner<alignedStart ; ++inner) {
+        dstEvaluator.copyCoeffByOuterInner(outer, inner, srcEvaluator);
+      }
+
+      // do the vectorizable part of the assignment
+      for(Index inner = alignedStart; inner<alignedEnd; inner+=packetSize) {
+        dstEvaluator.template copyPacketByOuterInner<dstAlignment, srcAlignment>(outer, inner, srcEvaluator);
+      }
+
+      // do the non-vectorizable part of the assignment
+      for(Index inner = alignedEnd; inner<innerSize ; ++inner) {
+        dstEvaluator.copyCoeffByOuterInner(outer, inner, srcEvaluator);
+      }
+
+      alignedStart = std::min<Index>((alignedStart+alignedStep)%packetSize, innerSize);
+    }
+  }
+};
+
+/***************************************************************************
+* Part 4 : Entry points
+***************************************************************************/
+
+// Based on DenseBase::LazyAssign()
+
+template<typename DstXprType, typename SrcXprType>
+const DstXprType& copy_using_evaluator(const DstXprType& dst, const SrcXprType& src)
+{
+#ifdef EIGEN_DEBUG_ASSIGN
+  internal::copy_using_evaluator_traits<DstXprType, SrcXprType>::debug();
+#endif
+  copy_using_evaluator_impl<DstXprType, SrcXprType>::run(const_cast<DstXprType&>(dst), src);
+  return dst;
+}
+
+// Based on DenseBase::swap()
+// TODO: Chech whether we need to do something special for swapping two
+//       Arrays or Matrices.
+
+template<typename DstXprType, typename SrcXprType>
+void swap_using_evaluator(const DstXprType& dst, const SrcXprType& src)
+{
+  copy_using_evaluator(SwapWrapper<DstXprType>(const_cast<DstXprType&>(dst)), src);
+}
+
+// Based on MatrixBase::operator+= (in CwiseBinaryOp.h)
+template<typename DstXprType, typename SrcXprType>
+void add_assign_using_evaluator(const MatrixBase<DstXprType>& dst, const MatrixBase<SrcXprType>& src)
+{
+  typedef typename DstXprType::Scalar Scalar;
+  SelfCwiseBinaryOp<internal::scalar_sum_op<Scalar>, DstXprType, SrcXprType> tmp(dst.const_cast_derived());
+  copy_using_evaluator(tmp, src.derived());
+}
+
+// Based on ArrayBase::operator+=
+template<typename DstXprType, typename SrcXprType>
+void add_assign_using_evaluator(const ArrayBase<DstXprType>& dst, const ArrayBase<SrcXprType>& src)
+{
+  typedef typename DstXprType::Scalar Scalar;
+  SelfCwiseBinaryOp<internal::scalar_sum_op<Scalar>, DstXprType, SrcXprType> tmp(dst.const_cast_derived());
+  copy_using_evaluator(tmp, src.derived());
+}
+
+// TODO: Add add_assign_using_evaluator for EigenBase ?
+
+template<typename DstXprType, typename SrcXprType>
+void subtract_assign_using_evaluator(const MatrixBase<DstXprType>& dst, const MatrixBase<SrcXprType>& src)
+{
+  typedef typename DstXprType::Scalar Scalar;
+  SelfCwiseBinaryOp<internal::scalar_difference_op<Scalar>, DstXprType, SrcXprType> tmp(dst.const_cast_derived());
+  copy_using_evaluator(tmp, src.derived());
+}
+
+template<typename DstXprType, typename SrcXprType>
+void subtract_assign_using_evaluator(const ArrayBase<DstXprType>& dst, const ArrayBase<SrcXprType>& src)
+{
+  typedef typename DstXprType::Scalar Scalar;
+  SelfCwiseBinaryOp<internal::scalar_difference_op<Scalar>, DstXprType, SrcXprType> tmp(dst.const_cast_derived());
+  copy_using_evaluator(tmp, src.derived());
+}
+
+template<typename DstXprType, typename SrcXprType>
+void multiply_assign_using_evaluator(const ArrayBase<DstXprType>& dst, const ArrayBase<SrcXprType>& src)
+{
+  typedef typename DstXprType::Scalar Scalar;
+  SelfCwiseBinaryOp<internal::scalar_product_op<Scalar>, DstXprType, SrcXprType> tmp(dst.const_cast_derived());
+  copy_using_evaluator(tmp, src.derived());
+}
+
+template<typename DstXprType, typename SrcXprType>
+void divide_assign_using_evaluator(const ArrayBase<DstXprType>& dst, const ArrayBase<SrcXprType>& src)
+{
+  typedef typename DstXprType::Scalar Scalar;
+  SelfCwiseBinaryOp<internal::scalar_quotient_op<Scalar>, DstXprType, SrcXprType> tmp(dst.const_cast_derived());
+  copy_using_evaluator(tmp, src.derived());
+}
+
+
+} // namespace internal
+
+#endif // EIGEN_ASSIGN_EVALUATOR_H

Eigen/src/Core/BandMatrix.h

@@ -87,7 +87,7 @@ class BandMatrixBase : public EigenBase<Derived>
       if (i<=supers())
       {
         start = supers()-i;
-        len = std::min(rows(),std::max<Index>(0,coeffs().rows() - (supers()-i)));
+        len = (std::min)(rows(),std::max<Index>(0,coeffs().rows() - (supers()-i)));
       }
       else if (i>=rows()-subs())
         len = std::max<Index>(0,coeffs().rows() - (i + 1 - rows() + subs()));
@@ -96,11 +96,11 @@ class BandMatrixBase : public EigenBase<Derived>
 
     /** \returns a vector expression of the main diagonal */
     inline Block<CoefficientsType,1,SizeAtCompileTime> diagonal()
-    { return Block<CoefficientsType,1,SizeAtCompileTime>(coeffs(),supers(),0,1,std::min(rows(),cols())); }
+    { return Block<CoefficientsType,1,SizeAtCompileTime>(coeffs(),supers(),0,1,(std::min)(rows(),cols())); }
 
     /** \returns a vector expression of the main diagonal (const version) */
     inline const Block<const CoefficientsType,1,SizeAtCompileTime> diagonal() const
-    { return Block<const CoefficientsType,1,SizeAtCompileTime>(coeffs(),supers(),0,1,std::min(rows(),cols())); }
+    { return Block<const CoefficientsType,1,SizeAtCompileTime>(coeffs(),supers(),0,1,(std::min)(rows(),cols())); }
 
     template<int Index> struct DiagonalIntReturnType {
       enum {
@@ -122,13 +122,13 @@ class BandMatrixBase : public EigenBase<Derived>
     /** \returns a vector expression of the \a N -th sub or super diagonal */
     template<int N> inline typename DiagonalIntReturnType<N>::Type diagonal()
     {
-      return typename DiagonalIntReturnType<N>::BuildType(coeffs(), supers()-N, std::max(0,N), 1, diagonalLength(N));
+      return typename DiagonalIntReturnType<N>::BuildType(coeffs(), supers()-N, (std::max)(0,N), 1, diagonalLength(N));
     }
 
     /** \returns a vector expression of the \a N -th sub or super diagonal */
     template<int N> inline const typename DiagonalIntReturnType<N>::Type diagonal() const
     {
-      return typename DiagonalIntReturnType<N>::BuildType(coeffs(), supers()-N, std::max(0,N), 1, diagonalLength(N));
+      return typename DiagonalIntReturnType<N>::BuildType(coeffs(), supers()-N, (std::max)(0,N), 1, diagonalLength(N));
     }
 
     /** \returns a vector expression of the \a i -th sub or super diagonal */
@@ -166,7 +166,7 @@ class BandMatrixBase : public EigenBase<Derived>
   protected:
 
     inline Index diagonalLength(Index i) const
-    { return i<0 ? std::min(cols(),rows()+i) : std::min(rows(),cols()-i); }
+    { return i<0 ? (std::min)(cols(),rows()+i) : (std::min)(rows(),cols()-i); }
 };
 
 /**
@@ -180,7 +180,7 @@ class BandMatrixBase : public EigenBase<Derived>
   * \param Cols Number of columns, or \b Dynamic
   * \param Supers Number of super diagonal
   * \param Subs Number of sub diagonal
-  * \param _Options A combination of either \b RowMajor or \b ColMajor, and of \b SelfAdjoint
+  * \param _Options A combination of either \b #RowMajor or \b #ColMajor, and of \b #SelfAdjoint
   *                 The former controls \ref TopicStorageOrders "storage order", and defaults to
   *                 column-major. The latter controls whether the matrix represents a selfadjoint 
   *                 matrix in which case either Supers of Subs have to be null.
@@ -284,6 +284,7 @@ class BandMatrixWrapper : public BandMatrixBase<BandMatrixWrapper<_CoefficientsT
       : m_coeffs(coeffs),
         m_rows(rows), m_supers(supers), m_subs(subs)
     {
+      EIGEN_UNUSED_VARIABLE(cols);
       //internal::assert(coeffs.cols()==cols() && (supers()+subs()+1)==coeffs.rows());
     }
 

Eigen/src/Core/Block.h

@@ -94,7 +94,7 @@ struct traits<Block<XprType, BlockRows, BlockCols, InnerPanel, HasDirectAccess>
     MaskPacketAccessBit = (InnerSize == Dynamic || (InnerSize % packet_traits<Scalar>::size) == 0)
                        && (InnerStrideAtCompileTime == 1)
                         ? PacketAccessBit : 0,
-    MaskAlignedBit = (InnerPanel && (OuterStrideAtCompileTime!=Dynamic) && ((OuterStrideAtCompileTime % packet_traits<Scalar>::size) == 0)) ? AlignedBit : 0,
+    MaskAlignedBit = (InnerPanel && (OuterStrideAtCompileTime!=Dynamic) && (((OuterStrideAtCompileTime * sizeof(Scalar)) % 16) == 0)) ? AlignedBit : 0,
     FlagsLinearAccessBit = (RowsAtCompileTime == 1 || ColsAtCompileTime == 1) ? LinearAccessBit : 0,
     FlagsLvalueBit = is_lvalue<XprType>::value ? LvalueBit : 0,
     FlagsRowMajorBit = IsRowMajor ? RowMajorBit : 0,
@@ -242,6 +242,21 @@ template<typename XprType, int BlockRows, int BlockCols, bool InnerPanel, bool H
     inline Index outerStride() const;
     #endif
 
+    const typename internal::remove_all<typename XprType::Nested>::type& nestedExpression() const 
+    { 
+      return m_xpr; 
+    }
+      
+    Index startRow() const 
+    { 
+      return m_startRow.value(); 
+    }
+      
+    Index startCol() const 
+    { 
+      return m_startCol.value(); 
+    }
+
   protected:
 
     const typename XprType::Nested m_xpr;
@@ -304,6 +319,11 @@ class Block<XprType,BlockRows,BlockCols, InnerPanel,true>
       init();
     }
 
+    const typename internal::remove_all<typename XprType::Nested>::type& nestedExpression() const 
+    { 
+      return m_xpr; 
+    }
+      
     /** \sa MapBase::innerStride() */
     inline Index innerStride() const
     {

Eigen/src/Core/CwiseNullaryOp.h

@@ -101,6 +101,9 @@ class CwiseNullaryOp : internal::no_assignment_operator,
       return m_functor.packetOp(index);
     }
 
+    /** \returns the functor representing the nullary operation */
+    const NullaryOp& functor() const { return m_functor; }
+
   protected:
     const internal::variable_if_dynamic<Index, RowsAtCompileTime> m_rows;
     const internal::variable_if_dynamic<Index, ColsAtCompileTime> m_cols;
@@ -742,7 +745,7 @@ struct setIdentity_impl<Derived, true>
   static EIGEN_STRONG_INLINE Derived& run(Derived& m)
   {
     m.setZero();
-    const Index size = std::min(m.rows(), m.cols());
+    const Index size = (std::min)(m.rows(), m.cols());
     for(Index i = 0; i < size; ++i) m.coeffRef(i,i) = typename Derived::Scalar(1);
     return m;
   }

Eigen/src/Core/DenseBase.h

@@ -34,7 +34,10 @@
   * This class is the base that is inherited by all dense objects (matrix, vector, arrays,
   * and related expression types). The common Eigen API for dense objects is contained in this class.
   *
-  * \param Derived is the derived type, e.g., a matrix type or an expression.
+  * \tparam Derived is the derived type, e.g., a matrix type or an expression.
+  *
+  * This class can be extended with the help of the plugin mechanism described on the page
+  * \ref TopicCustomizingEigen by defining the preprocessor symbol \c EIGEN_DENSEBASE_PLUGIN.
   *
   * \sa \ref TopicClassHierarchy
   */
@@ -53,7 +56,13 @@ template<typename Derived> class DenseBase
     class InnerIterator;
 
     typedef typename internal::traits<Derived>::StorageKind StorageKind;
-    typedef typename internal::traits<Derived>::Index Index; /**< The type of indices */
+
+    /** \brief The type of indices 
+      * \details To change this, \c \#define the preprocessor symbol \c EIGEN_DEFAULT_DENSE_INDEX_TYPE.
+      * \sa \ref TopicPreprocessorDirectives.
+      */
+    typedef typename internal::traits<Derived>::Index Index; 
+
     typedef typename internal::traits<Derived>::Scalar Scalar;
     typedef typename internal::packet_traits<Scalar>::type PacketScalar;
     typedef typename NumTraits<Scalar>::Real RealScalar;

Eigen/src/Core/DenseCoeffsBase.h

@@ -35,7 +35,7 @@ template<typename T> struct add_const_on_value_type_if_arithmetic
 /** \brief Base class providing read-only coefficient access to matrices and arrays.
   * \ingroup Core_Module
   * \tparam Derived Type of the derived class
-  * \tparam ReadOnlyAccessors Constant indicating read-only access
+  * \tparam #ReadOnlyAccessors Constant indicating read-only access
   *
   * This class defines the \c operator() \c const function and friends, which can be used to read specific
   * entries of a matrix or array.
@@ -212,7 +212,7 @@ class DenseCoeffsBase<Derived,ReadOnlyAccessors> : public EigenBase<Derived>
       * to ensure that a packet really starts there. This method is only available on expressions having the
       * PacketAccessBit.
       *
-      * The \a LoadMode parameter may have the value \a Aligned or \a Unaligned. Its effect is to select
+      * The \a LoadMode parameter may have the value \a #Aligned or \a #Unaligned. Its effect is to select
       * the appropriate vectorization instruction. Aligned access is faster, but is only possible for packets
       * starting at an address which is a multiple of the packet size.
       */
@@ -239,7 +239,7 @@ class DenseCoeffsBase<Derived,ReadOnlyAccessors> : public EigenBase<Derived>
       * to ensure that a packet really starts there. This method is only available on expressions having the
       * PacketAccessBit and the LinearAccessBit.
       *
-      * The \a LoadMode parameter may have the value \a Aligned or \a Unaligned. Its effect is to select
+      * The \a LoadMode parameter may have the value \a #Aligned or \a #Unaligned. Its effect is to select
       * the appropriate vectorization instruction. Aligned access is faster, but is only possible for packets
       * starting at an address which is a multiple of the packet size.
       */
@@ -275,7 +275,7 @@ class DenseCoeffsBase<Derived,ReadOnlyAccessors> : public EigenBase<Derived>
 /** \brief Base class providing read/write coefficient access to matrices and arrays.
   * \ingroup Core_Module
   * \tparam Derived Type of the derived class
-  * \tparam WriteAccessors Constant indicating read/write access
+  * \tparam #WriteAccessors Constant indicating read/write access
   *
   * This class defines the non-const \c operator() function and friends, which can be used to write specific
   * entries of a matrix or array. This class inherits DenseCoeffsBase<Derived, ReadOnlyAccessors> which
@@ -433,7 +433,7 @@ class DenseCoeffsBase<Derived, WriteAccessors> : public DenseCoeffsBase<Derived,
       * to ensure that a packet really starts there. This method is only available on expressions having the
       * PacketAccessBit.
       *
-      * The \a LoadMode parameter may have the value \a Aligned or \a Unaligned. Its effect is to select
+      * The \a LoadMode parameter may have the value \a #Aligned or \a #Unaligned. Its effect is to select
       * the appropriate vectorization instruction. Aligned access is faster, but is only possible for packets
       * starting at an address which is a multiple of the packet size.
       */
@@ -567,7 +567,7 @@ class DenseCoeffsBase<Derived, WriteAccessors> : public DenseCoeffsBase<Derived,
 /** \brief Base class providing direct read-only coefficient access to matrices and arrays.
   * \ingroup Core_Module
   * \tparam Derived Type of the derived class
-  * \tparam DirectAccessors Constant indicating direct access
+  * \tparam #DirectAccessors Constant indicating direct access
   *
   * This class defines functions to work with strides which can be used to access entries directly. This class
   * inherits DenseCoeffsBase<Derived, ReadOnlyAccessors> which defines functions to access entries read-only using
@@ -637,7 +637,7 @@ class DenseCoeffsBase<Derived, DirectAccessors> : public DenseCoeffsBase<Derived
 /** \brief Base class providing direct read/write coefficient access to matrices and arrays.
   * \ingroup Core_Module
   * \tparam Derived Type of the derived class
-  * \tparam DirectAccessors Constant indicating direct access
+  * \tparam #DirectWriteAccessors Constant indicating direct access
   *
   * This class defines functions to work with strides which can be used to access entries directly. This class
   * inherits DenseCoeffsBase<Derived, WriteAccessors> which defines functions to access entries read/write using

Eigen/src/Core/DenseStorage.h

@@ -58,14 +58,14 @@ struct plain_array
   #define EIGEN_MAKE_UNALIGNED_ARRAY_ASSERT(sizemask) \
     eigen_assert((reinterpret_cast<size_t>(array) & sizemask) == 0 \
               && "this assertion is explained here: " \
-              "http://eigen.tuxfamily.org/dox/UnalignedArrayAssert.html" \
+              "http://eigen.tuxfamily.org/dox-devel/TopicUnalignedArrayAssert.html" \
               " **** READ THIS WEB PAGE !!! ****");
 #endif
 
 template <typename T, int Size, int MatrixOrArrayOptions>
 struct plain_array<T, Size, MatrixOrArrayOptions, 16>
 {
-  EIGEN_ALIGN16 T array[Size];
+  EIGEN_USER_ALIGN16 T array[Size];
   plain_array() { EIGEN_MAKE_UNALIGNED_ARRAY_ASSERT(0xf) }
   plain_array(constructor_without_unaligned_array_assert) {}
 };
@@ -73,7 +73,7 @@ struct plain_array<T, Size, MatrixOrArrayOptions, 16>
 template <typename T, int MatrixOrArrayOptions, int Alignment>
 struct plain_array<T, 0, MatrixOrArrayOptions, Alignment>
 {
-  EIGEN_ALIGN16 T array[1];
+  EIGEN_USER_ALIGN16 T array[1];
   plain_array() {}
   plain_array(constructor_without_unaligned_array_assert) {}
 };
@@ -128,6 +128,16 @@ template<typename T, int _Rows, int _Cols, int _Options> class DenseStorage<T, 0
     inline T *data() { return 0; }
 };
 
+// more specializations for null matrices; these are necessary to resolve ambiguities
+template<typename T, int _Options> class DenseStorage<T, 0, Dynamic, Dynamic, _Options>
+: public DenseStorage<T, 0, 0, 0, _Options> { };
+
+template<typename T, int _Rows, int _Options> class DenseStorage<T, 0, _Rows, Dynamic, _Options>
+: public DenseStorage<T, 0, 0, 0, _Options> { };
+
+template<typename T, int _Cols, int _Options> class DenseStorage<T, 0, Dynamic, _Cols, _Options>
+: public DenseStorage<T, 0, 0, 0, _Options> { };
+
 // dynamic-size matrix with fixed-size storage
 template<typename T, int Size, int _Options> class DenseStorage<T, Size, Dynamic, Dynamic, _Options>
 {

Eigen/src/Core/Diagonal.h

@@ -87,7 +87,7 @@ template<typename MatrixType, int DiagIndex> class Diagonal
     EIGEN_INHERIT_ASSIGNMENT_OPERATORS(Diagonal)
 
     inline Index rows() const
-    { return m_index.value()<0 ? std::min(m_matrix.cols(),m_matrix.rows()+m_index.value()) : std::min(m_matrix.rows(),m_matrix.cols()-m_index.value()); }
+    { return m_index.value()<0 ? (std::min)(m_matrix.cols(),m_matrix.rows()+m_index.value()) : (std::min)(m_matrix.rows(),m_matrix.cols()-m_index.value()); }
 
     inline Index cols() const { return 1; }
 
@@ -133,6 +133,17 @@ template<typename MatrixType, int DiagIndex> class Diagonal
       return m_matrix.coeff(index+rowOffset(), index+colOffset());
     }
 
+    const typename internal::remove_all<typename MatrixType::Nested>::type& 
+    nestedExpression() const 
+    {
+      return m_matrix;
+    }
+
+    int index() const
+    {
+      return m_index.value();
+    }
+
   protected:
     const typename MatrixType::Nested m_matrix;
     const internal::variable_if_dynamic<Index, DiagIndex> m_index;

Eigen/src/Core/Dot.h

@@ -116,7 +116,9 @@ MatrixBase<Derived>::eigen2_dot(const MatrixBase<OtherDerived>& other) const
 
 //---------- implementation of L2 norm and related functions ----------
 
-/** \returns the squared \em l2 norm of *this, i.e., for vectors, the dot product of *this with itself.
+/** \returns, for vectors, the squared \em l2 norm of \c *this, and for matrices the Frobenius norm.
+  * In both cases, it consists in the sum of the square of all the matrix entries.
+  * For vectors, this is also equals to the dot product of \c *this with itself.
   *
   * \sa dot(), norm()
   */
@@ -126,7 +128,9 @@ EIGEN_STRONG_INLINE typename NumTraits<typename internal::traits<Derived>::Scala
   return internal::real((*this).cwiseAbs2().sum());
 }
 
-/** \returns the \em l2 norm of *this, i.e., for vectors, the square root of the dot product of *this with itself.
+/** \returns, for vectors, the \em l2 norm of \c *this, and for matrices the Frobenius norm.
+  * In both cases, it consists in the square root of the sum of the square of all the matrix entries.
+  * For vectors, this is also equals to the square root of the dot product of \c *this with itself.
   *
   * \sa dot(), squaredNorm()
   */

Eigen/src/Core/Functors.h

@@ -116,7 +116,7 @@ struct functor_traits<scalar_conj_product_op<LhsScalar,RhsScalar> > {
   */
 template<typename Scalar> struct scalar_min_op {
   EIGEN_EMPTY_STRUCT_CTOR(scalar_min_op)
-  EIGEN_STRONG_INLINE const Scalar operator() (const Scalar& a, const Scalar& b) const { return std::min(a, b); }
+  EIGEN_STRONG_INLINE const Scalar operator() (const Scalar& a, const Scalar& b) const { using std::min; return (min)(a, b); }
   template<typename Packet>
   EIGEN_STRONG_INLINE const Packet packetOp(const Packet& a, const Packet& b) const
   { return internal::pmin(a,b); }
@@ -139,7 +139,7 @@ struct functor_traits<scalar_min_op<Scalar> > {
   */
 template<typename Scalar> struct scalar_max_op {
   EIGEN_EMPTY_STRUCT_CTOR(scalar_max_op)
-  EIGEN_STRONG_INLINE const Scalar operator() (const Scalar& a, const Scalar& b) const { return std::max(a, b); }
+  EIGEN_STRONG_INLINE const Scalar operator() (const Scalar& a, const Scalar& b) const { using std::max; return (max)(a, b); }
   template<typename Packet>
   EIGEN_STRONG_INLINE const Packet packetOp(const Packet& a, const Packet& b) const
   { return internal::pmax(a,b); }
@@ -165,8 +165,10 @@ template<typename Scalar> struct scalar_hypot_op {
 //   typedef typename NumTraits<Scalar>::Real result_type;
   EIGEN_STRONG_INLINE const Scalar operator() (const Scalar& _x, const Scalar& _y) const
   {
-    Scalar p = std::max(_x, _y);
-    Scalar q = std::min(_x, _y);
+    using std::max;
+    using std::min;
+    Scalar p = (max)(_x, _y);
+    Scalar q = (min)(_x, _y);
     Scalar qp = q/p;
     return p * sqrt(Scalar(1) + qp*qp);
   }
@@ -218,6 +220,38 @@ struct functor_traits<scalar_quotient_op<Scalar> > {
   };
 };
 
+/** \internal
+  * \brief Template functor to compute the and of two booleans
+  *
+  * \sa class CwiseBinaryOp, ArrayBase::operator&&
+  */
+struct scalar_boolean_and_op {
+  EIGEN_EMPTY_STRUCT_CTOR(scalar_boolean_and_op)
+  EIGEN_STRONG_INLINE bool operator() (const bool& a, const bool& b) const { return a && b; }
+};
+template<> struct functor_traits<scalar_boolean_and_op> {
+  enum {
+    Cost = NumTraits<bool>::AddCost,
+    PacketAccess = false
+  };
+};
+
+/** \internal
+  * \brief Template functor to compute the or of two booleans
+  *
+  * \sa class CwiseBinaryOp, ArrayBase::operator||
+  */
+struct scalar_boolean_or_op {
+  EIGEN_EMPTY_STRUCT_CTOR(scalar_boolean_or_op)
+  EIGEN_STRONG_INLINE bool operator() (const bool& a, const bool& b) const { return a || b; }
+};
+template<> struct functor_traits<scalar_boolean_or_op> {
+  enum {
+    Cost = NumTraits<bool>::AddCost,
+    PacketAccess = false
+  };
+};
+
 // unary functors:
 
 /** \internal
@@ -605,7 +639,7 @@ template <typename Scalar, bool RandomAccess> struct linspaced_op
   EIGEN_STRONG_INLINE const Packet packetOp(Index row, Index col) const
   {
     eigen_assert(col==0 || row==0);
-    return impl(col + row);
+    return impl.packetOp(col + row);
   }
 
   // This proxy object handles the actual required temporaries, the different
@@ -669,7 +703,7 @@ struct functor_traits<scalar_sqrt_op<Scalar> >
 
 /** \internal
   * \brief Template functor to compute the cosine of a scalar
-  * \sa class CwiseUnaryOp, Cwise::cos()
+  * \sa class CwiseUnaryOp, ArrayBase::cos()
   */
 template<typename Scalar> struct scalar_cos_op {
   EIGEN_EMPTY_STRUCT_CTOR(scalar_cos_op)
@@ -688,7 +722,7 @@ struct functor_traits<scalar_cos_op<Scalar> >
 
 /** \internal
   * \brief Template functor to compute the sine of a scalar
-  * \sa class CwiseUnaryOp, Cwise::sin()
+  * \sa class CwiseUnaryOp, ArrayBase::sin()
   */
 template<typename Scalar> struct scalar_sin_op {
   EIGEN_EMPTY_STRUCT_CTOR(scalar_sin_op)
@@ -708,7 +742,7 @@ struct functor_traits<scalar_sin_op<Scalar> >
 
 /** \internal
   * \brief Template functor to compute the tan of a scalar
-  * \sa class CwiseUnaryOp, Cwise::tan()
+  * \sa class CwiseUnaryOp, ArrayBase::tan()
   */
 template<typename Scalar> struct scalar_tan_op {
   EIGEN_EMPTY_STRUCT_CTOR(scalar_tan_op)
@@ -725,6 +759,44 @@ struct functor_traits<scalar_tan_op<Scalar> >
   };
 };
 
+/** \internal
+  * \brief Template functor to compute the arc cosine of a scalar
+  * \sa class CwiseUnaryOp, ArrayBase::acos()
+  */
+template<typename Scalar> struct scalar_acos_op {
+  EIGEN_EMPTY_STRUCT_CTOR(scalar_acos_op)
+  inline const Scalar operator() (const Scalar& a) const { return acos(a); }
+  typedef typename packet_traits<Scalar>::type Packet;
+  inline Packet packetOp(const Packet& a) const { return internal::pacos(a); }
+};
+template<typename Scalar>
+struct functor_traits<scalar_acos_op<Scalar> >
+{
+  enum {
+    Cost = 5 * NumTraits<Scalar>::MulCost,
+    PacketAccess = packet_traits<Scalar>::HasACos
+  };
+};
+
+/** \internal
+  * \brief Template functor to compute the arc sine of a scalar
+  * \sa class CwiseUnaryOp, ArrayBase::asin()
+  */
+template<typename Scalar> struct scalar_asin_op {
+  EIGEN_EMPTY_STRUCT_CTOR(scalar_asin_op)
+  inline const Scalar operator() (const Scalar& a) const { return asin(a); }
+  typedef typename packet_traits<Scalar>::type Packet;
+  inline Packet packetOp(const Packet& a) const { return internal::pasin(a); }
+};
+template<typename Scalar>
+struct functor_traits<scalar_asin_op<Scalar> >
+{
+  enum {
+    Cost = 5 * NumTraits<Scalar>::MulCost,
+    PacketAccess = packet_traits<Scalar>::HasASin
+  };
+};
+
 /** \internal
   * \brief Template functor to raise a scalar to a power
   * \sa class CwiseUnaryOp, Cwise::pow

Eigen/src/Core/Fuzzy.h

@@ -34,9 +34,10 @@ struct isApprox_selector
 {
   static bool run(const Derived& x, const OtherDerived& y, typename Derived::RealScalar prec)
   {
+    using std::min;
     const typename internal::nested<Derived,2>::type nested(x);
     const typename internal::nested<OtherDerived,2>::type otherNested(y);
-    return (nested - otherNested).cwiseAbs2().sum() <= prec * prec * std::min(nested.cwiseAbs2().sum(), otherNested.cwiseAbs2().sum());
+    return (nested - otherNested).cwiseAbs2().sum() <= prec * prec * (min)(nested.cwiseAbs2().sum(), otherNested.cwiseAbs2().sum());
   }
 };
 

Eigen/src/Core/GeneralProduct.h

@@ -0,0 +1,624 @@
+// This file is part of Eigen, a lightweight C++ template library
+// for linear algebra.
+//
+// Copyright (C) 2006-2008 Benoit Jacob <jacob.benoit.1@gmail.com>
+// Copyright (C) 2008-2011 Gael Guennebaud <gael.guennebaud@inria.fr>
+//
+// Eigen is free software; you can redistribute it and/or
+// modify it under the terms of the GNU Lesser General Public
+// License as published by the Free Software Foundation; either
+// version 3 of the License, or (at your option) any later version.
+//
+// Alternatively, you can redistribute it and/or
+// modify it under the terms of the GNU General Public License as
+// published by the Free Software Foundation; either version 2 of
+// the License, or (at your option) any later version.
+//
+// Eigen is distributed in the hope that it will be useful, but WITHOUT ANY
+// WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
+// FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License or the
+// GNU General Public License for more details.
+//
+// You should have received a copy of the GNU Lesser General Public
+// License and a copy of the GNU General Public License along with
+// Eigen. If not, see <http://www.gnu.org/licenses/>.
+
+#ifndef EIGEN_GENERAL_PRODUCT_H
+#define EIGEN_GENERAL_PRODUCT_H
+
+/** \class GeneralProduct
+  * \ingroup Core_Module
+  *
+  * \brief Expression of the product of two general matrices or vectors
+  *
+  * \param LhsNested the type used to store the left-hand side
+  * \param RhsNested the type used to store the right-hand side
+  * \param ProductMode the type of the product
+  *
+  * This class represents an expression of the product of two general matrices.
+  * We call a general matrix, a dense matrix with full storage. For instance,
+  * This excludes triangular, selfadjoint, and sparse matrices.
+  * It is the return type of the operator* between general matrices. Its template
+  * arguments are determined automatically by ProductReturnType. Therefore,
+  * GeneralProduct should never be used direclty. To determine the result type of a
+  * function which involves a matrix product, use ProductReturnType::Type.
+  *
+  * \sa ProductReturnType, MatrixBase::operator*(const MatrixBase<OtherDerived>&)
+  */
+template<typename Lhs, typename Rhs, int ProductType = internal::product_type<Lhs,Rhs>::value>
+class GeneralProduct;
+
+enum {
+  Large = 2,
+  Small = 3
+};
+
+namespace internal {
+
+template<int Rows, int Cols, int Depth> struct product_type_selector;
+
+template<int Size, int MaxSize> struct product_size_category
+{
+  enum { is_large = MaxSize == Dynamic ||
+                    Size >= EIGEN_CACHEFRIENDLY_PRODUCT_THRESHOLD,
+         value = is_large  ? Large
+               : Size == 1 ? 1
+                           : Small
+  };
+};
+
+template<typename Lhs, typename Rhs> struct product_type
+{
+  typedef typename remove_all<Lhs>::type _Lhs;
+  typedef typename remove_all<Rhs>::type _Rhs;
+  enum {
+    MaxRows  = _Lhs::MaxRowsAtCompileTime,
+    Rows  = _Lhs::RowsAtCompileTime,
+    MaxCols  = _Rhs::MaxColsAtCompileTime,
+    Cols  = _Rhs::ColsAtCompileTime,
+    MaxDepth = EIGEN_SIZE_MIN_PREFER_FIXED(_Lhs::MaxColsAtCompileTime,
+                                           _Rhs::MaxRowsAtCompileTime),
+    Depth = EIGEN_SIZE_MIN_PREFER_FIXED(_Lhs::ColsAtCompileTime,
+                                        _Rhs::RowsAtCompileTime),
+    LargeThreshold = EIGEN_CACHEFRIENDLY_PRODUCT_THRESHOLD
+  };
+
+  // the splitting into different lines of code here, introducing the _select enums and the typedef below,
+  // is to work around an internal compiler error with gcc 4.1 and 4.2.
+private:
+  enum {
+    rows_select = product_size_category<Rows,MaxRows>::value,
+    cols_select = product_size_category<Cols,MaxCols>::value,
+    depth_select = product_size_category<Depth,MaxDepth>::value
+  };
+  typedef product_type_selector<rows_select, cols_select, depth_select> selector;
+
+public:
+  enum {
+    value = selector::ret
+  };
+#ifdef EIGEN_DEBUG_PRODUCT
+  static void debug()
+  {
+      EIGEN_DEBUG_VAR(Rows);
+      EIGEN_DEBUG_VAR(Cols);
+      EIGEN_DEBUG_VAR(Depth);
+      EIGEN_DEBUG_VAR(rows_select);
+      EIGEN_DEBUG_VAR(cols_select);
+      EIGEN_DEBUG_VAR(depth_select);
+      EIGEN_DEBUG_VAR(value);
+  }
+#endif
+};
+
+
+/* The following allows to select the kind of product at compile time
+ * based on the three dimensions of the product.
+ * This is a compile time mapping from {1,Small,Large}^3 -> {product types} */
+// FIXME I'm not sure the current mapping is the ideal one.
+template<int M, int N>  struct product_type_selector<M,N,1>              { enum { ret = OuterProduct }; };
+template<int Depth>     struct product_type_selector<1,    1,    Depth>  { enum { ret = InnerProduct }; };
+template<>              struct product_type_selector<1,    1,    1>      { enum { ret = InnerProduct }; };
+template<>              struct product_type_selector<Small,1,    Small>  { enum { ret = CoeffBasedProductMode }; };
+template<>              struct product_type_selector<1,    Small,Small>  { enum { ret = CoeffBasedProductMode }; };
+template<>              struct product_type_selector<Small,Small,Small>  { enum { ret = CoeffBasedProductMode }; };
+template<>              struct product_type_selector<Small, Small, 1>    { enum { ret = LazyCoeffBasedProductMode }; };
+template<>              struct product_type_selector<Small, Large, 1>    { enum { ret = LazyCoeffBasedProductMode }; };
+template<>              struct product_type_selector<Large, Small, 1>    { enum { ret = LazyCoeffBasedProductMode }; };
+template<>              struct product_type_selector<1,    Large,Small>  { enum { ret = CoeffBasedProductMode }; };
+template<>              struct product_type_selector<1,    Large,Large>  { enum { ret = GemvProduct }; };
+template<>              struct product_type_selector<1,    Small,Large>  { enum { ret = CoeffBasedProductMode }; };
+template<>              struct product_type_selector<Large,1,    Small>  { enum { ret = CoeffBasedProductMode }; };
+template<>              struct product_type_selector<Large,1,    Large>  { enum { ret = GemvProduct }; };
+template<>              struct product_type_selector<Small,1,    Large>  { enum { ret = CoeffBasedProductMode }; };
+template<>              struct product_type_selector<Small,Small,Large>  { enum { ret = GemmProduct }; };
+template<>              struct product_type_selector<Large,Small,Large>  { enum { ret = GemmProduct }; };
+template<>              struct product_type_selector<Small,Large,Large>  { enum { ret = GemmProduct }; };
+template<>              struct product_type_selector<Large,Large,Large>  { enum { ret = GemmProduct }; };
+template<>              struct product_type_selector<Large,Small,Small>  { enum { ret = GemmProduct }; };
+template<>              struct product_type_selector<Small,Large,Small>  { enum { ret = GemmProduct }; };
+template<>              struct product_type_selector<Large,Large,Small>  { enum { ret = GemmProduct }; };
+
+} // end namespace internal
+
+/** \class ProductReturnType
+  * \ingroup Core_Module
+  *
+  * \brief Helper class to get the correct and optimized returned type of operator*
+  *
+  * \param Lhs the type of the left-hand side
+  * \param Rhs the type of the right-hand side
+  * \param ProductMode the type of the product (determined automatically by internal::product_mode)
+  *
+  * This class defines the typename Type representing the optimized product expression
+  * between two matrix expressions. In practice, using ProductReturnType<Lhs,Rhs>::Type
+  * is the recommended way to define the result type of a function returning an expression
+  * which involve a matrix product. The class Product should never be
+  * used directly.
+  *
+  * \sa class Product, MatrixBase::operator*(const MatrixBase<OtherDerived>&)
+  */
+template<typename Lhs, typename Rhs, int ProductType>
+struct ProductReturnType
+{
+  // TODO use the nested type to reduce instanciations ????
+//   typedef typename internal::nested<Lhs,Rhs::ColsAtCompileTime>::type LhsNested;
+//   typedef typename internal::nested<Rhs,Lhs::RowsAtCompileTime>::type RhsNested;
+
+  typedef GeneralProduct<Lhs/*Nested*/, Rhs/*Nested*/, ProductType> Type;
+};
+
+template<typename Lhs, typename Rhs>
+struct ProductReturnType<Lhs,Rhs,CoeffBasedProductMode>
+{
+  typedef typename internal::nested<Lhs, Rhs::ColsAtCompileTime, typename internal::plain_matrix_type<Lhs>::type >::type LhsNested;
+  typedef typename internal::nested<Rhs, Lhs::RowsAtCompileTime, typename internal::plain_matrix_type<Rhs>::type >::type RhsNested;
+  typedef CoeffBasedProduct<LhsNested, RhsNested, EvalBeforeAssigningBit | EvalBeforeNestingBit> Type;
+};
+
+template<typename Lhs, typename Rhs>
+struct ProductReturnType<Lhs,Rhs,LazyCoeffBasedProductMode>
+{
+  typedef typename internal::nested<Lhs, Rhs::ColsAtCompileTime, typename internal::plain_matrix_type<Lhs>::type >::type LhsNested;
+  typedef typename internal::nested<Rhs, Lhs::RowsAtCompileTime, typename internal::plain_matrix_type<Rhs>::type >::type RhsNested;
+  typedef CoeffBasedProduct<LhsNested, RhsNested, NestByRefBit> Type;
+};
+
+// this is a workaround for sun CC
+template<typename Lhs, typename Rhs>
+struct LazyProductReturnType : public ProductReturnType<Lhs,Rhs,LazyCoeffBasedProductMode>
+{};
+
+/***********************************************************************
+*  Implementation of Inner Vector Vector Product
+***********************************************************************/
+
+// FIXME : maybe the "inner product" could return a Scalar
+// instead of a 1x1 matrix ??
+// Pro: more natural for the user
+// Cons: this could be a problem if in a meta unrolled algorithm a matrix-matrix
+// product ends up to a row-vector times col-vector product... To tackle this use
+// case, we could have a specialization for Block<MatrixType,1,1> with: operator=(Scalar x);
+
+namespace internal {
+
+template<typename Lhs, typename Rhs>
+struct traits<GeneralProduct<Lhs,Rhs,InnerProduct> >
+ : traits<Matrix<typename scalar_product_traits<typename Lhs::Scalar, typename Rhs::Scalar>::ReturnType,1,1> >
+{};
+
+}
+
+template<typename Lhs, typename Rhs>
+class GeneralProduct<Lhs, Rhs, InnerProduct>
+  : internal::no_assignment_operator,
+    public Matrix<typename internal::scalar_product_traits<typename Lhs::Scalar, typename Rhs::Scalar>::ReturnType,1,1>
+{
+    typedef Matrix<typename internal::scalar_product_traits<typename Lhs::Scalar, typename Rhs::Scalar>::ReturnType,1,1> Base;
+  public:
+    GeneralProduct(const Lhs& lhs, const Rhs& rhs)
+    {
+      EIGEN_STATIC_ASSERT((internal::is_same<typename Lhs::RealScalar, typename Rhs::RealScalar>::value),
+        YOU_MIXED_DIFFERENT_NUMERIC_TYPES__YOU_NEED_TO_USE_THE_CAST_METHOD_OF_MATRIXBASE_TO_CAST_NUMERIC_TYPES_EXPLICITLY)
+
+      Base::coeffRef(0,0) = (lhs.transpose().cwiseProduct(rhs)).sum();
+    }
+
+    /** Convertion to scalar */
+    operator const typename Base::Scalar() const {
+      return Base::coeff(0,0);
+    }
+};
+
+/***********************************************************************
+*  Implementation of Outer Vector Vector Product
+***********************************************************************/
+
+namespace internal {
+template<int StorageOrder> struct outer_product_selector;
+
+template<typename Lhs, typename Rhs>
+struct traits<GeneralProduct<Lhs,Rhs,OuterProduct> >
+ : traits<ProductBase<GeneralProduct<Lhs,Rhs,OuterProduct>, Lhs, Rhs> >
+{};
+
+}
+
+template<typename Lhs, typename Rhs>
+class GeneralProduct<Lhs, Rhs, OuterProduct>
+  : public ProductBase<GeneralProduct<Lhs,Rhs,OuterProduct>, Lhs, Rhs>
+{
+  public:
+    EIGEN_PRODUCT_PUBLIC_INTERFACE(GeneralProduct)
+
+    GeneralProduct(const Lhs& lhs, const Rhs& rhs) : Base(lhs,rhs)
+    {
+      EIGEN_STATIC_ASSERT((internal::is_same<typename Lhs::RealScalar, typename Rhs::RealScalar>::value),
+        YOU_MIXED_DIFFERENT_NUMERIC_TYPES__YOU_NEED_TO_USE_THE_CAST_METHOD_OF_MATRIXBASE_TO_CAST_NUMERIC_TYPES_EXPLICITLY)
+    }
+
+    template<typename Dest> void scaleAndAddTo(Dest& dest, Scalar alpha) const
+    {
+      internal::outer_product_selector<(int(Dest::Flags)&RowMajorBit) ? RowMajor : ColMajor>::run(*this, dest, alpha);
+    }
+};
+
+namespace internal {
+
+template<> struct outer_product_selector<ColMajor> {
+  template<typename ProductType, typename Dest>
+  static EIGEN_DONT_INLINE void run(const ProductType& prod, Dest& dest, typename ProductType::Scalar alpha) {
+    typedef typename Dest::Index Index;
+    // FIXME make sure lhs is sequentially stored
+    // FIXME not very good if rhs is real and lhs complex while alpha is real too
+    const Index cols = dest.cols();
+    for (Index j=0; j<cols; ++j)
+      dest.col(j) += (alpha * prod.rhs().coeff(j)) * prod.lhs();
+  }
+};
+
+template<> struct outer_product_selector<RowMajor> {
+  template<typename ProductType, typename Dest>
+  static EIGEN_DONT_INLINE void run(const ProductType& prod, Dest& dest, typename ProductType::Scalar alpha) {
+    typedef typename Dest::Index Index;
+    // FIXME make sure rhs is sequentially stored
+    // FIXME not very good if lhs is real and rhs complex while alpha is real too
+    const Index rows = dest.rows();
+    for (Index i=0; i<rows; ++i)
+      dest.row(i) += (alpha * prod.lhs().coeff(i)) * prod.rhs();
+  }
+};
+
+} // end namespace internal
+
+/***********************************************************************
+*  Implementation of General Matrix Vector Product
+***********************************************************************/
+
+/*  According to the shape/flags of the matrix we have to distinghish 3 different cases:
+ *   1 - the matrix is col-major, BLAS compatible and M is large => call fast BLAS-like colmajor routine
+ *   2 - the matrix is row-major, BLAS compatible and N is large => call fast BLAS-like rowmajor routine
+ *   3 - all other cases are handled using a simple loop along the outer-storage direction.
+ *  Therefore we need a lower level meta selector.
+ *  Furthermore, if the matrix is the rhs, then the product has to be transposed.
+ */
+namespace internal {
+
+template<typename Lhs, typename Rhs>
+struct traits<GeneralProduct<Lhs,Rhs,GemvProduct> >
+ : traits<ProductBase<GeneralProduct<Lhs,Rhs,GemvProduct>, Lhs, Rhs> >
+{};
+
+template<int Side, int StorageOrder, bool BlasCompatible>
+struct gemv_selector;
+
+} // end namespace internal
+
+template<typename Lhs, typename Rhs>
+class GeneralProduct<Lhs, Rhs, GemvProduct>
+  : public ProductBase<GeneralProduct<Lhs,Rhs,GemvProduct>, Lhs, Rhs>
+{
+  public:
+    EIGEN_PRODUCT_PUBLIC_INTERFACE(GeneralProduct)
+
+    typedef typename Lhs::Scalar LhsScalar;
+    typedef typename Rhs::Scalar RhsScalar;
+
+    GeneralProduct(const Lhs& lhs, const Rhs& rhs) : Base(lhs,rhs)
+    {
+//       EIGEN_STATIC_ASSERT((internal::is_same<typename Lhs::Scalar, typename Rhs::Scalar>::value),
+//         YOU_MIXED_DIFFERENT_NUMERIC_TYPES__YOU_NEED_TO_USE_THE_CAST_METHOD_OF_MATRIXBASE_TO_CAST_NUMERIC_TYPES_EXPLICITLY)
+    }
+
+    enum { Side = Lhs::IsVectorAtCompileTime ? OnTheLeft : OnTheRight };
+    typedef typename internal::conditional<int(Side)==OnTheRight,_LhsNested,_RhsNested>::type MatrixType;
+
+    template<typename Dest> void scaleAndAddTo(Dest& dst, Scalar alpha) const
+    {
+      eigen_assert(m_lhs.rows() == dst.rows() && m_rhs.cols() == dst.cols());
+      internal::gemv_selector<Side,(int(MatrixType::Flags)&RowMajorBit) ? RowMajor : ColMajor,
+                       bool(internal::blas_traits<MatrixType>::HasUsableDirectAccess)>::run(*this, dst, alpha);
+    }
+};
+
+namespace internal {
+
+// The vector is on the left => transposition
+template<int StorageOrder, bool BlasCompatible>
+struct gemv_selector<OnTheLeft,StorageOrder,BlasCompatible>
+{
+  template<typename ProductType, typename Dest>
+  static void run(const ProductType& prod, Dest& dest, typename ProductType::Scalar alpha)
+  {
+    Transpose<Dest> destT(dest);
+    enum { OtherStorageOrder = StorageOrder == RowMajor ? ColMajor : RowMajor };
+    gemv_selector<OnTheRight,OtherStorageOrder,BlasCompatible>
+      ::run(GeneralProduct<Transpose<const typename ProductType::_RhsNested>,Transpose<const typename ProductType::_LhsNested>, GemvProduct>
+        (prod.rhs().transpose(), prod.lhs().transpose()), destT, alpha);
+  }
+};
+
+template<typename Scalar,int Size,int MaxSize,bool Cond> struct gemv_static_vector_if;
+
+template<typename Scalar,int Size,int MaxSize>
+struct gemv_static_vector_if<Scalar,Size,MaxSize,false>
+{
+  EIGEN_STRONG_INLINE  Scalar* data() { eigen_internal_assert(false && "should never be called"); return 0; }
+};
+
+template<typename Scalar,int Size>
+struct gemv_static_vector_if<Scalar,Size,Dynamic,true>
+{
+  EIGEN_STRONG_INLINE Scalar* data() { return 0; }
+};
+
+template<typename Scalar,int Size,int MaxSize>
+struct gemv_static_vector_if<Scalar,Size,MaxSize,true>
+{
+  #if EIGEN_ALIGN_STATICALLY
+  internal::plain_array<Scalar,EIGEN_SIZE_MIN_PREFER_FIXED(Size,MaxSize),0> m_data;
+  EIGEN_STRONG_INLINE Scalar* data() { return m_data.array; }
+  #else
+  // Some architectures cannot align on the stack,
+  // => let's manually enforce alignment by allocating more data and return the address of the first aligned element.
+  enum {
+    ForceAlignment  = internal::packet_traits<Scalar>::Vectorizable,
+    PacketSize      = internal::packet_traits<Scalar>::size
+  };
+  internal::plain_array<Scalar,EIGEN_SIZE_MIN_PREFER_FIXED(Size,MaxSize)+(ForceAlignment?PacketSize:0),0> m_data;
+  EIGEN_STRONG_INLINE Scalar* data() {
+    return ForceAlignment
+            ? reinterpret_cast<Scalar*>((reinterpret_cast<size_t>(m_data.array) & ~(size_t(15))) + 16)
+            : m_data.array;
+  }
+  #endif
+};
+
+template<> struct gemv_selector<OnTheRight,ColMajor,true>
+{
+  template<typename ProductType, typename Dest>
+  static inline void run(const ProductType& prod, Dest& dest, typename ProductType::Scalar alpha)
+  {
+    typedef typename ProductType::Index Index;
+    typedef typename ProductType::LhsScalar   LhsScalar;
+    typedef typename ProductType::RhsScalar   RhsScalar;
+    typedef typename ProductType::Scalar      ResScalar;
+    typedef typename ProductType::RealScalar  RealScalar;
+    typedef typename ProductType::ActualLhsType ActualLhsType;
+    typedef typename ProductType::ActualRhsType ActualRhsType;
+    typedef typename ProductType::LhsBlasTraits LhsBlasTraits;
+    typedef typename ProductType::RhsBlasTraits RhsBlasTraits;
+    typedef Map<Matrix<ResScalar,Dynamic,1>, Aligned> MappedDest;
+
+    const ActualLhsType actualLhs = LhsBlasTraits::extract(prod.lhs());
+    const ActualRhsType actualRhs = RhsBlasTraits::extract(prod.rhs());
+
+    ResScalar actualAlpha = alpha * LhsBlasTraits::extractScalarFactor(prod.lhs())
+                                  * RhsBlasTraits::extractScalarFactor(prod.rhs());
+
+    enum {
+      // FIXME find a way to allow an inner stride on the result if packet_traits<Scalar>::size==1
+      // on, the other hand it is good for the cache to pack the vector anyways...
+      EvalToDestAtCompileTime = Dest::InnerStrideAtCompileTime==1,
+      ComplexByReal = (NumTraits<LhsScalar>::IsComplex) && (!NumTraits<RhsScalar>::IsComplex),
+      MightCannotUseDest = (Dest::InnerStrideAtCompileTime!=1) || ComplexByReal
+    };
+
+    gemv_static_vector_if<ResScalar,Dest::SizeAtCompileTime,Dest::MaxSizeAtCompileTime,MightCannotUseDest> static_dest;
+
+    bool alphaIsCompatible = (!ComplexByReal) || (imag(actualAlpha)==RealScalar(0));
+    bool evalToDest = EvalToDestAtCompileTime && alphaIsCompatible;
+    
+    RhsScalar compatibleAlpha = get_factor<ResScalar,RhsScalar>::run(actualAlpha);
+
+    ei_declare_aligned_stack_constructed_variable(ResScalar,actualDestPtr,dest.size(),
+                                                  evalToDest ? dest.data() : static_dest.data());
+    
+    if(!evalToDest)
+    {
+      #ifdef EIGEN_DENSE_STORAGE_CTOR_PLUGIN
+      int size = dest.size();
+      EIGEN_DENSE_STORAGE_CTOR_PLUGIN
+      #endif
+      if(!alphaIsCompatible)
+      {
+        MappedDest(actualDestPtr, dest.size()).setZero();
+        compatibleAlpha = RhsScalar(1);
+      }
+      else
+        MappedDest(actualDestPtr, dest.size()) = dest;
+    }
+
+    general_matrix_vector_product
+      <Index,LhsScalar,ColMajor,LhsBlasTraits::NeedToConjugate,RhsScalar,RhsBlasTraits::NeedToConjugate>::run(
+        actualLhs.rows(), actualLhs.cols(),
+        &actualLhs.coeffRef(0,0), actualLhs.outerStride(),
+        actualRhs.data(), actualRhs.innerStride(),
+        actualDestPtr, 1,
+        compatibleAlpha);
+
+    if (!evalToDest)
+    {
+      if(!alphaIsCompatible)
+        dest += actualAlpha * MappedDest(actualDestPtr, dest.size());
+      else
+        dest = MappedDest(actualDestPtr, dest.size());
+    }
+  }
+};
+
+template<> struct gemv_selector<OnTheRight,RowMajor,true>
+{
+  template<typename ProductType, typename Dest>
+  static void run(const ProductType& prod, Dest& dest, typename ProductType::Scalar alpha)
+  {
+    typedef typename ProductType::LhsScalar LhsScalar;
+    typedef typename ProductType::RhsScalar RhsScalar;
+    typedef typename ProductType::Scalar    ResScalar;
+    typedef typename ProductType::Index Index;
+    typedef typename ProductType::ActualLhsType ActualLhsType;
+    typedef typename ProductType::ActualRhsType ActualRhsType;
+    typedef typename ProductType::_ActualRhsType _ActualRhsType;
+    typedef typename ProductType::LhsBlasTraits LhsBlasTraits;
+    typedef typename ProductType::RhsBlasTraits RhsBlasTraits;
+
+    typename add_const<ActualLhsType>::type actualLhs = LhsBlasTraits::extract(prod.lhs());
+    typename add_const<ActualRhsType>::type actualRhs = RhsBlasTraits::extract(prod.rhs());
+
+    ResScalar actualAlpha = alpha * LhsBlasTraits::extractScalarFactor(prod.lhs())
+                                  * RhsBlasTraits::extractScalarFactor(prod.rhs());
+
+    enum {
+      // FIXME find a way to allow an inner stride on the result if packet_traits<Scalar>::size==1
+      // on, the other hand it is good for the cache to pack the vector anyways...
+      DirectlyUseRhs = _ActualRhsType::InnerStrideAtCompileTime==1
+    };
+
+    gemv_static_vector_if<RhsScalar,_ActualRhsType::SizeAtCompileTime,_ActualRhsType::MaxSizeAtCompileTime,!DirectlyUseRhs> static_rhs;
+
+    ei_declare_aligned_stack_constructed_variable(RhsScalar,actualRhsPtr,actualRhs.size(),
+        DirectlyUseRhs ? const_cast<RhsScalar*>(actualRhs.data()) : static_rhs.data());
+
+    if(!DirectlyUseRhs)
+    {
+      #ifdef EIGEN_DENSE_STORAGE_CTOR_PLUGIN
+      int size = actualRhs.size();
+      EIGEN_DENSE_STORAGE_CTOR_PLUGIN
+      #endif
+      Map<typename _ActualRhsType::PlainObject>(actualRhsPtr, actualRhs.size()) = actualRhs;
+    }
+
+    general_matrix_vector_product
+      <Index,LhsScalar,RowMajor,LhsBlasTraits::NeedToConjugate,RhsScalar,RhsBlasTraits::NeedToConjugate>::run(
+        actualLhs.rows(), actualLhs.cols(),
+        &actualLhs.coeffRef(0,0), actualLhs.outerStride(),
+        actualRhsPtr, 1,
+        &dest.coeffRef(0,0), dest.innerStride(),
+        actualAlpha);
+  }
+};
+
+template<> struct gemv_selector<OnTheRight,ColMajor,false>
+{
+  template<typename ProductType, typename Dest>
+  static void run(const ProductType& prod, Dest& dest, typename ProductType::Scalar alpha)
+  {
+    typedef typename Dest::Index Index;
+    // TODO makes sure dest is sequentially stored in memory, otherwise use a temp
+    const Index size = prod.rhs().rows();
+    for(Index k=0; k<size; ++k)
+      dest += (alpha*prod.rhs().coeff(k)) * prod.lhs().col(k);
+  }
+};
+
+template<> struct gemv_selector<OnTheRight,RowMajor,false>
+{
+  template<typename ProductType, typename Dest>
+  static void run(const ProductType& prod, Dest& dest, typename ProductType::Scalar alpha)
+  {
+    typedef typename Dest::Index Index;
+    // TODO makes sure rhs is sequentially stored in memory, otherwise use a temp
+    const Index rows = prod.rows();
+    for(Index i=0; i<rows; ++i)
+      dest.coeffRef(i) += alpha * (prod.lhs().row(i).cwiseProduct(prod.rhs().transpose())).sum();
+  }
+};
+
+} // end namespace internal
+
+/***************************************************************************
+* Implementation of matrix base methods
+***************************************************************************/
+
+/** \returns the matrix product of \c *this and \a other.
+  *
+  * \note If instead of the matrix product you want the coefficient-wise product, see Cwise::operator*().
+  *
+  * \sa lazyProduct(), operator*=(const MatrixBase&), Cwise::operator*()
+  */
+template<typename Derived>
+template<typename OtherDerived>
+inline const typename ProductReturnType<Derived,OtherDerived>::Type
+MatrixBase<Derived>::operator*(const MatrixBase<OtherDerived> &other) const
+{
+  // A note regarding the function declaration: In MSVC, this function will sometimes
+  // not be inlined since DenseStorage is an unwindable object for dynamic
+  // matrices and product types are holding a member to store the result.
+  // Thus it does not help tagging this function with EIGEN_STRONG_INLINE.
+  enum {
+    ProductIsValid =  Derived::ColsAtCompileTime==Dynamic
+                   || OtherDerived::RowsAtCompileTime==Dynamic
+                   || int(Derived::ColsAtCompileTime)==int(OtherDerived::RowsAtCompileTime),
+    AreVectors = Derived::IsVectorAtCompileTime && OtherDerived::IsVectorAtCompileTime,
+    SameSizes = EIGEN_PREDICATE_SAME_MATRIX_SIZE(Derived,OtherDerived)
+  };
+  // note to the lost user:
+  //    * for a dot product use: v1.dot(v2)
+  //    * for a coeff-wise product use: v1.cwiseProduct(v2)
+  EIGEN_STATIC_ASSERT(ProductIsValid || !(AreVectors && SameSizes),
+    INVALID_VECTOR_VECTOR_PRODUCT__IF_YOU_WANTED_A_DOT_OR_COEFF_WISE_PRODUCT_YOU_MUST_USE_THE_EXPLICIT_FUNCTIONS)
+  EIGEN_STATIC_ASSERT(ProductIsValid || !(SameSizes && !AreVectors),
+    INVALID_MATRIX_PRODUCT__IF_YOU_WANTED_A_COEFF_WISE_PRODUCT_YOU_MUST_USE_THE_EXPLICIT_FUNCTION)
+  EIGEN_STATIC_ASSERT(ProductIsValid || SameSizes, INVALID_MATRIX_PRODUCT)
+#ifdef EIGEN_DEBUG_PRODUCT
+  internal::product_type<Derived,OtherDerived>::debug();
+#endif
+  return typename ProductReturnType<Derived,OtherDerived>::Type(derived(), other.derived());
+}
+
+/** \returns an expression of the matrix product of \c *this and \a other without implicit evaluation.
+  *
+  * The returned product will behave like any other expressions: the coefficients of the product will be
+  * computed once at a time as requested. This might be useful in some extremely rare cases when only
+  * a small and no coherent fraction of the result's coefficients have to be computed.
+  *
+  * \warning This version of the matrix product can be much much slower. So use it only if you know
+  * what you are doing and that you measured a true speed improvement.
+  *
+  * \sa operator*(const MatrixBase&)
+  */
+template<typename Derived>
+template<typename OtherDerived>
+const typename LazyProductReturnType<Derived,OtherDerived>::Type
+MatrixBase<Derived>::lazyProduct(const MatrixBase<OtherDerived> &other) const
+{
+  enum {
+    ProductIsValid =  Derived::ColsAtCompileTime==Dynamic
+                   || OtherDerived::RowsAtCompileTime==Dynamic
+                   || int(Derived::ColsAtCompileTime)==int(OtherDerived::RowsAtCompileTime),
+    AreVectors = Derived::IsVectorAtCompileTime && OtherDerived::IsVectorAtCompileTime,
+    SameSizes = EIGEN_PREDICATE_SAME_MATRIX_SIZE(Derived,OtherDerived)
+  };
+  // note to the lost user:
+  //    * for a dot product use: v1.dot(v2)
+  //    * for a coeff-wise product use: v1.cwiseProduct(v2)
+  EIGEN_STATIC_ASSERT(ProductIsValid || !(AreVectors && SameSizes),
+    INVALID_VECTOR_VECTOR_PRODUCT__IF_YOU_WANTED_A_DOT_OR_COEFF_WISE_PRODUCT_YOU_MUST_USE_THE_EXPLICIT_FUNCTIONS)
+  EIGEN_STATIC_ASSERT(ProductIsValid || !(SameSizes && !AreVectors),
+    INVALID_MATRIX_PRODUCT__IF_YOU_WANTED_A_COEFF_WISE_PRODUCT_YOU_MUST_USE_THE_EXPLICIT_FUNCTION)
+  EIGEN_STATIC_ASSERT(ProductIsValid || SameSizes, INVALID_MATRIX_PRODUCT)
+
+  return typename LazyProductReturnType<Derived,OtherDerived>::Type(derived(), other.derived());
+}
+
+#endif // EIGEN_PRODUCT_H

Eigen/src/Core/GenericPacketMath.h

@@ -134,12 +134,12 @@ pdiv(const Packet& a,
 /** \internal \returns the min of \a a and \a b  (coeff-wise) */
 template<typename Packet> inline Packet
 pmin(const Packet& a,
-        const Packet& b) { return std::min(a, b); }
+        const Packet& b) { using std::min; return (min)(a, b); }
 
 /** \internal \returns the max of \a a and \a b  (coeff-wise) */
 template<typename Packet> inline Packet
 pmax(const Packet& a,
-        const Packet& b) { return std::max(a, b); }
+        const Packet& b) { using std::max; return (max)(a, b); }
 
 /** \internal \returns the absolute value of \a a */
 template<typename Packet> inline Packet
@@ -225,15 +225,20 @@ template<typename Packet> inline typename unpacket_traits<Packet>::type predux_m
 template<typename Packet> inline Packet preverse(const Packet& a)
 { return a; }
 
+
+/** \internal \returns \a a with real and imaginary part flipped (for complex type only) */
+template<typename Packet> inline Packet pcplxflip(const Packet& a)
+{ return Packet(imag(a),real(a)); }
+
 /**************************
 * Special math functions
 ***************************/
 
-/** \internal \returns the sin of \a a (coeff-wise) */
+/** \internal \returns the sine of \a a (coeff-wise) */
 template<typename Packet> EIGEN_DECLARE_FUNCTION_ALLOWING_MULTIPLE_DEFINITIONS
 Packet psin(const Packet& a) { return sin(a); }
 
-/** \internal \returns the cos of \a a (coeff-wise) */
+/** \internal \returns the cosine of \a a (coeff-wise) */
 template<typename Packet> EIGEN_DECLARE_FUNCTION_ALLOWING_MULTIPLE_DEFINITIONS
 Packet pcos(const Packet& a) { return cos(a); }
 
@@ -241,6 +246,14 @@ Packet pcos(const Packet& a) { return cos(a); }
 template<typename Packet> EIGEN_DECLARE_FUNCTION_ALLOWING_MULTIPLE_DEFINITIONS
 Packet ptan(const Packet& a) { return tan(a); }
 
+/** \internal \returns the arc sine of \a a (coeff-wise) */
+template<typename Packet> EIGEN_DECLARE_FUNCTION_ALLOWING_MULTIPLE_DEFINITIONS
+Packet pasin(const Packet& a) { return asin(a); }
+
+/** \internal \returns the arc cosine of \a a (coeff-wise) */
+template<typename Packet> EIGEN_DECLARE_FUNCTION_ALLOWING_MULTIPLE_DEFINITIONS
+Packet pacos(const Packet& a) { return acos(a); }
+
 /** \internal \returns the exp of \a a (coeff-wise) */
 template<typename Packet> EIGEN_DECLARE_FUNCTION_ALLOWING_MULTIPLE_DEFINITIONS
 Packet pexp(const Packet& a) { return exp(a); }
@@ -257,6 +270,14 @@ Packet psqrt(const Packet& a) { return sqrt(a); }
 * The following functions might not have to be overwritten for vectorized types
 ***************************************************************************/
 
+/** \internal copy a packet with constant coeficient \a a (e.g., [a,a,a,a]) to \a *to. \a to must be 16 bytes aligned */
+// NOTE: this function must really be templated on the packet type (think about different packet types for the same scalar type)
+template<typename Packet>
+inline void pstore1(typename unpacket_traits<Packet>::type* to, const typename unpacket_traits<Packet>::type& a)
+{
+  pstore(to, pset1<Packet>(a));
+}
+
 /** \internal \returns a * b + c (coeff-wise) */
 template<typename Packet> inline Packet
 pmadd(const Packet&  a,
@@ -265,7 +286,7 @@ pmadd(const Packet&  a,
 { return padd(pmul(a, b),c); }
 
 /** \internal \returns a packet version of \a *from.
-  * \If LoadMode equals Aligned, \a from must be 16 bytes aligned */
+  * If LoadMode equals #Aligned, \a from must be 16 bytes aligned */
 template<typename Packet, int LoadMode>
 inline Packet ploadt(const typename unpacket_traits<Packet>::type* from)
 {
@@ -276,7 +297,7 @@ inline Packet ploadt(const typename unpacket_traits<Packet>::type* from)
 }
 
 /** \internal copy the packet \a from to \a *to.
-  * If StoreMode equals Aligned, \a to must be 16 bytes aligned */
+  * If StoreMode equals #Aligned, \a to must be 16 bytes aligned */
 template<typename Scalar, typename Packet, int LoadMode>
 inline void pstoret(Scalar* to, const Packet& from)
 {

Eigen/src/Core/GlobalFunctions.h

@@ -56,6 +56,8 @@ namespace std
   EIGEN_ARRAY_DECLARE_GLOBAL_STD_UNARY(imag,scalar_imag_op)
   EIGEN_ARRAY_DECLARE_GLOBAL_STD_UNARY(sin,scalar_sin_op)
   EIGEN_ARRAY_DECLARE_GLOBAL_STD_UNARY(cos,scalar_cos_op)
+  EIGEN_ARRAY_DECLARE_GLOBAL_STD_UNARY(asin,scalar_asin_op)
+  EIGEN_ARRAY_DECLARE_GLOBAL_STD_UNARY(acos,scalar_acos_op)
   EIGEN_ARRAY_DECLARE_GLOBAL_STD_UNARY(tan,scalar_tan_op)
   EIGEN_ARRAY_DECLARE_GLOBAL_STD_UNARY(exp,scalar_exp_op)
   EIGEN_ARRAY_DECLARE_GLOBAL_STD_UNARY(log,scalar_log_op)
@@ -77,6 +79,8 @@ namespace Eigen
     EIGEN_ARRAY_DECLARE_GLOBAL_EIGEN_UNARY(imag,scalar_imag_op)
     EIGEN_ARRAY_DECLARE_GLOBAL_EIGEN_UNARY(sin,scalar_sin_op)
     EIGEN_ARRAY_DECLARE_GLOBAL_EIGEN_UNARY(cos,scalar_cos_op)
+    EIGEN_ARRAY_DECLARE_GLOBAL_EIGEN_UNARY(asin,scalar_asin_op)
+    EIGEN_ARRAY_DECLARE_GLOBAL_EIGEN_UNARY(acos,scalar_acos_op)
     EIGEN_ARRAY_DECLARE_GLOBAL_EIGEN_UNARY(tan,scalar_tan_op)
     EIGEN_ARRAY_DECLARE_GLOBAL_EIGEN_UNARY(exp,scalar_exp_op)
     EIGEN_ARRAY_DECLARE_GLOBAL_EIGEN_UNARY(log,scalar_log_op)

Eigen/src/Core/IO.h

@@ -141,7 +141,8 @@ struct significant_decimals_default_impl
   typedef typename NumTraits<Scalar>::Real RealScalar;
   static inline int run()
   {
-    return cast<RealScalar,int>(std::ceil(-log(NumTraits<RealScalar>::epsilon())/log(RealScalar(10))));
+    using std::ceil;
+    return cast<RealScalar,int>(ceil(-log(NumTraits<RealScalar>::epsilon())/log(RealScalar(10))));
   }
 };
 

Eigen/src/Core/Map.h

@@ -31,10 +31,10 @@
   *
   * \brief A matrix or vector expression mapping an existing array of data.
   *
-  * \param PlainObjectType the equivalent matrix type of the mapped data
-  * \param MapOptions specifies whether the pointer is \c Aligned, or \c Unaligned.
-  *                The default is \c Unaligned.
-  * \param StrideType optionnally specifies strides. By default, Map assumes the memory layout
+  * \tparam PlainObjectType the equivalent matrix type of the mapped data
+  * \tparam MapOptions specifies whether the pointer is \c #Aligned, or \c #Unaligned.
+  *                The default is \c #Unaligned.
+  * \tparam StrideType optionally specifies strides. By default, Map assumes the memory layout
   *                   of an ordinary, contiguous array. This can be overridden by specifying strides.
   *                   The type passed here must be a specialization of the Stride template, see examples below.
   *
@@ -72,9 +72,9 @@
   * Example: \include Map_placement_new.cpp
   * Output: \verbinclude Map_placement_new.out
   *
-  * This class is the return type of Matrix::Map() but can also be used directly.
+  * This class is the return type of PlainObjectBase::Map() but can also be used directly.
   *
-  * \sa Matrix::Map(), \ref TopicStorageOrders
+  * \sa PlainObjectBase::Map(), \ref TopicStorageOrders
   */
 
 namespace internal {
@@ -95,7 +95,7 @@ struct traits<Map<PlainObjectType, MapOptions, StrideType> >
     HasNoInnerStride = InnerStrideAtCompileTime == 1,
     HasNoOuterStride = StrideType::OuterStrideAtCompileTime == 0,
     HasNoStride = HasNoInnerStride && HasNoOuterStride,
-    IsAligned = int(int(MapOptions)&Aligned)==Aligned,
+    IsAligned = bool(EIGEN_ALIGN) && ((int(MapOptions)&Aligned)==Aligned),
     IsDynamicSize = PlainObjectType::SizeAtCompileTime==Dynamic,
     KeepsPacketAccess = bool(HasNoInnerStride)
                         && ( bool(IsDynamicSize)
@@ -192,14 +192,14 @@ template<typename _Scalar, int _Rows, int _Cols, int _Options, int _MaxRows, int
 inline Array<_Scalar, _Rows, _Cols, _Options, _MaxRows, _MaxCols>
   ::Array(const Scalar *data)
 {
-  _set_noalias(Eigen::Map<const Array>(data));
+  this->_set_noalias(Eigen::Map<const Array>(data));
 }
 
 template<typename _Scalar, int _Rows, int _Cols, int _Options, int _MaxRows, int _MaxCols>
 inline Matrix<_Scalar, _Rows, _Cols, _Options, _MaxRows, _MaxCols>
   ::Matrix(const Scalar *data)
 {
-  _set_noalias(Eigen::Map<const Matrix>(data));
+  this->_set_noalias(Eigen::Map<const Matrix>(data));
 }
 
 #endif // EIGEN_MAP_H

Eigen/src/Core/MapBase.h

@@ -85,6 +85,8 @@ template<typename Derived> class MapBase<Derived, ReadOnlyAccessors>
     using Base::rowStride;
     using Base::colStride;
 
+    // bug 217 - compile error on ICC 11.1
+    using Base::operator=;
 
     typedef typename Base::CoeffReturnType CoeffReturnType;
 
@@ -168,8 +170,8 @@ template<typename Derived> class MapBase<Derived, ReadOnlyAccessors>
       EIGEN_STATIC_ASSERT(EIGEN_IMPLIES(internal::traits<Derived>::Flags&PacketAccessBit,
                                         internal::inner_stride_at_compile_time<Derived>::ret==1),
                           PACKET_ACCESS_REQUIRES_TO_HAVE_INNER_STRIDE_FIXED_TO_1);
-      eigen_assert(EIGEN_IMPLIES(internal::traits<Derived>::Flags&AlignedBit, (size_t(m_data) % (sizeof(Scalar)*internal::packet_traits<Scalar>::size)) == 0)
-        && "data is not aligned");
+      eigen_assert(EIGEN_IMPLIES(internal::traits<Derived>::Flags&AlignedBit, (size_t(m_data) % 16) == 0)
+                   && "data is not aligned");
     }
 
     PointerType m_data;
@@ -236,7 +238,7 @@ template<typename Derived> class MapBase<Derived, WriteAccessors>
                 (this->m_data + index * innerStride(), x);
     }
 
-    inline MapBase(PointerType data) : Base(data) {}
+    explicit inline MapBase(PointerType data) : Base(data) {}
     inline MapBase(PointerType data, Index size) : Base(data, size) {}
     inline MapBase(PointerType data, Index rows, Index cols) : Base(data, rows, cols) {}
 

Eigen/src/Core/MathFunctions.h

@@ -87,7 +87,8 @@ struct real_impl<std::complex<RealScalar> >
 {
   static inline RealScalar run(const std::complex<RealScalar>& x)
   {
-    return std::real(x);
+    using std::real;
+    return real(x);
   }
 };
 
@@ -122,7 +123,8 @@ struct imag_impl<std::complex<RealScalar> >
 {
   static inline RealScalar run(const std::complex<RealScalar>& x)
   {
-    return std::imag(x);
+    using std::imag;
+    return imag(x);
   }
 };
 
@@ -244,7 +246,8 @@ struct conj_impl<std::complex<RealScalar> >
 {
   static inline std::complex<RealScalar> run(const std::complex<RealScalar>& x)
   {
-    return std::conj(x);
+    using std::conj;
+    return conj(x);
   }
 };
 
@@ -270,7 +273,8 @@ struct abs_impl
   typedef typename NumTraits<Scalar>::Real RealScalar;
   static inline RealScalar run(const Scalar& x)
   {
-    return std::abs(x);
+    using std::abs;
+    return abs(x);
   }
 };
 
@@ -305,7 +309,7 @@ struct abs2_impl<std::complex<RealScalar> >
 {
   static inline RealScalar run(const std::complex<RealScalar>& x)
   {
-    return std::norm(x);
+    return real(x)*real(x) + imag(x)*imag(x);
   }
 };
 
@@ -369,10 +373,12 @@ struct hypot_impl
   typedef typename NumTraits<Scalar>::Real RealScalar;
   static inline RealScalar run(const Scalar& x, const Scalar& y)
   {
+    using std::max;
+    using std::min;
     RealScalar _x = abs(x);
     RealScalar _y = abs(y);
-    RealScalar p = std::max(_x, _y);
-    RealScalar q = std::min(_x, _y);
+    RealScalar p = (max)(_x, _y);
+    RealScalar q = (min)(_x, _y);
     RealScalar qp = q/p;
     return p * sqrt(RealScalar(1) + qp*qp);
   }
@@ -420,7 +426,8 @@ struct sqrt_default_impl
 {
   static inline Scalar run(const Scalar& x)
   {
-    return std::sqrt(x);
+    using std::sqrt;
+    return sqrt(x);
   }
 };
 
@@ -454,194 +461,36 @@ inline EIGEN_MATHFUNC_RETVAL(sqrt, Scalar) sqrt(const Scalar& x)
 }
 
 /****************************************************************************
-* Implementation of exp                                                  *
+* Implementation of standard unary real functions (exp, log, sin, cos, ...  *
 ****************************************************************************/
 
-template<typename Scalar, bool IsInteger>
-struct exp_default_impl
-{
-  static inline Scalar run(const Scalar& x)
-  {
-    return std::exp(x);
+// This macro instanciate all the necessary template mechanism which is common to all unary real functions.
+#define EIGEN_MATHFUNC_STANDARD_REAL_UNARY(NAME) \
+  template<typename Scalar, bool IsInteger> struct NAME##_default_impl {            \
+    static inline Scalar run(const Scalar& x) { using std::NAME; return NAME(x); }  \
+  };                                                                                \
+  template<typename Scalar> struct NAME##_default_impl<Scalar, true> {              \
+    static inline Scalar run(const Scalar&) {                                       \
+      EIGEN_STATIC_ASSERT_NON_INTEGER(Scalar)                                       \
+      return Scalar(0);                                                             \
+    }                                                                               \
+  };                                                                                \
+  template<typename Scalar> struct NAME##_impl                                      \
+    : NAME##_default_impl<Scalar, NumTraits<Scalar>::IsInteger>                     \
+  {};                                                                               \
+  template<typename Scalar> struct NAME##_retval { typedef Scalar type; };          \
+  template<typename Scalar>                                                         \
+  inline EIGEN_MATHFUNC_RETVAL(NAME, Scalar) NAME(const Scalar& x) {                \
+    return EIGEN_MATHFUNC_IMPL(NAME, Scalar)::run(x);                               \
   }
-};
 
-template<typename Scalar>
-struct exp_default_impl<Scalar, true>
-{
-  static inline Scalar run(const Scalar&)
-  {
-    EIGEN_STATIC_ASSERT_NON_INTEGER(Scalar)
-    return Scalar(0);
-  }
-};
-
-template<typename Scalar>
-struct exp_impl : exp_default_impl<Scalar, NumTraits<Scalar>::IsInteger> {};
-
-template<typename Scalar>
-struct exp_retval
-{
-  typedef Scalar type;
-};
-
-template<typename Scalar>
-inline EIGEN_MATHFUNC_RETVAL(exp, Scalar) exp(const Scalar& x)
-{
-  return EIGEN_MATHFUNC_IMPL(exp, Scalar)::run(x);
-}
-
-/****************************************************************************
-* Implementation of cos                                                  *
-****************************************************************************/
-
-template<typename Scalar, bool IsInteger>
-struct cos_default_impl
-{
-  static inline Scalar run(const Scalar& x)
-  {
-    return std::cos(x);
-  }
-};
-
-template<typename Scalar>
-struct cos_default_impl<Scalar, true>
-{
-  static inline Scalar run(const Scalar&)
-  {
-    EIGEN_STATIC_ASSERT_NON_INTEGER(Scalar)
-    return Scalar(0);
-  }
-};
-
-template<typename Scalar>
-struct cos_impl : cos_default_impl<Scalar, NumTraits<Scalar>::IsInteger> {};
-
-template<typename Scalar>
-struct cos_retval
-{
-  typedef Scalar type;
-};
-
-template<typename Scalar>
-inline EIGEN_MATHFUNC_RETVAL(cos, Scalar) cos(const Scalar& x)
-{
-  return EIGEN_MATHFUNC_IMPL(cos, Scalar)::run(x);
-}
-
-/****************************************************************************
-* Implementation of sin                                                  *
-****************************************************************************/
-
-template<typename Scalar, bool IsInteger>
-struct sin_default_impl
-{
-  static inline Scalar run(const Scalar& x)
-  {
-    return std::sin(x);
-  }
-};
-
-template<typename Scalar>
-struct sin_default_impl<Scalar, true>
-{
-  static inline Scalar run(const Scalar&)
-  {
-    EIGEN_STATIC_ASSERT_NON_INTEGER(Scalar)
-    return Scalar(0);
-  }
-};
-
-template<typename Scalar>
-struct sin_impl : sin_default_impl<Scalar, NumTraits<Scalar>::IsInteger> {};
-
-template<typename Scalar>
-struct sin_retval
-{
-  typedef Scalar type;
-};
-
-template<typename Scalar>
-inline EIGEN_MATHFUNC_RETVAL(sin, Scalar) sin(const Scalar& x)
-{
-  return EIGEN_MATHFUNC_IMPL(sin, Scalar)::run(x);
-}
-
-/****************************************************************************
-* Implementation of tan                                                  *
-****************************************************************************/
-
-template<typename Scalar, bool IsInteger>
-struct tan_default_impl
-{
-  static inline Scalar run(const Scalar& x)
-  {
-    return std::tan(x);
-  }
-};
-
-template<typename Scalar>
-struct tan_default_impl<Scalar, true>
-{
-  static inline Scalar run(const Scalar&)
-  {
-    EIGEN_STATIC_ASSERT_NON_INTEGER(Scalar)
-    return Scalar(0);
-  }
-};
-
-template<typename Scalar>
-struct tan_impl : tan_default_impl<Scalar, NumTraits<Scalar>::IsInteger> {};
-
-template<typename Scalar>
-struct tan_retval
-{
-  typedef Scalar type;
-};
-
-template<typename Scalar>
-inline EIGEN_MATHFUNC_RETVAL(tan, Scalar) tan(const Scalar& x)
-{
-  return EIGEN_MATHFUNC_IMPL(tan, Scalar)::run(x);
-}
-
-/****************************************************************************
-* Implementation of log                                                  *
-****************************************************************************/
-
-template<typename Scalar, bool IsInteger>
-struct log_default_impl
-{
-  static inline Scalar run(const Scalar& x)
-  {
-    return std::log(x);
-  }
-};
-
-template<typename Scalar>
-struct log_default_impl<Scalar, true>
-{
-  static inline Scalar run(const Scalar&)
-  {
-    EIGEN_STATIC_ASSERT_NON_INTEGER(Scalar)
-    return Scalar(0);
-  }
-};
-
-template<typename Scalar>
-struct log_impl : log_default_impl<Scalar, NumTraits<Scalar>::IsInteger> {};
-
-template<typename Scalar>
-struct log_retval
-{
-  typedef Scalar type;
-};
-
-template<typename Scalar>
-inline EIGEN_MATHFUNC_RETVAL(log, Scalar) log(const Scalar& x)
-{
-  return EIGEN_MATHFUNC_IMPL(log, Scalar)::run(x);
-}
+EIGEN_MATHFUNC_STANDARD_REAL_UNARY(exp)
+EIGEN_MATHFUNC_STANDARD_REAL_UNARY(log)
+EIGEN_MATHFUNC_STANDARD_REAL_UNARY(sin)
+EIGEN_MATHFUNC_STANDARD_REAL_UNARY(cos)
+EIGEN_MATHFUNC_STANDARD_REAL_UNARY(tan)
+EIGEN_MATHFUNC_STANDARD_REAL_UNARY(asin)
+EIGEN_MATHFUNC_STANDARD_REAL_UNARY(acos)
 
 /****************************************************************************
 * Implementation of atan2                                                *
@@ -653,7 +502,8 @@ struct atan2_default_impl
   typedef Scalar retval;
   static inline Scalar run(const Scalar& x, const Scalar& y)
   {
-    return std::atan2(x, y);
+    using std::atan2;
+    return atan2(x, y);
   }
 };
 
@@ -692,7 +542,8 @@ struct pow_default_impl
   typedef Scalar retval;
   static inline Scalar run(const Scalar& x, const Scalar& y)
   {
-    return std::pow(x, y);
+    using std::pow;
+    return pow(x, y);
   }
 };
 
@@ -884,7 +735,8 @@ struct scalar_fuzzy_default_impl<Scalar, false, false>
   }
   static inline bool isApprox(const Scalar& x, const Scalar& y, const RealScalar& prec)
   {
-    return abs(x - y) <= std::min(abs(x), abs(y)) * prec;
+    using std::min;
+    return abs(x - y) <= (min)(abs(x), abs(y)) * prec;
   }
   static inline bool isApproxOrLessThan(const Scalar& x, const Scalar& y, const RealScalar& prec)
   {
@@ -922,7 +774,8 @@ struct scalar_fuzzy_default_impl<Scalar, true, false>
   }
   static inline bool isApprox(const Scalar& x, const Scalar& y, const RealScalar& prec)
   {
-    return abs2(x - y) <= std::min(abs2(x), abs2(y)) * prec * prec;
+    using std::min;
+    return abs2(x - y) <= (min)(abs2(x), abs2(y)) * prec * prec;
   }
 };
 

Eigen/src/Core/Matrix.h

@@ -43,8 +43,8 @@
   * \tparam _Cols Number of columns, or \b Dynamic
   *
   * The remaining template parameters are optional -- in most cases you don't have to worry about them.
-  * \tparam _Options \anchor matrix_tparam_options A combination of either \b RowMajor or \b ColMajor, and of either
-  *                 \b AutoAlign or \b DontAlign.
+  * \tparam _Options \anchor matrix_tparam_options A combination of either \b #RowMajor or \b #ColMajor, and of either
+  *                 \b #AutoAlign or \b #DontAlign.
   *                 The former controls \ref TopicStorageOrders "storage order", and defaults to column-major. The latter controls alignment, which is required
   *                 for vectorization. It defaults to aligning matrices except for fixed sizes that aren't a multiple of the packet size.
   * \tparam _MaxRows Maximum number of rows. Defaults to \a _Rows (\ref maxrows "note").
@@ -79,6 +79,9 @@
   * m(0, 3) = 3;
   * \endcode
   *
+  * This class can be extended with the help of the plugin mechanism described on the page
+  * \ref TopicCustomizingEigen by defining the preprocessor symbol \c EIGEN_MATRIX_PLUGIN.
+  *
   * <i><b>Some notes:</b></i>
   *
   * <dl>
@@ -150,10 +153,6 @@ class Matrix
 
     typedef typename Base::PlainObject PlainObject;
 
-    enum { NeedsToAlign = (!(Options&DontAlign))
-                          && SizeAtCompileTime!=Dynamic && ((static_cast<int>(sizeof(Scalar))*SizeAtCompileTime)%16)==0 };
-    EIGEN_MAKE_ALIGNED_OPERATOR_NEW_IF(NeedsToAlign)
-
     using Base::base;
     using Base::coeffRef;
 
@@ -412,25 +411,6 @@ EIGEN_MAKE_TYPEDEFS_ALL_SIZES(std::complex<double>, cd)
 
 #undef EIGEN_MAKE_TYPEDEFS_ALL_SIZES
 #undef EIGEN_MAKE_TYPEDEFS
-
-#undef EIGEN_MAKE_TYPEDEFS_LARGE
-
-#define EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE_AND_SIZE(TypeSuffix, SizeSuffix) \
-using Eigen::Matrix##SizeSuffix##TypeSuffix; \
-using Eigen::Vector##SizeSuffix##TypeSuffix; \
-using Eigen::RowVector##SizeSuffix##TypeSuffix;
-
-#define EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE(TypeSuffix) \
-EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE_AND_SIZE(TypeSuffix, 2) \
-EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE_AND_SIZE(TypeSuffix, 3) \
-EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE_AND_SIZE(TypeSuffix, 4) \
-EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE_AND_SIZE(TypeSuffix, X) \
-
-#define EIGEN_USING_MATRIX_TYPEDEFS \
-EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE(i) \
-EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE(f) \
-EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE(d) \
-EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE(cf) \
-EIGEN_USING_MATRIX_TYPEDEFS_FOR_TYPE(cd)
+#undef EIGEN_MAKE_FIXED_TYPEDEFS
 
 #endif // EIGEN_MATRIX_H

Eigen/src/Core/MatrixBase.h

@@ -38,7 +38,7 @@
   * Note that some methods are defined in other modules such as the \ref LU_Module LU module
   * for all functions related to matrix inversions.
   *
-  * \param Derived is the derived type, e.g. a matrix type, or an expression, etc.
+  * \tparam Derived is the derived type, e.g. a matrix type, or an expression, etc.
   *
   * When writing a function taking Eigen objects as argument, if you want your function
   * to take as argument any matrix, vector, or expression, just let it take a
@@ -53,6 +53,9 @@
     }
   * \endcode
   *
+  * This class can be extended with the help of the plugin mechanism described on the page
+  * \ref TopicCustomizingEigen by defining the preprocessor symbol \c EIGEN_MATRIXBASE_PLUGIN.
+  *
   * \sa \ref TopicClassHierarchy
   */
 template<typename Derived> class MatrixBase
@@ -108,7 +111,7 @@ template<typename Derived> class MatrixBase
 
     /** \returns the size of the main diagonal, which is min(rows(),cols()).
       * \sa rows(), cols(), SizeAtCompileTime. */
-    inline Index diagonalSize() const { return std::min(rows(),cols()); }
+    inline Index diagonalSize() const { return (std::min)(rows(),cols()); }
 
     /** \brief The plain matrix type corresponding to this expression.
       *
@@ -462,6 +465,8 @@ template<typename Derived> class MatrixBase
     const MatrixFunctionReturnValue<Derived> sinh() const;
     const MatrixFunctionReturnValue<Derived> cos() const;
     const MatrixFunctionReturnValue<Derived> sin() const;
+    const MatrixSquareRootReturnValue<Derived> sqrt() const;
+    const MatrixLogarithmReturnValue<Derived> log() const;
 
 #ifdef EIGEN2_SUPPORT
     template<typename ProductDerived, typename Lhs, typename Rhs>

Eigen/src/Core/NumTraits.h

@@ -87,8 +87,8 @@ template<typename T> struct GenericNumTraits
     // make sure to override this for floating-point types
     return Real(0);
   }
-  inline static T highest() { return std::numeric_limits<T>::max(); }
-  inline static T lowest()  { return IsInteger ? std::numeric_limits<T>::min() : (-std::numeric_limits<T>::max()); }
+  inline static T highest() { return (std::numeric_limits<T>::max)(); }
+  inline static T lowest()  { return IsInteger ? (std::numeric_limits<T>::min)() : (-(std::numeric_limits<T>::max)()); }
   
 #ifdef EIGEN2_SUPPORT
   enum {

Eigen/src/Core/PlainObjectBase.h

@@ -34,6 +34,19 @@
 
 namespace internal {
 
+template<typename Index>
+EIGEN_ALWAYS_INLINE void check_rows_cols_for_overflow(Index rows, Index cols)
+{
+  // http://hg.mozilla.org/mozilla-central/file/6c8a909977d3/xpcom/ds/CheckedInt.h#l242
+  // we assume Index is signed
+  Index max_index = (size_t(1) << (8 * sizeof(Index) - 1)) - 1; // assume Index is signed
+  bool error = (rows < 0  || cols < 0)  ? true
+             : (rows == 0 || cols == 0) ? false
+                                        : (rows > max_index / cols);
+  if (error)
+    throw_std_bad_alloc();
+}
+
 template <typename Derived, typename OtherDerived = Derived, bool IsVector = static_cast<bool>(Derived::IsVectorAtCompileTime)> struct conservative_resize_like_impl;
 
 template<typename MatrixTypeA, typename MatrixTypeB, bool SwapPointers> struct matrix_swap_impl;
@@ -42,6 +55,10 @@ template<typename MatrixTypeA, typename MatrixTypeB, bool SwapPointers> struct m
 
 /**
   * \brief %Dense storage base class for matrices and arrays.
+  *
+  * This class can be extended with the help of the plugin mechanism described on the page
+  * \ref TopicCustomizingEigen by defining the preprocessor symbol \c EIGEN_PLAINOBJECTBASE_PLUGIN.
+  *
   * \sa \ref TopicClassHierarchy
   */
 template<typename Derived>
@@ -80,14 +97,12 @@ class PlainObjectBase : public internal::dense_xpr_base<Derived>::type
     template<typename StrideType> struct StridedConstMapType { typedef Eigen::Map<const Derived, Unaligned, StrideType> type; };
     template<typename StrideType> struct StridedAlignedMapType { typedef Eigen::Map<Derived, Aligned, StrideType> type; };
     template<typename StrideType> struct StridedConstAlignedMapType { typedef Eigen::Map<const Derived, Aligned, StrideType> type; };
-    
 
   protected:
     DenseStorage<Scalar, Base::MaxSizeAtCompileTime, Base::RowsAtCompileTime, Base::ColsAtCompileTime, Options> m_storage;
 
   public:
-    enum { NeedsToAlign = (!(Options&DontAlign))
-                          && SizeAtCompileTime!=Dynamic && ((static_cast<int>(sizeof(Scalar))*SizeAtCompileTime)%16)==0 };
+    enum { NeedsToAlign = SizeAtCompileTime != Dynamic && (internal::traits<Derived>::Flags & AlignedBit) != 0 };
     EIGEN_MAKE_ALIGNED_OPERATOR_NEW_IF(NeedsToAlign)
 
     Base& base() { return *static_cast<Base*>(this); }
@@ -196,11 +211,13 @@ class PlainObjectBase : public internal::dense_xpr_base<Derived>::type
     EIGEN_STRONG_INLINE void resize(Index rows, Index cols)
     {
       #ifdef EIGEN_INITIALIZE_MATRICES_BY_ZERO
+        internal::check_rows_cols_for_overflow(rows, cols);
         Index size = rows*cols;
         bool size_changed = size != this->size();
         m_storage.resize(size, rows, cols);
         if(size_changed) EIGEN_INITIALIZE_BY_ZERO_IF_THAT_OPTION_IS_ENABLED
       #else
+        internal::check_rows_cols_for_overflow(rows, cols);
         m_storage.resize(rows*cols, rows, cols);
       #endif
     }
@@ -269,6 +286,7 @@ class PlainObjectBase : public internal::dense_xpr_base<Derived>::type
     EIGEN_STRONG_INLINE void resizeLike(const EigenBase<OtherDerived>& _other)
     {
       const OtherDerived& other = _other.derived();
+      internal::check_rows_cols_for_overflow(other.rows(), other.cols());
       const Index othersize = other.rows()*other.cols();
       if(RowsAtCompileTime == 1)
       {
@@ -283,33 +301,47 @@ class PlainObjectBase : public internal::dense_xpr_base<Derived>::type
       else resize(other.rows(), other.cols());
     }
 
-    /** Resizes \c *this to a \a rows x \a cols matrix while leaving old values of \c *this untouched.
+    /** Resizes the matrix to \a rows x \a cols while leaving old values untouched.
       *
-      * This method is intended for dynamic-size matrices. If you only want to change the number
-      * of rows and/or of columns, you can use conservativeResize(NoChange_t, Index),
+      * The method is intended for matrices of dynamic size. If you only want to change the number
+      * of rows and/or of columns, you can use conservativeResize(NoChange_t, Index) or
       * conservativeResize(Index, NoChange_t).
       *
-      * The top-left part of the resized matrix will be the same as the overlapping top-left corner
-      * of \c *this. In case values need to be appended to the matrix they will be uninitialized.
+      * Matrices are resized relative to the top-left element. In case values need to be 
+      * appended to the matrix they will be uninitialized.
       */
     EIGEN_STRONG_INLINE void conservativeResize(Index rows, Index cols)
     {
       internal::conservative_resize_like_impl<Derived>::run(*this, rows, cols);
     }
 
+    /** Resizes the matrix to \a rows x \a cols while leaving old values untouched.
+      *
+      * As opposed to conservativeResize(Index rows, Index cols), this version leaves
+      * the number of columns unchanged.
+      *
+      * In case the matrix is growing, new rows will be uninitialized.
+      */
     EIGEN_STRONG_INLINE void conservativeResize(Index rows, NoChange_t)
     {
       // Note: see the comment in conservativeResize(Index,Index)
       conservativeResize(rows, cols());
     }
 
+    /** Resizes the matrix to \a rows x \a cols while leaving old values untouched.
+      *
+      * As opposed to conservativeResize(Index rows, Index cols), this version leaves
+      * the number of rows unchanged.
+      *
+      * In case the matrix is growing, new columns will be uninitialized.
+      */
     EIGEN_STRONG_INLINE void conservativeResize(NoChange_t, Index cols)
     {
       // Note: see the comment in conservativeResize(Index,Index)
       conservativeResize(rows(), cols);
     }
 
-    /** Resizes \c *this to a vector of length \a size while retaining old values of *this.
+    /** Resizes the vector to \a size while retaining old values.
       *
       * \only_for_vectors. This method does not work for
       * partially dynamic matrices when the static dimension is anything other
@@ -322,6 +354,15 @@ class PlainObjectBase : public internal::dense_xpr_base<Derived>::type
       internal::conservative_resize_like_impl<Derived>::run(*this, size);
     }
 
+    /** Resizes the matrix to \a rows x \a cols of \c other, while leaving old values untouched.
+      *
+      * The method is intended for matrices of dynamic size. If you only want to change the number
+      * of rows and/or of columns, you can use conservativeResize(NoChange_t, Index) or
+      * conservativeResize(Index, NoChange_t).
+      *
+      * Matrices are resized relative to the top-left element. In case values need to be 
+      * appended to the matrix they will copied from \c other.
+      */
     template<typename OtherDerived>
     EIGEN_STRONG_INLINE void conservativeResizeLike(const DenseBase<OtherDerived>& other)
     {
@@ -390,6 +431,7 @@ class PlainObjectBase : public internal::dense_xpr_base<Derived>::type
       : m_storage(other.derived().rows() * other.derived().cols(), other.derived().rows(), other.derived().cols())
     {
       _check_template_params();
+      internal::check_rows_cols_for_overflow(other.derived().rows(), other.derived().cols());
       Base::operator=(other.derived());
     }
 
@@ -398,9 +440,6 @@ class PlainObjectBase : public internal::dense_xpr_base<Derived>::type
       * while the AlignedMap() functions return aligned Map objects and thus should be called only with 16-byte-aligned
       * \a data pointers.
       *
-      * These methods do not allow to specify strides. If you need to specify strides, you have to
-      * use the Map class directly.
-      *
       * \see class Map
       */
     //@{
@@ -555,8 +594,12 @@ class PlainObjectBase : public internal::dense_xpr_base<Derived>::type
     template<typename T0, typename T1>
     EIGEN_STRONG_INLINE void _init2(Index rows, Index cols, typename internal::enable_if<Base::SizeAtCompileTime!=2,T0>::type* = 0)
     {
+      EIGEN_STATIC_ASSERT(bool(NumTraits<T0>::IsInteger) &&
+                          bool(NumTraits<T1>::IsInteger),
+                          FLOATING_POINT_ARGUMENT_PASSED__INTEGER_WAS_EXPECTED)
       eigen_assert(rows >= 0 && (RowsAtCompileTime == Dynamic || RowsAtCompileTime == rows)
              && cols >= 0 && (ColsAtCompileTime == Dynamic || ColsAtCompileTime == cols));
+      internal::check_rows_cols_for_overflow(rows, cols);      
       m_storage.resize(rows*cols,rows,cols);
       EIGEN_INITIALIZE_BY_ZERO_IF_THAT_OPTION_IS_ENABLED
     }
@@ -614,14 +657,15 @@ struct internal::conservative_resize_like_impl
     if ( ( Derived::IsRowMajor && _this.cols() == cols) || // row-major and we change only the number of rows
          (!Derived::IsRowMajor && _this.rows() == rows) )  // column-major and we change only the number of columns
     {
+      internal::check_rows_cols_for_overflow(rows, cols);
       _this.derived().m_storage.conservativeResize(rows*cols,rows,cols);
     }
     else
     {
       // The storage order does not allow us to use reallocation.
       typename Derived::PlainObject tmp(rows,cols);
-      const Index common_rows = std::min(rows, _this.rows());
-      const Index common_cols = std::min(cols, _this.cols());
+      const Index common_rows = (std::min)(rows, _this.rows());
+      const Index common_cols = (std::min)(cols, _this.cols());
       tmp.block(0,0,common_rows,common_cols) = _this.block(0,0,common_rows,common_cols);
       _this.derived().swap(tmp);
     }
@@ -654,8 +698,8 @@ struct internal::conservative_resize_like_impl
     {
       // The storage order does not allow us to use reallocation.
       typename Derived::PlainObject tmp(other);
-      const Index common_rows = std::min(tmp.rows(), _this.rows());
-      const Index common_cols = std::min(tmp.cols(), _this.cols());
+      const Index common_rows = (std::min)(tmp.rows(), _this.rows());
+      const Index common_cols = (std::min)(tmp.cols(), _this.cols());
       tmp.block(0,0,common_rows,common_cols) = _this.block(0,0,common_rows,common_cols);
       _this.derived().swap(tmp);
     }

Eigen/src/Core/Product.h

@@ -1,8 +1,7 @@
 // This file is part of Eigen, a lightweight C++ template library
 // for linear algebra.
 //
-// Copyright (C) 2006-2008 Benoit Jacob <jacob.benoit.1@gmail.com>
-// Copyright (C) 2008 Gael Guennebaud <gael.guennebaud@inria.fr>
+// Copyright (C) 2008-2011 Gael Guennebaud <gael.guennebaud@inria.fr>
 //
 // Eigen is free software; you can redistribute it and/or
 // modify it under the terms of the GNU Lesser General Public
@@ -26,603 +25,103 @@
 #ifndef EIGEN_PRODUCT_H
 #define EIGEN_PRODUCT_H
 
-/** \class GeneralProduct
+template<typename Lhs, typename Rhs> class Product;
+template<typename Lhs, typename Rhs, typename StorageKind> class ProductImpl;
+
+/** \class Product
   * \ingroup Core_Module
   *
-  * \brief Expression of the product of two general matrices or vectors
+  * \brief Expression of the product of two arbitrary matrices or vectors
   *
-  * \param LhsNested the type used to store the left-hand side
-  * \param RhsNested the type used to store the right-hand side
-  * \param ProductMode the type of the product
+  * \param Lhs the type of the left-hand side expression
+  * \param Rhs the type of the right-hand side expression
   *
-  * This class represents an expression of the product of two general matrices.
-  * We call a general matrix, a dense matrix with full storage. For instance,
-  * This excludes triangular, selfadjoint, and sparse matrices.
-  * It is the return type of the operator* between general matrices. Its template
-  * arguments are determined automatically by ProductReturnType. Therefore,
-  * GeneralProduct should never be used direclty. To determine the result type of a
-  * function which involves a matrix product, use ProductReturnType::Type.
+  * This class represents an expression of the product of two arbitrary matrices.
   *
-  * \sa ProductReturnType, MatrixBase::operator*(const MatrixBase<OtherDerived>&)
   */
-template<typename Lhs, typename Rhs, int ProductType = internal::product_type<Lhs,Rhs>::value>
-class GeneralProduct;
-
-enum {
-  Large = 2,
-  Small = 3
-};
 
 namespace internal {
-
-template<int Rows, int Cols, int Depth> struct product_type_selector;
-
-template<int Size, int MaxSize> struct product_size_category
+template<typename Lhs, typename Rhs>
+struct traits<Product<Lhs, Rhs> >
 {
-  enum { is_large = MaxSize == Dynamic ||
-                    Size >= EIGEN_CACHEFRIENDLY_PRODUCT_THRESHOLD,
-         value = is_large  ? Large
-               : Size == 1 ? 1
-                           : Small
+  typedef MatrixXpr XprKind;
+  typedef typename remove_all<Lhs>::type LhsCleaned;
+  typedef typename remove_all<Rhs>::type RhsCleaned;
+  typedef typename scalar_product_traits<typename traits<LhsCleaned>::Scalar, typename traits<RhsCleaned>::Scalar>::ReturnType Scalar;
+  typedef typename promote_storage_type<typename traits<LhsCleaned>::StorageKind,
+                                        typename traits<RhsCleaned>::StorageKind>::ret StorageKind;
+  typedef typename promote_index_type<typename traits<LhsCleaned>::Index,
+                                      typename traits<RhsCleaned>::Index>::type Index;
+  enum {
+    RowsAtCompileTime = LhsCleaned::RowsAtCompileTime,
+    ColsAtCompileTime = RhsCleaned::ColsAtCompileTime,
+    MaxRowsAtCompileTime = LhsCleaned::MaxRowsAtCompileTime,
+    MaxColsAtCompileTime = RhsCleaned::MaxColsAtCompileTime,
+    Flags = (MaxRowsAtCompileTime==1 ? RowMajorBit : 0), // TODO should be no storage order
+    CoeffReadCost = 0 // TODO CoeffReadCost should not be part of the expression traits
   };
 };
-
-template<typename Lhs, typename Rhs> struct product_type
-{
-  typedef typename remove_all<Lhs>::type _Lhs;
-  typedef typename remove_all<Rhs>::type _Rhs;
-  enum {
-    MaxRows  = _Lhs::MaxRowsAtCompileTime,
-    Rows  = _Lhs::RowsAtCompileTime,
-    MaxCols  = _Rhs::MaxColsAtCompileTime,
-    Cols  = _Rhs::ColsAtCompileTime,
-    MaxDepth = EIGEN_SIZE_MIN_PREFER_FIXED(_Lhs::MaxColsAtCompileTime,
-                                           _Rhs::MaxRowsAtCompileTime),
-    Depth = EIGEN_SIZE_MIN_PREFER_FIXED(_Lhs::ColsAtCompileTime,
-                                        _Rhs::RowsAtCompileTime),
-    LargeThreshold = EIGEN_CACHEFRIENDLY_PRODUCT_THRESHOLD
-  };
-
-  // the splitting into different lines of code here, introducing the _select enums and the typedef below,
-  // is to work around an internal compiler error with gcc 4.1 and 4.2.
-private:
-  enum {
-    rows_select = product_size_category<Rows,MaxRows>::value,
-    cols_select = product_size_category<Cols,MaxCols>::value,
-    depth_select = product_size_category<Depth,MaxDepth>::value
-  };
-  typedef product_type_selector<rows_select, cols_select, depth_select> selector;
-
-public:
-  enum {
-    value = selector::ret
-  };
-#ifdef EIGEN_DEBUG_PRODUCT
-  static void debug()
-  {
-      EIGEN_DEBUG_VAR(Rows);
-      EIGEN_DEBUG_VAR(Cols);
-      EIGEN_DEBUG_VAR(Depth);
-      EIGEN_DEBUG_VAR(rows_select);
-      EIGEN_DEBUG_VAR(cols_select);
-      EIGEN_DEBUG_VAR(depth_select);
-      EIGEN_DEBUG_VAR(value);
-  }
-#endif
-};
-
-
-/* The following allows to select the kind of product at compile time
- * based on the three dimensions of the product.
- * This is a compile time mapping from {1,Small,Large}^3 -> {product types} */
-// FIXME I'm not sure the current mapping is the ideal one.
-template<int M, int N>  struct product_type_selector<M,N,1>              { enum { ret = OuterProduct }; };
-template<int Depth>     struct product_type_selector<1,    1,    Depth>  { enum { ret = InnerProduct }; };
-template<>              struct product_type_selector<1,    1,    1>      { enum { ret = InnerProduct }; };
-template<>              struct product_type_selector<Small,1,    Small>  { enum { ret = CoeffBasedProductMode }; };
-template<>              struct product_type_selector<1,    Small,Small>  { enum { ret = CoeffBasedProductMode }; };
-template<>              struct product_type_selector<Small,Small,Small>  { enum { ret = CoeffBasedProductMode }; };
-template<>              struct product_type_selector<Small, Small, 1>    { enum { ret = LazyCoeffBasedProductMode }; };
-template<>              struct product_type_selector<Small, Large, 1>    { enum { ret = LazyCoeffBasedProductMode }; };
-template<>              struct product_type_selector<Large, Small, 1>    { enum { ret = LazyCoeffBasedProductMode }; };
-template<>              struct product_type_selector<1,    Large,Small>  { enum { ret = CoeffBasedProductMode }; };
-template<>              struct product_type_selector<1,    Large,Large>  { enum { ret = GemvProduct }; };
-template<>              struct product_type_selector<1,    Small,Large>  { enum { ret = CoeffBasedProductMode }; };
-template<>              struct product_type_selector<Large,1,    Small>  { enum { ret = CoeffBasedProductMode }; };
-template<>              struct product_type_selector<Large,1,    Large>  { enum { ret = GemvProduct }; };
-template<>              struct product_type_selector<Small,1,    Large>  { enum { ret = CoeffBasedProductMode }; };
-template<>              struct product_type_selector<Small,Small,Large>  { enum { ret = GemmProduct }; };
-template<>              struct product_type_selector<Large,Small,Large>  { enum { ret = GemmProduct }; };
-template<>              struct product_type_selector<Small,Large,Large>  { enum { ret = GemmProduct }; };
-template<>              struct product_type_selector<Large,Large,Large>  { enum { ret = GemmProduct }; };
-template<>              struct product_type_selector<Large,Small,Small>  { enum { ret = GemmProduct }; };
-template<>              struct product_type_selector<Small,Large,Small>  { enum { ret = GemmProduct }; };
-template<>              struct product_type_selector<Large,Large,Small>  { enum { ret = GemmProduct }; };
-
 } // end namespace internal
 
-/** \class ProductReturnType
-  * \ingroup Core_Module
-  *
-  * \brief Helper class to get the correct and optimized returned type of operator*
-  *
-  * \param Lhs the type of the left-hand side
-  * \param Rhs the type of the right-hand side
-  * \param ProductMode the type of the product (determined automatically by internal::product_mode)
-  *
-  * This class defines the typename Type representing the optimized product expression
-  * between two matrix expressions. In practice, using ProductReturnType<Lhs,Rhs>::Type
-  * is the recommended way to define the result type of a function returning an expression
-  * which involve a matrix product. The class Product should never be
-  * used directly.
-  *
-  * \sa class Product, MatrixBase::operator*(const MatrixBase<OtherDerived>&)
-  */
-template<typename Lhs, typename Rhs, int ProductType>
-struct ProductReturnType
-{
-  // TODO use the nested type to reduce instanciations ????
-//   typedef typename internal::nested<Lhs,Rhs::ColsAtCompileTime>::type LhsNested;
-//   typedef typename internal::nested<Rhs,Lhs::RowsAtCompileTime>::type RhsNested;
-
-  typedef GeneralProduct<Lhs/*Nested*/, Rhs/*Nested*/, ProductType> Type;
-};
 
 template<typename Lhs, typename Rhs>
-struct ProductReturnType<Lhs,Rhs,CoeffBasedProductMode>
-{
-  typedef typename internal::nested<Lhs, Rhs::ColsAtCompileTime, typename internal::plain_matrix_type<Lhs>::type >::type LhsNested;
-  typedef typename internal::nested<Rhs, Lhs::RowsAtCompileTime, typename internal::plain_matrix_type<Rhs>::type >::type RhsNested;
-  typedef CoeffBasedProduct<LhsNested, RhsNested, EvalBeforeAssigningBit | EvalBeforeNestingBit> Type;
-};
-
-template<typename Lhs, typename Rhs>
-struct ProductReturnType<Lhs,Rhs,LazyCoeffBasedProductMode>
-{
-  typedef typename internal::nested<Lhs, Rhs::ColsAtCompileTime, typename internal::plain_matrix_type<Lhs>::type >::type LhsNested;
-  typedef typename internal::nested<Rhs, Lhs::RowsAtCompileTime, typename internal::plain_matrix_type<Rhs>::type >::type RhsNested;
-  typedef CoeffBasedProduct<LhsNested, RhsNested, NestByRefBit> Type;
-};
-
-// this is a workaround for sun CC
-template<typename Lhs, typename Rhs>
-struct LazyProductReturnType : public ProductReturnType<Lhs,Rhs,LazyCoeffBasedProductMode>
-{};
-
-/***********************************************************************
-*  Implementation of Inner Vector Vector Product
-***********************************************************************/
-
-// FIXME : maybe the "inner product" could return a Scalar
-// instead of a 1x1 matrix ??
-// Pro: more natural for the user
-// Cons: this could be a problem if in a meta unrolled algorithm a matrix-matrix
-// product ends up to a row-vector times col-vector product... To tackle this use
-// case, we could have a specialization for Block<MatrixType,1,1> with: operator=(Scalar x);
-
-namespace internal {
-
-template<typename Lhs, typename Rhs>
-struct traits<GeneralProduct<Lhs,Rhs,InnerProduct> >
- : traits<Matrix<typename scalar_product_traits<typename Lhs::Scalar, typename Rhs::Scalar>::ReturnType,1,1> >
-{};
-
-}
-
-template<typename Lhs, typename Rhs>
-class GeneralProduct<Lhs, Rhs, InnerProduct>
-  : internal::no_assignment_operator,
-    public Matrix<typename internal::scalar_product_traits<typename Lhs::Scalar, typename Rhs::Scalar>::ReturnType,1,1>
-{
-    typedef Matrix<typename internal::scalar_product_traits<typename Lhs::Scalar, typename Rhs::Scalar>::ReturnType,1,1> Base;
-  public:
-    GeneralProduct(const Lhs& lhs, const Rhs& rhs)
-    {
-      EIGEN_STATIC_ASSERT((internal::is_same<typename Lhs::RealScalar, typename Rhs::RealScalar>::value),
-        YOU_MIXED_DIFFERENT_NUMERIC_TYPES__YOU_NEED_TO_USE_THE_CAST_METHOD_OF_MATRIXBASE_TO_CAST_NUMERIC_TYPES_EXPLICITLY)
-
-      Base::coeffRef(0,0) = (lhs.transpose().cwiseProduct(rhs)).sum();
-    }
-
-    /** Convertion to scalar */
-    operator const typename Base::Scalar() const {
-      return Base::coeff(0,0);
-    }
-};
-
-/***********************************************************************
-*  Implementation of Outer Vector Vector Product
-***********************************************************************/
-
-namespace internal {
-template<int StorageOrder> struct outer_product_selector;
-
-template<typename Lhs, typename Rhs>
-struct traits<GeneralProduct<Lhs,Rhs,OuterProduct> >
- : traits<ProductBase<GeneralProduct<Lhs,Rhs,OuterProduct>, Lhs, Rhs> >
-{};
-
-}
-
-template<typename Lhs, typename Rhs>
-class GeneralProduct<Lhs, Rhs, OuterProduct>
-  : public ProductBase<GeneralProduct<Lhs,Rhs,OuterProduct>, Lhs, Rhs>
+class Product : public ProductImpl<Lhs,Rhs,typename internal::promote_storage_type<typename internal::traits<Lhs>::StorageKind,
+                                                                            typename internal::traits<Rhs>::StorageKind>::ret>
 {
   public:
-    EIGEN_PRODUCT_PUBLIC_INTERFACE(GeneralProduct)
-
-    GeneralProduct(const Lhs& lhs, const Rhs& rhs) : Base(lhs,rhs)
-    {
-      EIGEN_STATIC_ASSERT((internal::is_same<typename Lhs::RealScalar, typename Rhs::RealScalar>::value),
-        YOU_MIXED_DIFFERENT_NUMERIC_TYPES__YOU_NEED_TO_USE_THE_CAST_METHOD_OF_MATRIXBASE_TO_CAST_NUMERIC_TYPES_EXPLICITLY)
-    }
-
-    template<typename Dest> void scaleAndAddTo(Dest& dest, Scalar alpha) const
-    {
-      internal::outer_product_selector<(int(Dest::Flags)&RowMajorBit) ? RowMajor : ColMajor>::run(*this, dest, alpha);
-    }
-};
-
-namespace internal {
-
-template<> struct outer_product_selector<ColMajor> {
-  template<typename ProductType, typename Dest>
-  static EIGEN_DONT_INLINE void run(const ProductType& prod, Dest& dest, typename ProductType::Scalar alpha) {
-    typedef typename Dest::Index Index;
-    // FIXME make sure lhs is sequentially stored
-    // FIXME not very good if rhs is real and lhs complex while alpha is real too
-    const Index cols = dest.cols();
-    for (Index j=0; j<cols; ++j)
-      dest.col(j) += (alpha * prod.rhs().coeff(j)) * prod.lhs();
-  }
-};
-
-template<> struct outer_product_selector<RowMajor> {
-  template<typename ProductType, typename Dest>
-  static EIGEN_DONT_INLINE void run(const ProductType& prod, Dest& dest, typename ProductType::Scalar alpha) {
-    typedef typename Dest::Index Index;
-    // FIXME make sure rhs is sequentially stored
-    // FIXME not very good if lhs is real and rhs complex while alpha is real too
-    const Index rows = dest.rows();
-    for (Index i=0; i<rows; ++i)
-      dest.row(i) += (alpha * prod.lhs().coeff(i)) * prod.rhs();
-  }
-};
-
-} // end namespace internal
-
-/***********************************************************************
-*  Implementation of General Matrix Vector Product
-***********************************************************************/
-
-/*  According to the shape/flags of the matrix we have to distinghish 3 different cases:
- *   1 - the matrix is col-major, BLAS compatible and M is large => call fast BLAS-like colmajor routine
- *   2 - the matrix is row-major, BLAS compatible and N is large => call fast BLAS-like rowmajor routine
- *   3 - all other cases are handled using a simple loop along the outer-storage direction.
- *  Therefore we need a lower level meta selector.
- *  Furthermore, if the matrix is the rhs, then the product has to be transposed.
- */
-namespace internal {
-
-template<typename Lhs, typename Rhs>
-struct traits<GeneralProduct<Lhs,Rhs,GemvProduct> >
- : traits<ProductBase<GeneralProduct<Lhs,Rhs,GemvProduct>, Lhs, Rhs> >
-{};
-
-template<int Side, int StorageOrder, bool BlasCompatible>
-struct gemv_selector;
-
-} // end namespace internal
-
-template<typename Lhs, typename Rhs>
-class GeneralProduct<Lhs, Rhs, GemvProduct>
-  : public ProductBase<GeneralProduct<Lhs,Rhs,GemvProduct>, Lhs, Rhs>
-{
-  public:
-    EIGEN_PRODUCT_PUBLIC_INTERFACE(GeneralProduct)
-
-    typedef typename Lhs::Scalar LhsScalar;
-    typedef typename Rhs::Scalar RhsScalar;
-
-    GeneralProduct(const Lhs& lhs, const Rhs& rhs) : Base(lhs,rhs)
-    {
-//       EIGEN_STATIC_ASSERT((internal::is_same<typename Lhs::Scalar, typename Rhs::Scalar>::value),
-//         YOU_MIXED_DIFFERENT_NUMERIC_TYPES__YOU_NEED_TO_USE_THE_CAST_METHOD_OF_MATRIXBASE_TO_CAST_NUMERIC_TYPES_EXPLICITLY)
-    }
-
-    enum { Side = Lhs::IsVectorAtCompileTime ? OnTheLeft : OnTheRight };
-    typedef typename internal::conditional<int(Side)==OnTheRight,_LhsNested,_RhsNested>::type MatrixType;
-
-    template<typename Dest> void scaleAndAddTo(Dest& dst, Scalar alpha) const
-    {
-      eigen_assert(m_lhs.rows() == dst.rows() && m_rhs.cols() == dst.cols());
-      internal::gemv_selector<Side,(int(MatrixType::Flags)&RowMajorBit) ? RowMajor : ColMajor,
-                       bool(internal::blas_traits<MatrixType>::HasUsableDirectAccess)>::run(*this, dst, alpha);
-    }
-};
-
-namespace internal {
-
-// The vector is on the left => transposition
-template<int StorageOrder, bool BlasCompatible>
-struct gemv_selector<OnTheLeft,StorageOrder,BlasCompatible>
-{
-  template<typename ProductType, typename Dest>
-  static void run(const ProductType& prod, Dest& dest, typename ProductType::Scalar alpha)
-  {
-    Transpose<Dest> destT(dest);
-    enum { OtherStorageOrder = StorageOrder == RowMajor ? ColMajor : RowMajor };
-    gemv_selector<OnTheRight,OtherStorageOrder,BlasCompatible>
-      ::run(GeneralProduct<Transpose<const typename ProductType::_RhsNested>,Transpose<const typename ProductType::_LhsNested>, GemvProduct>
-        (prod.rhs().transpose(), prod.lhs().transpose()), destT, alpha);
-  }
-};
-
-template<typename Scalar,int Size,int MaxSize,bool Cond> struct gemv_static_vector_if;
-
-template<typename Scalar,int Size,int MaxSize>
-struct gemv_static_vector_if<Scalar,Size,MaxSize,false>
-{
-  EIGEN_STRONG_INLINE  Scalar* data() { eigen_internal_assert(false && "should never be called"); return 0; }
-};
-
-template<typename Scalar,int Size>
-struct gemv_static_vector_if<Scalar,Size,Dynamic,true>
-{
-  EIGEN_STRONG_INLINE Scalar* data() { return 0; }
-};
-
-template<typename Scalar,int Size,int MaxSize>
-struct gemv_static_vector_if<Scalar,Size,MaxSize,true>
-{
-  internal::plain_array<Scalar,EIGEN_SIZE_MIN_PREFER_FIXED(Size,MaxSize),0> m_data;
-  EIGEN_STRONG_INLINE Scalar* data() { return m_data.array; }
-};
-
-template<> struct gemv_selector<OnTheRight,ColMajor,true>
-{
-  template<typename ProductType, typename Dest>
-  static inline void run(const ProductType& prod, Dest& dest, typename ProductType::Scalar alpha)
-  {
-    typedef typename ProductType::Index Index;
-    typedef typename ProductType::LhsScalar   LhsScalar;
-    typedef typename ProductType::RhsScalar   RhsScalar;
-    typedef typename ProductType::Scalar      ResScalar;
-    typedef typename ProductType::RealScalar  RealScalar;
-    typedef typename ProductType::ActualLhsType ActualLhsType;
-    typedef typename ProductType::ActualRhsType ActualRhsType;
-    typedef typename ProductType::LhsBlasTraits LhsBlasTraits;
-    typedef typename ProductType::RhsBlasTraits RhsBlasTraits;
-    typedef Map<Matrix<ResScalar,Dynamic,1>, Aligned> MappedDest;
-
-    const ActualLhsType actualLhs = LhsBlasTraits::extract(prod.lhs());
-    const ActualRhsType actualRhs = RhsBlasTraits::extract(prod.rhs());
-
-    ResScalar actualAlpha = alpha * LhsBlasTraits::extractScalarFactor(prod.lhs())
-                                  * RhsBlasTraits::extractScalarFactor(prod.rhs());
-
-    enum {
-      // FIXME find a way to allow an inner stride on the result if packet_traits<Scalar>::size==1
-      // on, the other hand it is good for the cache to pack the vector anyways...
-      EvalToDestAtCompileTime = Dest::InnerStrideAtCompileTime==1,
-      ComplexByReal = (NumTraits<LhsScalar>::IsComplex) && (!NumTraits<RhsScalar>::IsComplex),
-      MightCannotUseDest = (Dest::InnerStrideAtCompileTime!=1) || ComplexByReal
-    };
-
-    gemv_static_vector_if<ResScalar,Dest::SizeAtCompileTime,Dest::MaxSizeAtCompileTime,MightCannotUseDest> static_dest;
-
-    bool alphaIsCompatible = (!ComplexByReal) || (imag(actualAlpha)==RealScalar(0));
-    bool evalToDest = EvalToDestAtCompileTime && alphaIsCompatible;
     
-    RhsScalar compatibleAlpha = get_factor<ResScalar,RhsScalar>::run(actualAlpha);
+    typedef typename ProductImpl<
+        Lhs, Rhs,
+        typename internal::promote_storage_type<typename Lhs::StorageKind,
+                                                typename Rhs::StorageKind>::ret>::Base Base;
+    EIGEN_GENERIC_PUBLIC_INTERFACE(Product)
 
-    ResScalar* actualDestPtr;
-    bool freeDestPtr = false;
-    if (evalToDest)
+    typedef typename Lhs::Nested LhsNested;
+    typedef typename Rhs::Nested RhsNested;
+    typedef typename internal::remove_all<LhsNested>::type LhsNestedCleaned;
+    typedef typename internal::remove_all<RhsNested>::type RhsNestedCleaned;
+
+    Product(const Lhs& lhs, const Rhs& rhs) : m_lhs(lhs), m_rhs(rhs)
     {
-      actualDestPtr = &dest.coeffRef(0);
-    }
-    else
-    {
-      #ifdef EIGEN_DENSE_STORAGE_CTOR_PLUGIN
-      int size = dest.size();
-      EIGEN_DENSE_STORAGE_CTOR_PLUGIN
-      #endif
-      if((actualDestPtr = static_dest.data())==0)
-      {
-        freeDestPtr = true;
-        actualDestPtr = ei_aligned_stack_new(ResScalar,dest.size());
-      }
-      if(!alphaIsCompatible)
-      {
-        MappedDest(actualDestPtr, dest.size()).setZero();
-        compatibleAlpha = RhsScalar(1);
-      }
-      else
-        MappedDest(actualDestPtr, dest.size()) = dest;
+      eigen_assert(lhs.cols() == rhs.rows()
+        && "invalid matrix product"
+        && "if you wanted a coeff-wise or a dot product use the respective explicit functions");
     }
 
-    general_matrix_vector_product
-      <Index,LhsScalar,ColMajor,LhsBlasTraits::NeedToConjugate,RhsScalar,RhsBlasTraits::NeedToConjugate>::run(
-        actualLhs.rows(), actualLhs.cols(),
-        &actualLhs.coeffRef(0,0), actualLhs.outerStride(),
-        actualRhs.data(), actualRhs.innerStride(),
-        actualDestPtr, 1,
-        compatibleAlpha);
+    inline Index rows() const { return m_lhs.rows(); }
+    inline Index cols() const { return m_rhs.cols(); }
 
-    if (!evalToDest)
-    {
-      if(!alphaIsCompatible)
-        dest += actualAlpha * MappedDest(actualDestPtr, dest.size());
-      else
-        dest = MappedDest(actualDestPtr, dest.size());
-      if(freeDestPtr) ei_aligned_stack_delete(ResScalar, actualDestPtr, dest.size());
-    }
-  }
+    const LhsNestedCleaned& lhs() const { return m_lhs; }
+    const RhsNestedCleaned& rhs() const { return m_rhs; }
+
+  protected:
+
+    const LhsNested m_lhs;
+    const RhsNested m_rhs;
 };
 
-template<> struct gemv_selector<OnTheRight,RowMajor,true>
+template<typename Lhs, typename Rhs>
+class ProductImpl<Lhs,Rhs,Dense> : public internal::dense_xpr_base<Product<Lhs,Rhs> >::type
 {
-  template<typename ProductType, typename Dest>
-  static void run(const ProductType& prod, Dest& dest, typename ProductType::Scalar alpha)
-  {
-    typedef typename ProductType::LhsScalar LhsScalar;
-    typedef typename ProductType::RhsScalar RhsScalar;
-    typedef typename ProductType::Scalar    ResScalar;
-    typedef typename ProductType::Index Index;
-    typedef typename ProductType::ActualLhsType ActualLhsType;
-    typedef typename ProductType::ActualRhsType ActualRhsType;
-    typedef typename ProductType::_ActualRhsType _ActualRhsType;
-    typedef typename ProductType::LhsBlasTraits LhsBlasTraits;
-    typedef typename ProductType::RhsBlasTraits RhsBlasTraits;
+    typedef Product<Lhs, Rhs> Derived;
+  public:
 
-    typename add_const<ActualLhsType>::type actualLhs = LhsBlasTraits::extract(prod.lhs());
-    typename add_const<ActualRhsType>::type actualRhs = RhsBlasTraits::extract(prod.rhs());
-
-    ResScalar actualAlpha = alpha * LhsBlasTraits::extractScalarFactor(prod.lhs())
-                                  * RhsBlasTraits::extractScalarFactor(prod.rhs());
-
-    enum {
-      // FIXME find a way to allow an inner stride on the result if packet_traits<Scalar>::size==1
-      // on, the other hand it is good for the cache to pack the vector anyways...
-      DirectlyUseRhs = _ActualRhsType::InnerStrideAtCompileTime==1
-    };
-
-    gemv_static_vector_if<RhsScalar,_ActualRhsType::SizeAtCompileTime,_ActualRhsType::MaxSizeAtCompileTime,!DirectlyUseRhs> static_rhs;
-
-    RhsScalar* actualRhsPtr;
-    bool freeRhsPtr = false;
-    if (DirectlyUseRhs)
-    {
-      actualRhsPtr = const_cast<RhsScalar*>(&actualRhs.coeffRef(0));
-    }
-    else
-    {
-      #ifdef EIGEN_DENSE_STORAGE_CTOR_PLUGIN
-      int size = actualRhs.size();
-      EIGEN_DENSE_STORAGE_CTOR_PLUGIN
-      #endif
-      if((actualRhsPtr = static_rhs.data())==0)
-      {
-        freeRhsPtr = true;
-        actualRhsPtr = ei_aligned_stack_new(RhsScalar, actualRhs.size());
-      }
-      Map<typename _ActualRhsType::PlainObject>(actualRhsPtr, actualRhs.size()) = actualRhs;
-    }
-
-    general_matrix_vector_product
-      <Index,LhsScalar,RowMajor,LhsBlasTraits::NeedToConjugate,RhsScalar,RhsBlasTraits::NeedToConjugate>::run(
-        actualLhs.rows(), actualLhs.cols(),
-        &actualLhs.coeffRef(0,0), actualLhs.outerStride(),
-        actualRhsPtr, 1,
-        &dest.coeffRef(0,0), dest.innerStride(),
-        actualAlpha);
-
-    if((!DirectlyUseRhs) && freeRhsPtr) ei_aligned_stack_delete(RhsScalar, actualRhsPtr, prod.rhs().size());
-  }
+    typedef typename internal::dense_xpr_base<Product<Lhs, Rhs> >::type Base;
+    EIGEN_DENSE_PUBLIC_INTERFACE(Derived)
 };
 
-template<> struct gemv_selector<OnTheRight,ColMajor,false>
-{
-  template<typename ProductType, typename Dest>
-  static void run(const ProductType& prod, Dest& dest, typename ProductType::Scalar alpha)
-  {
-    typedef typename Dest::Index Index;
-    // TODO makes sure dest is sequentially stored in memory, otherwise use a temp
-    const Index size = prod.rhs().rows();
-    for(Index k=0; k<size; ++k)
-      dest += (alpha*prod.rhs().coeff(k)) * prod.lhs().col(k);
-  }
-};
-
-template<> struct gemv_selector<OnTheRight,RowMajor,false>
-{
-  template<typename ProductType, typename Dest>
-  static void run(const ProductType& prod, Dest& dest, typename ProductType::Scalar alpha)
-  {
-    typedef typename Dest::Index Index;
-    // TODO makes sure rhs is sequentially stored in memory, otherwise use a temp
-    const Index rows = prod.rows();
-    for(Index i=0; i<rows; ++i)
-      dest.coeffRef(i) += alpha * (prod.lhs().row(i).cwiseProduct(prod.rhs().transpose())).sum();
-  }
-};
-
-} // end namespace internal
-
 /***************************************************************************
 * Implementation of matrix base methods
 ***************************************************************************/
 
-/** \returns the matrix product of \c *this and \a other.
-  *
-  * \note If instead of the matrix product you want the coefficient-wise product, see Cwise::operator*().
-  *
-  * \sa lazyProduct(), operator*=(const MatrixBase&), Cwise::operator*()
-  */
-template<typename Derived>
-template<typename OtherDerived>
-inline const typename ProductReturnType<Derived,OtherDerived>::Type
-MatrixBase<Derived>::operator*(const MatrixBase<OtherDerived> &other) const
-{
-  // A note regarding the function declaration: In MSVC, this function will sometimes
-  // not be inlined since DenseStorage is an unwindable object for dynamic
-  // matrices and product types are holding a member to store the result.
-  // Thus it does not help tagging this function with EIGEN_STRONG_INLINE.
-  enum {
-    ProductIsValid =  Derived::ColsAtCompileTime==Dynamic
-                   || OtherDerived::RowsAtCompileTime==Dynamic
-                   || int(Derived::ColsAtCompileTime)==int(OtherDerived::RowsAtCompileTime),
-    AreVectors = Derived::IsVectorAtCompileTime && OtherDerived::IsVectorAtCompileTime,
-    SameSizes = EIGEN_PREDICATE_SAME_MATRIX_SIZE(Derived,OtherDerived)
-  };
-  // note to the lost user:
-  //    * for a dot product use: v1.dot(v2)
-  //    * for a coeff-wise product use: v1.cwiseProduct(v2)
-  EIGEN_STATIC_ASSERT(ProductIsValid || !(AreVectors && SameSizes),
-    INVALID_VECTOR_VECTOR_PRODUCT__IF_YOU_WANTED_A_DOT_OR_COEFF_WISE_PRODUCT_YOU_MUST_USE_THE_EXPLICIT_FUNCTIONS)
-  EIGEN_STATIC_ASSERT(ProductIsValid || !(SameSizes && !AreVectors),
-    INVALID_MATRIX_PRODUCT__IF_YOU_WANTED_A_COEFF_WISE_PRODUCT_YOU_MUST_USE_THE_EXPLICIT_FUNCTION)
-  EIGEN_STATIC_ASSERT(ProductIsValid || SameSizes, INVALID_MATRIX_PRODUCT)
-#ifdef EIGEN_DEBUG_PRODUCT
-  internal::product_type<Derived,OtherDerived>::debug();
-#endif
-  return typename ProductReturnType<Derived,OtherDerived>::Type(derived(), other.derived());
-}
 
-/** \returns an expression of the matrix product of \c *this and \a other without implicit evaluation.
-  *
-  * The returned product will behave like any other expressions: the coefficients of the product will be
-  * computed once at a time as requested. This might be useful in some extremely rare cases when only
-  * a small and no coherent fraction of the result's coefficients have to be computed.
-  *
-  * \warning This version of the matrix product can be much much slower. So use it only if you know
-  * what you are doing and that you measured a true speed improvement.
-  *
-  * \sa operator*(const MatrixBase&)
+/** \internal used to test the evaluator only
   */
-template<typename Derived>
-template<typename OtherDerived>
-const typename LazyProductReturnType<Derived,OtherDerived>::Type
-MatrixBase<Derived>::lazyProduct(const MatrixBase<OtherDerived> &other) const
+template<typename Lhs,typename Rhs>
+const Product<Lhs,Rhs>
+prod(const Lhs& lhs, const Rhs& rhs)
 {
-  enum {
-    ProductIsValid =  Derived::ColsAtCompileTime==Dynamic
-                   || OtherDerived::RowsAtCompileTime==Dynamic
-                   || int(Derived::ColsAtCompileTime)==int(OtherDerived::RowsAtCompileTime),
-    AreVectors = Derived::IsVectorAtCompileTime && OtherDerived::IsVectorAtCompileTime,
-    SameSizes = EIGEN_PREDICATE_SAME_MATRIX_SIZE(Derived,OtherDerived)
-  };
-  // note to the lost user:
-  //    * for a dot product use: v1.dot(v2)
-  //    * for a coeff-wise product use: v1.cwiseProduct(v2)
-  EIGEN_STATIC_ASSERT(ProductIsValid || !(AreVectors && SameSizes),
-    INVALID_VECTOR_VECTOR_PRODUCT__IF_YOU_WANTED_A_DOT_OR_COEFF_WISE_PRODUCT_YOU_MUST_USE_THE_EXPLICIT_FUNCTIONS)
-  EIGEN_STATIC_ASSERT(ProductIsValid || !(SameSizes && !AreVectors),
-    INVALID_MATRIX_PRODUCT__IF_YOU_WANTED_A_COEFF_WISE_PRODUCT_YOU_MUST_USE_THE_EXPLICIT_FUNCTION)
-  EIGEN_STATIC_ASSERT(ProductIsValid || SameSizes, INVALID_MATRIX_PRODUCT)
-
-  return typename LazyProductReturnType<Derived,OtherDerived>::Type(derived(), other.derived());
+  return Product<Lhs,Rhs>(lhs,rhs);
 }
 
 #endif // EIGEN_PRODUCT_H

Eigen/src/Core/ProductBase.h

@@ -256,16 +256,16 @@ class ScaledProduct
     : Base(prod.lhs(),prod.rhs()), m_prod(prod), m_alpha(x) {}
 
     template<typename Dest>
-    inline void evalTo(Dest& dst) const { dst.setZero(); scaleAndAddTo(dst,m_alpha); }
+    inline void evalTo(Dest& dst) const { dst.setZero(); scaleAndAddTo(dst, Scalar(1)); }
 
     template<typename Dest>
-    inline void addTo(Dest& dst) const { scaleAndAddTo(dst,m_alpha); }
+    inline void addTo(Dest& dst) const { scaleAndAddTo(dst, Scalar(1)); }
 
     template<typename Dest>
-    inline void subTo(Dest& dst) const { scaleAndAddTo(dst,-m_alpha); }
+    inline void subTo(Dest& dst) const { scaleAndAddTo(dst, Scalar(-1)); }
 
     template<typename Dest>
-    inline void scaleAndAddTo(Dest& dst,Scalar alpha) const { m_prod.derived().scaleAndAddTo(dst,alpha); }
+    inline void scaleAndAddTo(Dest& dst,Scalar alpha) const { m_prod.derived().scaleAndAddTo(dst,alpha * m_alpha); }
 
     const Scalar& alpha() const { return m_alpha; }
     

Eigen/src/Core/Redux.h

@@ -216,18 +216,31 @@ struct redux_impl<Func, Derived, LinearVectorizedTraversal, NoUnrolling>
     const Index packetSize = packet_traits<Scalar>::size;
     const Index alignedStart = first_aligned(mat);
     enum {
-      alignment = (Derived::Flags & DirectAccessBit) || (Derived::Flags & AlignedBit)
+      alignment = bool(Derived::Flags & DirectAccessBit) || bool(Derived::Flags & AlignedBit)
                 ? Aligned : Unaligned
     };
-    const Index alignedSize = ((size-alignedStart)/packetSize)*packetSize;
-    const Index alignedEnd = alignedStart + alignedSize;
+    const Index alignedSize2 = ((size-alignedStart)/(2*packetSize))*(2*packetSize);
+    const Index alignedSize = ((size-alignedStart)/(packetSize))*(packetSize);
+    const Index alignedEnd2 = alignedStart + alignedSize2;
+    const Index alignedEnd  = alignedStart + alignedSize;
     Scalar res;
     if(alignedSize)
     {
-      PacketScalar packet_res = mat.template packet<alignment>(alignedStart);
-      for(Index index = alignedStart + packetSize; index < alignedEnd; index += packetSize)
-        packet_res = func.packetOp(packet_res, mat.template packet<alignment>(index));
-      res = func.predux(packet_res);
+      PacketScalar packet_res0 = mat.template packet<alignment>(alignedStart);
+      if(alignedSize>packetSize) // we have at least two packets to partly unroll the loop
+      {
+        PacketScalar packet_res1 = mat.template packet<alignment>(alignedStart+packetSize);
+        for(Index index = alignedStart + 2*packetSize; index < alignedEnd2; index += 2*packetSize)
+        {
+          packet_res0 = func.packetOp(packet_res0, mat.template packet<alignment>(index));
+          packet_res1 = func.packetOp(packet_res1, mat.template packet<alignment>(index+packetSize));
+        }
+
+        packet_res0 = func.packetOp(packet_res0,packet_res1);
+        if(alignedEnd>alignedEnd2)
+          packet_res0 = func.packetOp(packet_res0, mat.template packet<alignment>(alignedEnd2));
+      }
+      res = func.predux(packet_res0);
 
       for(Index index = 0; index < alignedStart; ++index)
         res = func(res,mat.coeff(index));

Eigen/src/Core/Replicate.h

@@ -122,6 +122,10 @@ template<typename MatrixType,int RowFactor,int ColFactor> class Replicate
       return m_matrix.template packet<LoadMode>(actual_row, actual_col);
     }
 
+    const typename internal::remove_all<typename MatrixType::Nested>::type& nestedExpression() const 
+    { 
+      return m_matrix; 
+    }
 
   protected:
     const typename MatrixType::Nested m_matrix;

Eigen/src/Core/ReturnByValue.h

@@ -71,9 +71,9 @@ template<typename Derived> class ReturnByValue
 
     template<typename Dest>
     inline void evalTo(Dest& dst) const
-    { static_cast<const Derived* const>(this)->evalTo(dst); }
-    inline Index rows() const { return static_cast<const Derived* const>(this)->rows(); }
-    inline Index cols() const { return static_cast<const Derived* const>(this)->cols(); }
+    { static_cast<const Derived*>(this)->evalTo(dst); }
+    inline Index rows() const { return static_cast<const Derived*>(this)->rows(); }
+    inline Index cols() const { return static_cast<const Derived*>(this)->cols(); }
 
 #ifndef EIGEN_PARSED_BY_DOXYGEN
 #define Unusable YOU_ARE_TRYING_TO_ACCESS_A_SINGLE_COEFFICIENT_IN_A_SPECIAL_EXPRESSION_WHERE_THAT_IS_NOT_ALLOWED_BECAUSE_THAT_WOULD_BE_INEFFICIENT

Eigen/src/Core/Reverse.h

@@ -183,6 +183,12 @@ template<typename MatrixType, int Direction> class Reverse
       m_matrix.const_cast_derived().template writePacket<LoadMode>(m_matrix.size() - index - PacketSize, internal::preverse(x));
     }
 
+    const typename internal::remove_all<typename MatrixType::Nested>::type& 
+    nestedExpression() const 
+    {
+      return m_matrix;
+    }
+
   protected:
     const typename MatrixType::Nested m_matrix;
 };

Eigen/src/Core/Select.h

@@ -101,6 +101,21 @@ class Select : internal::no_assignment_operator,
         return m_else.coeff(i);
     }
 
+    const ConditionMatrixType& conditionMatrix() const
+    {
+      return m_condition;
+    }
+
+    const ThenMatrixType& thenMatrix() const
+    {
+      return m_then;
+    }
+
+    const ElseMatrixType& elseMatrix() const
+    {
+      return m_else;
+    }
+
   protected:
     const typename ConditionMatrixType::Nested m_condition;
     const typename ThenMatrixType::Nested m_then;

Eigen/src/Core/SelfAdjointView.h

@@ -32,13 +32,13 @@
   * \brief Expression of a selfadjoint matrix from a triangular part of a dense matrix
   *
   * \param MatrixType the type of the dense matrix storing the coefficients
-  * \param TriangularPart can be either \c Lower or \c Upper
+  * \param TriangularPart can be either \c #Lower or \c #Upper
   *
   * This class is an expression of a sefladjoint matrix from a triangular part of a matrix
   * with given dense storage of the coefficients. It is the return type of MatrixBase::selfadjointView()
   * and most of the time this is the only way that it is used.
   *
-  * \sa class TriangularBase, MatrixBase::selfAdjointView()
+  * \sa class TriangularBase, MatrixBase::selfadjointView()
   */
 
 namespace internal {

Eigen/src/Core/SelfCwiseBinaryOp.h

@@ -163,6 +163,16 @@ template<typename BinaryOp, typename Lhs, typename Rhs> class SelfCwiseBinaryOp
       return Base::operator=(rhs);
     }
 
+    Lhs& expression() const 
+    { 
+      return m_matrix;
+    }
+
+    const BinaryOp& functor() const 
+    { 
+      return m_functor;
+    }
+
   protected:
     Lhs& m_matrix;
     const BinaryOp& m_functor;

Eigen/src/Core/SolveTriangular.h

@@ -74,26 +74,19 @@ struct triangular_solver_selector<Lhs,Rhs,Side,Mode,NoUnrolling,1>
     // FIXME find a way to allow an inner stride if packet_traits<Scalar>::size==1
 
     bool useRhsDirectly = Rhs::InnerStrideAtCompileTime==1 || rhs.innerStride()==1;
-    RhsScalar* actualRhs;
-    if(useRhsDirectly)
-    {
-      actualRhs = &rhs.coeffRef(0);
-    }
-    else
-    {
-      actualRhs = ei_aligned_stack_new(RhsScalar,rhs.size());
+
+    ei_declare_aligned_stack_constructed_variable(RhsScalar,actualRhs,rhs.size(),
+                                                  (useRhsDirectly ? rhs.data() : 0));
+                                                  
+    if(!useRhsDirectly)
       MappedRhs(actualRhs,rhs.size()) = rhs;
-    }
 
     triangular_solve_vector<LhsScalar, RhsScalar, typename Lhs::Index, Side, Mode, LhsProductTraits::NeedToConjugate,
                             (int(Lhs::Flags) & RowMajorBit) ? RowMajor : ColMajor>
       ::run(actualLhs.cols(), actualLhs.data(), actualLhs.outerStride(), actualRhs);
 
     if(!useRhsDirectly)
-    {
       rhs = MappedRhs(actualRhs, rhs.size());
-      ei_aligned_stack_delete(RhsScalar, actualRhs, rhs.size());
-    }
   }
 };
 

Eigen/src/Core/StableNorm.h

@@ -56,6 +56,7 @@ template<typename Derived>
 inline typename NumTraits<typename internal::traits<Derived>::Scalar>::Real
 MatrixBase<Derived>::stableNorm() const
 {
+  using std::min;
   const Index blockSize = 4096;
   RealScalar scale = 0;
   RealScalar invScale = 1;
@@ -68,7 +69,7 @@ MatrixBase<Derived>::stableNorm() const
   if (bi>0)
     internal::stable_norm_kernel(this->head(bi), ssq, scale, invScale);
   for (; bi<n; bi+=blockSize)
-    internal::stable_norm_kernel(this->segment(bi,std::min(blockSize, n - bi)).template forceAlignedAccessIf<Alignment>(), ssq, scale, invScale);
+    internal::stable_norm_kernel(this->segment(bi,(min)(blockSize, n - bi)).template forceAlignedAccessIf<Alignment>(), ssq, scale, invScale);
   return scale * internal::sqrt(ssq);
 }
 
@@ -85,6 +86,9 @@ template<typename Derived>
 inline typename NumTraits<typename internal::traits<Derived>::Scalar>::Real
 MatrixBase<Derived>::blueNorm() const
 {
+  using std::pow;
+  using std::min;
+  using std::max;
   static Index nmax = -1;
   static RealScalar b1, b2, s1m, s2m, overfl, rbig, relerr;
   if(nmax <= 0)
@@ -99,25 +103,25 @@ MatrixBase<Derived>::blueNorm() const
     // For portability, the PORT subprograms "ilmaeh" and "rlmach"
     // are used. For any specific computer, each of the assignment
     // statements can be replaced
-    nbig  = std::numeric_limits<Index>::max();            // largest integer
+    nbig  = (std::numeric_limits<Index>::max)();            // largest integer
     ibeta = std::numeric_limits<RealScalar>::radix;         // base for floating-point numbers
     it    = std::numeric_limits<RealScalar>::digits;        // number of base-beta digits in mantissa
     iemin = std::numeric_limits<RealScalar>::min_exponent;  // minimum exponent
     iemax = std::numeric_limits<RealScalar>::max_exponent;  // maximum exponent
-    rbig  = std::numeric_limits<RealScalar>::max();         // largest floating-point number
+    rbig  = (std::numeric_limits<RealScalar>::max)();         // largest floating-point number
 
     iexp  = -((1-iemin)/2);
-    b1    = RealScalar(std::pow(RealScalar(ibeta),RealScalar(iexp)));  // lower boundary of midrange
+    b1    = RealScalar(pow(RealScalar(ibeta),RealScalar(iexp)));  // lower boundary of midrange
     iexp  = (iemax + 1 - it)/2;
-    b2    = RealScalar(std::pow(RealScalar(ibeta),RealScalar(iexp)));   // upper boundary of midrange
+    b2    = RealScalar(pow(RealScalar(ibeta),RealScalar(iexp)));   // upper boundary of midrange
 
     iexp  = (2-iemin)/2;
-    s1m   = RealScalar(std::pow(RealScalar(ibeta),RealScalar(iexp)));   // scaling factor for lower range
+    s1m   = RealScalar(pow(RealScalar(ibeta),RealScalar(iexp)));   // scaling factor for lower range
     iexp  = - ((iemax+it)/2);
-    s2m   = RealScalar(std::pow(RealScalar(ibeta),RealScalar(iexp)));   // scaling factor for upper range
+    s2m   = RealScalar(pow(RealScalar(ibeta),RealScalar(iexp)));   // scaling factor for upper range
 
     overfl  = rbig*s2m;             // overflow boundary for abig
-    eps     = RealScalar(std::pow(double(ibeta), 1-it));
+    eps     = RealScalar(pow(double(ibeta), 1-it));
     relerr  = internal::sqrt(eps);         // tolerance for neglecting asml
     abig    = RealScalar(1.0/eps - 1.0);
     if (RealScalar(nbig)>abig)  nmax = int(abig);  // largest safe n
@@ -163,8 +167,8 @@ MatrixBase<Derived>::blueNorm() const
   }
   else
     return internal::sqrt(amed);
-  asml = std::min(abig, amed);
-  abig = std::max(abig, amed);
+  asml = (min)(abig, amed);
+  abig = (max)(abig, amed);
   if(asml <= abig*relerr)
     return abig;
   else

Eigen/src/Core/Swap.h

@@ -119,6 +119,8 @@ template<typename ExpressionType> class SwapWrapper
       _other.template writePacket<LoadMode>(index, tmp);
     }
 
+    ExpressionType& expression() const { return m_expression; }
+
   protected:
     ExpressionType& m_expression;
 };

Eigen/src/Core/Transpose.h

@@ -350,15 +350,14 @@ struct blas_traits<SelfCwiseBinaryOp<BinOp,NestedXpr,Rhs> >
 template<bool DestIsTransposed, typename OtherDerived>
 struct check_transpose_aliasing_compile_time_selector
 {
-  enum { ret = blas_traits<OtherDerived>::IsTransposed != DestIsTransposed
-  };
+  enum { ret = bool(blas_traits<OtherDerived>::IsTransposed) != DestIsTransposed };
 };
 
 template<bool DestIsTransposed, typename BinOp, typename DerivedA, typename DerivedB>
 struct check_transpose_aliasing_compile_time_selector<DestIsTransposed,CwiseBinaryOp<BinOp,DerivedA,DerivedB> >
 {
-  enum { ret =    blas_traits<DerivedA>::IsTransposed != DestIsTransposed
-               || blas_traits<DerivedB>::IsTransposed != DestIsTransposed
+  enum { ret =    bool(blas_traits<DerivedA>::IsTransposed) != DestIsTransposed
+               || bool(blas_traits<DerivedB>::IsTransposed) != DestIsTransposed
   };
 };
 
@@ -367,7 +366,7 @@ struct check_transpose_aliasing_run_time_selector
 {
   static bool run(const Scalar* dest, const OtherDerived& src)
   {
-    return (blas_traits<OtherDerived>::IsTransposed != DestIsTransposed) && (dest!=0 && dest==(Scalar*)extract_data(src));
+    return (bool(blas_traits<OtherDerived>::IsTransposed) != DestIsTransposed) && (dest!=0 && dest==(Scalar*)extract_data(src));
   }
 };
 

Eigen/src/Core/TriangularMatrix.h

@@ -111,6 +111,7 @@ template<typename Derived> class TriangularBase : public EigenBase<Derived>
       EIGEN_ONLY_USED_FOR_DEBUG(col);
       eigen_assert(col>=0 && col<cols() && row>=0 && row<rows());
       const int mode = int(Mode) & ~SelfAdjoint;
+      EIGEN_ONLY_USED_FOR_DEBUG(mode);
       eigen_assert((mode==Upper && col>=row)
                 || (mode==Lower && col<=row)
                 || ((mode==StrictlyUpper || mode==UnitUpper) && col>row)
@@ -134,13 +135,13 @@ template<typename Derived> class TriangularBase : public EigenBase<Derived>
   * \brief Base class for triangular part in a matrix
   *
   * \param MatrixType the type of the object in which we are taking the triangular part
-  * \param Mode the kind of triangular matrix expression to construct. Can be Upper,
-  *             Lower, UpperSelfadjoint, or LowerSelfadjoint. This is in fact a bit field;
-  *             it must have either Upper or Lower, and additionnaly it may have either
-  *             UnitDiag or Selfadjoint.
+  * \param Mode the kind of triangular matrix expression to construct. Can be #Upper,
+  *             #Lower, #UnitUpper, #UnitLower, #StrictlyUpper, or #StrictlyLower.
+  *             This is in fact a bit field; it must have either #Upper or #Lower, 
+  *             and additionnaly it may have #UnitDiag or #ZeroDiag or neither.
   *
   * This class represents a triangular part of a matrix, not necessarily square. Strictly speaking, for rectangular
-  * matrices one should speak ok "trapezoid" parts. This class is the return type
+  * matrices one should speak of "trapezoid" parts. This class is the return type
   * of MatrixBase::triangularView() and most of the time this is the only way it is used.
   *
   * \sa MatrixBase::triangularView()
@@ -448,6 +449,8 @@ struct triangular_assignment_selector
     col = (UnrollCount-1) / Derived1::RowsAtCompileTime,
     row = (UnrollCount-1) % Derived1::RowsAtCompileTime
   };
+  
+  typedef typename Derived1::Scalar Scalar;
 
   inline static void run(Derived1 &dst, const Derived2 &src)
   {
@@ -466,9 +469,9 @@ struct triangular_assignment_selector
     else if(ClearOpposite)
     {
       if (Mode&UnitDiag && row==col)
-        dst.coeffRef(row, col) = 1;
+        dst.coeffRef(row, col) = Scalar(1);
       else
-        dst.coeffRef(row, col) = 0;
+        dst.coeffRef(row, col) = Scalar(0);
     }
   }
 };
@@ -484,16 +487,17 @@ template<typename Derived1, typename Derived2, bool ClearOpposite>
 struct triangular_assignment_selector<Derived1, Derived2, Upper, Dynamic, ClearOpposite>
 {
   typedef typename Derived1::Index Index;
+  typedef typename Derived1::Scalar Scalar;
   inline static void run(Derived1 &dst, const Derived2 &src)
   {
     for(Index j = 0; j < dst.cols(); ++j)
     {
-      Index maxi = std::min(j, dst.rows()-1);
+      Index maxi = (std::min)(j, dst.rows()-1);
       for(Index i = 0; i <= maxi; ++i)
         dst.copyCoeff(i, j, src);
       if (ClearOpposite)
         for(Index i = maxi+1; i < dst.rows(); ++i)
-          dst.coeffRef(i, j) = 0;
+          dst.coeffRef(i, j) = Scalar(0);
     }
   }
 };
@@ -508,10 +512,10 @@ struct triangular_assignment_selector<Derived1, Derived2, Lower, Dynamic, ClearO
     {
       for(Index i = j; i < dst.rows(); ++i)
         dst.copyCoeff(i, j, src);
-      Index maxi = std::min(j, dst.rows());
+      Index maxi = (std::min)(j, dst.rows());
       if (ClearOpposite)
         for(Index i = 0; i < maxi; ++i)
-          dst.coeffRef(i, j) = 0;
+          dst.coeffRef(i, j) = static_cast<typename Derived1::Scalar>(0);
     }
   }
 };
@@ -524,7 +528,7 @@ struct triangular_assignment_selector<Derived1, Derived2, StrictlyUpper, Dynamic
   {
     for(Index j = 0; j < dst.cols(); ++j)
     {
-      Index maxi = std::min(j, dst.rows());
+      Index maxi = (std::min)(j, dst.rows());
       for(Index i = 0; i < maxi; ++i)
         dst.copyCoeff(i, j, src);
       if (ClearOpposite)
@@ -544,10 +548,10 @@ struct triangular_assignment_selector<Derived1, Derived2, StrictlyLower, Dynamic
     {
       for(Index i = j+1; i < dst.rows(); ++i)
         dst.copyCoeff(i, j, src);
-      Index maxi = std::min(j, dst.rows()-1);
+      Index maxi = (std::min)(j, dst.rows()-1);
       if (ClearOpposite)
         for(Index i = 0; i <= maxi; ++i)
-          dst.coeffRef(i, j) = 0;
+          dst.coeffRef(i, j) = static_cast<typename Derived1::Scalar>(0);
     }
   }
 };
@@ -560,7 +564,7 @@ struct triangular_assignment_selector<Derived1, Derived2, UnitUpper, Dynamic, Cl
   {
     for(Index j = 0; j < dst.cols(); ++j)
     {
-      Index maxi = std::min(j, dst.rows());
+      Index maxi = (std::min)(j, dst.rows());
       for(Index i = 0; i < maxi; ++i)
         dst.copyCoeff(i, j, src);
       if (ClearOpposite)
@@ -580,7 +584,7 @@ struct triangular_assignment_selector<Derived1, Derived2, UnitLower, Dynamic, Cl
   {
     for(Index j = 0; j < dst.cols(); ++j)
     {
-      Index maxi = std::min(j, dst.rows());
+      Index maxi = (std::min)(j, dst.rows());
       for(Index i = maxi+1; i < dst.rows(); ++i)
         dst.copyCoeff(i, j, src);
       if (ClearOpposite)
@@ -756,8 +760,8 @@ typename internal::eigen2_part_return_type<Derived, Mode>::type MatrixBase<Deriv
 /**
   * \returns an expression of a triangular view extracted from the current matrix
   *
-  * The parameter \a Mode can have the following values: \c Upper, \c StrictlyUpper, \c UnitUpper,
-  * \c Lower, \c StrictlyLower, \c UnitLower.
+  * The parameter \a Mode can have the following values: \c #Upper, \c #StrictlyUpper, \c #UnitUpper,
+  * \c #Lower, \c #StrictlyLower, \c #UnitLower.
   *
   * Example: \include MatrixBase_extract.cpp
   * Output: \verbinclude MatrixBase_extract.out
@@ -792,7 +796,7 @@ bool MatrixBase<Derived>::isUpperTriangular(RealScalar prec) const
   RealScalar maxAbsOnUpperPart = static_cast<RealScalar>(-1);
   for(Index j = 0; j < cols(); ++j)
   {
-    Index maxi = std::min(j, rows()-1);
+    Index maxi = (std::min)(j, rows()-1);
     for(Index i = 0; i <= maxi; ++i)
     {
       RealScalar absValue = internal::abs(coeff(i,j));
@@ -824,7 +828,7 @@ bool MatrixBase<Derived>::isLowerTriangular(RealScalar prec) const
   RealScalar threshold = maxAbsOnLowerPart * prec;
   for(Index j = 1; j < cols(); ++j)
   {
-    Index maxi = std::min(j, rows()-1);
+    Index maxi = (std::min)(j, rows()-1);
     for(Index i = 0; i < maxi; ++i)
       if(internal::abs(coeff(i, j)) > threshold) return false;
   }

Eigen/src/Core/VectorwiseOp.h

@@ -31,9 +31,9 @@
   *
   * \brief Generic expression of a partially reduxed matrix
   *
-  * \param MatrixType the type of the matrix we are applying the redux operation
-  * \param MemberOp type of the member functor
-  * \param Direction indicates the direction of the redux (Vertical or Horizontal)
+  * \tparam MatrixType the type of the matrix we are applying the redux operation
+  * \tparam MemberOp type of the member functor
+  * \tparam Direction indicates the direction of the redux (#Vertical or #Horizontal)
   *
   * This class represents an expression of a partial redux operator of a matrix.
   * It is the return type of some VectorwiseOp functions,
@@ -164,7 +164,7 @@ struct member_redux {
   * \brief Pseudo expression providing partial reduction operations
   *
   * \param ExpressionType the type of the object on which to do partial reductions
-  * \param Direction indicates the direction of the redux (Vertical or Horizontal)
+  * \param Direction indicates the direction of the redux (#Vertical or #Horizontal)
   *
   * This class represents a pseudo expression with partial reduction features.
   * It is the return type of DenseBase::colwise() and DenseBase::rowwise()
@@ -237,7 +237,10 @@ template<typename ExpressionType, int Direction> class VectorwiseOp
     typename ExtendedType<OtherDerived>::Type
     extendedTo(const DenseBase<OtherDerived>& other) const
     {
-      EIGEN_STATIC_ASSERT_VECTOR_ONLY(OtherDerived);
+      EIGEN_STATIC_ASSERT(EIGEN_IMPLIES(Direction==Vertical, OtherDerived::MaxColsAtCompileTime==1),
+                          YOU_PASSED_A_ROW_VECTOR_BUT_A_COLUMN_VECTOR_WAS_EXPECTED)
+      EIGEN_STATIC_ASSERT(EIGEN_IMPLIES(Direction==Horizontal, OtherDerived::MaxRowsAtCompileTime==1),
+                          YOU_PASSED_A_COLUMN_VECTOR_BUT_A_ROW_VECTOR_WAS_EXPECTED)
       return typename ExtendedType<OtherDerived>::Type
                       (other.derived(),
                        Direction==Vertical   ? 1 : m_matrix.rows(),
@@ -418,10 +421,9 @@ template<typename ExpressionType, int Direction> class VectorwiseOp
     ExpressionType& operator=(const DenseBase<OtherDerived>& other)
     {
       EIGEN_STATIC_ASSERT_VECTOR_ONLY(OtherDerived)
+      EIGEN_STATIC_ASSERT_SAME_XPR_KIND(ExpressionType, OtherDerived)
       //eigen_assert((m_matrix.isNull()) == (other.isNull())); FIXME
-      for(Index j=0; j<subVectors(); ++j)
-        subVector(j) = other;
-      return const_cast<ExpressionType&>(m_matrix);
+      return const_cast<ExpressionType&>(m_matrix = extendedTo(other.derived()));
     }
 
     /** Adds the vector \a other to each subvector of \c *this */
@@ -429,9 +431,8 @@ template<typename ExpressionType, int Direction> class VectorwiseOp
     ExpressionType& operator+=(const DenseBase<OtherDerived>& other)
     {
       EIGEN_STATIC_ASSERT_VECTOR_ONLY(OtherDerived)
-      for(Index j=0; j<subVectors(); ++j)
-        subVector(j) += other.derived();
-      return const_cast<ExpressionType&>(m_matrix);
+      EIGEN_STATIC_ASSERT_SAME_XPR_KIND(ExpressionType, OtherDerived)
+      return const_cast<ExpressionType&>(m_matrix += extendedTo(other.derived()));
     }
 
     /** Substracts the vector \a other to each subvector of \c *this */
@@ -439,8 +440,29 @@ template<typename ExpressionType, int Direction> class VectorwiseOp
     ExpressionType& operator-=(const DenseBase<OtherDerived>& other)
     {
       EIGEN_STATIC_ASSERT_VECTOR_ONLY(OtherDerived)
-      for(Index j=0; j<subVectors(); ++j)
-        subVector(j) -= other.derived();
+      EIGEN_STATIC_ASSERT_SAME_XPR_KIND(ExpressionType, OtherDerived)
+      return const_cast<ExpressionType&>(m_matrix -= extendedTo(other.derived()));
+    }
+
+    /** Multiples each subvector of \c *this by the vector \a other */
+    template<typename OtherDerived>
+    ExpressionType& operator*=(const DenseBase<OtherDerived>& other)
+    {
+      EIGEN_STATIC_ASSERT_VECTOR_ONLY(OtherDerived)
+      EIGEN_STATIC_ASSERT_ARRAYXPR(ExpressionType)
+      EIGEN_STATIC_ASSERT_SAME_XPR_KIND(ExpressionType, OtherDerived)
+      m_matrix *= extendedTo(other.derived());
+      return const_cast<ExpressionType&>(m_matrix);
+    }
+
+    /** Divides each subvector of \c *this by the vector \a other */
+    template<typename OtherDerived>
+    ExpressionType& operator/=(const DenseBase<OtherDerived>& other)
+    {
+      EIGEN_STATIC_ASSERT_VECTOR_ONLY(OtherDerived)
+      EIGEN_STATIC_ASSERT_ARRAYXPR(ExpressionType)
+      EIGEN_STATIC_ASSERT_SAME_XPR_KIND(ExpressionType, OtherDerived)
+      m_matrix /= extendedTo(other.derived());
       return const_cast<ExpressionType&>(m_matrix);
     }
 
@@ -451,7 +473,8 @@ template<typename ExpressionType, int Direction> class VectorwiseOp
                   const typename ExtendedType<OtherDerived>::Type>
     operator+(const DenseBase<OtherDerived>& other) const
     {
-      EIGEN_STATIC_ASSERT_VECTOR_ONLY(OtherDerived);
+      EIGEN_STATIC_ASSERT_VECTOR_ONLY(OtherDerived)
+      EIGEN_STATIC_ASSERT_SAME_XPR_KIND(ExpressionType, OtherDerived)
       return m_matrix + extendedTo(other.derived());
     }
 
@@ -462,10 +485,39 @@ template<typename ExpressionType, int Direction> class VectorwiseOp
                   const typename ExtendedType<OtherDerived>::Type>
     operator-(const DenseBase<OtherDerived>& other) const
     {
-      EIGEN_STATIC_ASSERT_VECTOR_ONLY(OtherDerived);
+      EIGEN_STATIC_ASSERT_VECTOR_ONLY(OtherDerived)
+      EIGEN_STATIC_ASSERT_SAME_XPR_KIND(ExpressionType, OtherDerived)
       return m_matrix - extendedTo(other.derived());
     }
 
+    /** Returns the expression where each subvector is the product of the vector \a other
+      * by the corresponding subvector of \c *this */
+    template<typename OtherDerived> EIGEN_STRONG_INLINE
+    CwiseBinaryOp<internal::scalar_product_op<Scalar>,
+                  const ExpressionTypeNestedCleaned,
+                  const typename ExtendedType<OtherDerived>::Type>
+    operator*(const DenseBase<OtherDerived>& other) const
+    {
+      EIGEN_STATIC_ASSERT_VECTOR_ONLY(OtherDerived)
+      EIGEN_STATIC_ASSERT_ARRAYXPR(ExpressionType)
+      EIGEN_STATIC_ASSERT_SAME_XPR_KIND(ExpressionType, OtherDerived)
+      return m_matrix * extendedTo(other.derived());
+    }
+
+    /** Returns the expression where each subvector is the quotient of the corresponding
+      * subvector of \c *this by the vector \a other */
+    template<typename OtherDerived>
+    CwiseBinaryOp<internal::scalar_quotient_op<Scalar>,
+                  const ExpressionTypeNestedCleaned,
+                  const typename ExtendedType<OtherDerived>::Type>
+    operator/(const DenseBase<OtherDerived>& other) const
+    {
+      EIGEN_STATIC_ASSERT_VECTOR_ONLY(OtherDerived)
+      EIGEN_STATIC_ASSERT_ARRAYXPR(ExpressionType)
+      EIGEN_STATIC_ASSERT_SAME_XPR_KIND(ExpressionType, OtherDerived)
+      return m_matrix / extendedTo(other.derived());
+    }
+
 /////////// Geometry module ///////////
 
     #if EIGEN2_SUPPORT_STAGE > STAGE20_RESOLVE_API_CONFLICTS
@@ -509,7 +561,7 @@ template<typename ExpressionType, int Direction> class VectorwiseOp
   * Example: \include MatrixBase_colwise.cpp
   * Output: \verbinclude MatrixBase_colwise.out
   *
-  * \sa rowwise(), class VectorwiseOp
+  * \sa rowwise(), class VectorwiseOp, \ref TutorialReductionsVisitorsBroadcasting
   */
 template<typename Derived>
 inline const typename DenseBase<Derived>::ConstColwiseReturnType
@@ -520,7 +572,7 @@ DenseBase<Derived>::colwise() const
 
 /** \returns a writable VectorwiseOp wrapper of *this providing additional partial reduction operations
   *
-  * \sa rowwise(), class VectorwiseOp
+  * \sa rowwise(), class VectorwiseOp, \ref TutorialReductionsVisitorsBroadcasting
   */
 template<typename Derived>
 inline typename DenseBase<Derived>::ColwiseReturnType
@@ -534,7 +586,7 @@ DenseBase<Derived>::colwise()
   * Example: \include MatrixBase_rowwise.cpp
   * Output: \verbinclude MatrixBase_rowwise.out
   *
-  * \sa colwise(), class VectorwiseOp
+  * \sa colwise(), class VectorwiseOp, \ref TutorialReductionsVisitorsBroadcasting
   */
 template<typename Derived>
 inline const typename DenseBase<Derived>::ConstRowwiseReturnType
@@ -545,7 +597,7 @@ DenseBase<Derived>::rowwise() const
 
 /** \returns a writable VectorwiseOp wrapper of *this providing additional partial reduction operations
   *
-  * \sa colwise(), class VectorwiseOp
+  * \sa colwise(), class VectorwiseOp, \ref TutorialReductionsVisitorsBroadcasting
   */
 template<typename Derived>
 inline typename DenseBase<Derived>::RowwiseReturnType

Eigen/src/Core/arch/AltiVec/Complex.h

@@ -48,6 +48,7 @@ template<> struct packet_traits<std::complex<float> >  : default_packet_traits
   typedef Packet2cf type;
   enum {
     Vectorizable = 1,
+    AlignedOnScalar = 1,
     size = 2,
 
     HasAdd    = 1,
@@ -69,19 +70,17 @@ template<> EIGEN_STRONG_INLINE Packet2cf pset1<Packet2cf>(const std::complex<flo
 {
   Packet2cf res;
   /* On AltiVec we cannot load 64-bit registers, so wa have to take care of alignment */
-  if ((ptrdiff_t)&from % 16 == 0) {
-    res.v = pload((const float *)&from);
-    res.v = vec_perm(res.v, res.v, p16uc_PSET_HI);
-  } else {
-    res.v = ploadu((const float *)&from);
-    res.v = vec_perm(res.v, res.v, p16uc_PSET_LO);
-  }
+  if((ptrdiff_t(&from) % 16) == 0)
+    res.v = pload<Packet4f>((const float *)&from);
+  else
+    res.v = ploadu<Packet4f>((const float *)&from);
+  res.v = vec_perm(res.v, res.v, p16uc_PSET_HI);
   return res;
 }
 
 template<> EIGEN_STRONG_INLINE Packet2cf padd<Packet2cf>(const Packet2cf& a, const Packet2cf& b) { return Packet2cf(vec_add(a.v,b.v)); }
 template<> EIGEN_STRONG_INLINE Packet2cf psub<Packet2cf>(const Packet2cf& a, const Packet2cf& b) { return Packet2cf(vec_sub(a.v,b.v)); }
-template<> EIGEN_STRONG_INLINE Packet2cf pnegate(const Packet2cf& a) { return Packet2cf(psub<Packet4f>(p4f_ZERO, a.v)); }
+template<> EIGEN_STRONG_INLINE Packet2cf pnegate(const Packet2cf& a) { return Packet2cf(pnegate(a.v)); }
 template<> EIGEN_STRONG_INLINE Packet2cf pconj(const Packet2cf& a) { return Packet2cf((Packet4f)vec_xor((Packet4ui)a.v, p4ui_CONJ_XOR)); }
 
 template<> EIGEN_STRONG_INLINE Packet2cf pmul<Packet2cf>(const Packet2cf& a, const Packet2cf& b)
@@ -108,8 +107,13 @@ template<> EIGEN_STRONG_INLINE Packet2cf por    <Packet2cf>(const Packet2cf& a,
 template<> EIGEN_STRONG_INLINE Packet2cf pxor   <Packet2cf>(const Packet2cf& a, const Packet2cf& b) { return Packet2cf(vec_xor(a.v,b.v)); }
 template<> EIGEN_STRONG_INLINE Packet2cf pandnot<Packet2cf>(const Packet2cf& a, const Packet2cf& b) { return Packet2cf(vec_and(a.v, vec_nor(b.v,b.v))); }
 
-template<> EIGEN_STRONG_INLINE Packet2cf pload <std::complex<float> >(const std::complex<float>* from) { EIGEN_DEBUG_ALIGNED_LOAD return Packet2cf(pload((const float*)from)); }
-template<> EIGEN_STRONG_INLINE Packet2cf ploadu<std::complex<float> >(const std::complex<float>* from) { EIGEN_DEBUG_UNALIGNED_LOAD return Packet2cf(ploadu((const float*)from)); }
+template<> EIGEN_STRONG_INLINE Packet2cf pload <Packet2cf>(const std::complex<float>* from) { EIGEN_DEBUG_ALIGNED_LOAD return Packet2cf(pload<Packet4f>((const float*)from)); }
+template<> EIGEN_STRONG_INLINE Packet2cf ploadu<Packet2cf>(const std::complex<float>* from) { EIGEN_DEBUG_UNALIGNED_LOAD return Packet2cf(ploadu<Packet4f>((const float*)from)); }
+
+template<> EIGEN_STRONG_INLINE Packet2cf ploaddup<Packet2cf>(const std::complex<float>*     from)
+{
+  return pset1<Packet2cf>(*from);
+}
 
 template<> EIGEN_STRONG_INLINE void pstore <std::complex<float> >(std::complex<float> *   to, const Packet2cf& from) { EIGEN_DEBUG_ALIGNED_STORE pstore((float*)to, from.v); }
 template<> EIGEN_STRONG_INLINE void pstoreu<std::complex<float> >(std::complex<float> *   to, const Packet2cf& from) { EIGEN_DEBUG_UNALIGNED_STORE pstoreu((float*)to, from.v); }
@@ -136,7 +140,7 @@ template<> EIGEN_STRONG_INLINE std::complex<float> predux<Packet2cf>(const Packe
   Packet4f b;
   b = (Packet4f) vec_sld(a.v, a.v, 8);
   b = padd(a.v, b);
-  return pfirst(Packet2cf(sum));
+  return pfirst(Packet2cf(b));
 }
 
 template<> EIGEN_STRONG_INLINE Packet2cf preduxp<Packet2cf>(const Packet2cf* vecs)
@@ -180,7 +184,7 @@ template<> struct conj_helper<Packet2cf, Packet2cf, false,true>
 
   EIGEN_STRONG_INLINE Packet2cf pmul(const Packet2cf& a, const Packet2cf& b) const
   {
-    return pmul(a, pconj(b));
+    return internal::pmul(a, pconj(b));
   }
 };
 
@@ -191,7 +195,7 @@ template<> struct conj_helper<Packet2cf, Packet2cf, true,false>
 
   EIGEN_STRONG_INLINE Packet2cf pmul(const Packet2cf& a, const Packet2cf& b) const
   {
-    return pmul(pconj(a), b);
+    return internal::pmul(pconj(a), b);
   }
 };
 
@@ -202,7 +206,7 @@ template<> struct conj_helper<Packet2cf, Packet2cf, true,true>
 
   EIGEN_STRONG_INLINE Packet2cf pmul(const Packet2cf& a, const Packet2cf& b) const
   {
-    return pconj(pmul(a, b));
+    return pconj(internal::pmul(a, b));
   }
 };
 
@@ -214,6 +218,11 @@ template<> EIGEN_STRONG_INLINE Packet2cf pdiv<Packet2cf>(const Packet2cf& a, con
   return Packet2cf(pdiv(res.v, vec_add(s,vec_perm(s, s, p16uc_COMPLEX_REV))));
 }
 
+template<> EIGEN_STRONG_INLINE Packet2cf pcplxflip<Packet2cf>(const Packet2cf& x)
+{
+  return Packet2cf(vec_perm(x.v, x.v, p16uc_COMPLEX_REV));
+}
+
 } // end namespace internal
 
 #endif // EIGEN_COMPLEX_ALTIVEC_H

Eigen/src/Core/arch/AltiVec/PacketMath.h

@@ -73,6 +73,7 @@ static Packet4f p4f_COUNTDOWN = { 3.0, 2.0, 1.0, 0.0 };
 static Packet4i p4i_COUNTDOWN = { 3, 2, 1, 0 };
 static Packet16uc p16uc_REVERSE = {12,13,14,15, 8,9,10,11, 4,5,6,7, 0,1,2,3};
 static Packet16uc p16uc_FORWARD = vec_lvsl(0, (float*)0);
+static Packet16uc p16uc_DUPLICATE = {0,1,2,3, 0,1,2,3, 4,5,6,7, 4,5,6,7};
 
 static _EIGEN_DECLARE_CONST_FAST_Packet4f(ZERO, 0);
 static _EIGEN_DECLARE_CONST_FAST_Packet4i(ZERO, 0);
@@ -292,6 +293,21 @@ template<> EIGEN_STRONG_INLINE Packet4i ploadu<Packet4i>(const int* from)
   return (Packet4i) vec_perm(MSQ, LSQ, mask);    // align the data
 }
 
+template<> EIGEN_STRONG_INLINE Packet4f ploaddup<Packet4f>(const float*   from)
+{
+  Packet4f p;
+  if((ptrdiff_t(&from) % 16) == 0)  p = pload<Packet4f>(from);
+  else                              p = ploadu<Packet4f>(from);
+  return vec_perm(p, p, p16uc_DUPLICATE);
+}
+template<> EIGEN_STRONG_INLINE Packet4i ploaddup<Packet4i>(const int*     from)
+{
+  Packet4i p;
+  if((ptrdiff_t(&from) % 16) == 0)  p = pload<Packet4i>(from);
+  else                              p = ploadu<Packet4i>(from);
+  return vec_perm(p, p, p16uc_DUPLICATE);
+}
+
 template<> EIGEN_STRONG_INLINE void pstore<float>(float*   to, const Packet4f& from) { EIGEN_DEBUG_ALIGNED_STORE vec_st(from, 0, to); }
 template<> EIGEN_STRONG_INLINE void pstore<int>(int*       to, const Packet4i& from) { EIGEN_DEBUG_ALIGNED_STORE vec_st(from, 0, to); }
 

Eigen/src/Core/arch/NEON/Complex.h

@@ -27,8 +27,8 @@
 
 namespace internal {
 
-static uint32x4_t p4ui_CONJ_XOR = { 0x00000000, 0x80000000, 0x00000000, 0x80000000 };
-static uint32x2_t p2ui_CONJ_XOR = { 0x00000000, 0x80000000 };
+static uint32x4_t p4ui_CONJ_XOR = EIGEN_INIT_NEON_PACKET4(0x00000000, 0x80000000, 0x00000000, 0x80000000);
+static uint32x2_t p2ui_CONJ_XOR = EIGEN_INIT_NEON_PACKET2(0x00000000, 0x80000000);
 
 //---------- float ----------
 struct Packet2cf
@@ -43,6 +43,7 @@ template<> struct packet_traits<std::complex<float> >  : default_packet_traits
   typedef Packet2cf type;
   enum {
     Vectorizable = 1,
+    AlignedOnScalar = 1,
     size = 2,
 
     HasAdd    = 1,
@@ -120,6 +121,8 @@ template<> EIGEN_STRONG_INLINE Packet2cf pandnot<Packet2cf>(const Packet2cf& a,
 template<> EIGEN_STRONG_INLINE Packet2cf pload<Packet2cf>(const std::complex<float>* from) { EIGEN_DEBUG_ALIGNED_LOAD return Packet2cf(pload<Packet4f>((const float*)from)); }
 template<> EIGEN_STRONG_INLINE Packet2cf ploadu<Packet2cf>(const std::complex<float>* from) { EIGEN_DEBUG_UNALIGNED_LOAD return Packet2cf(ploadu<Packet4f>((const float*)from)); }
 
+template<> EIGEN_STRONG_INLINE Packet2cf ploaddup<Packet2cf>(const std::complex<float>* from) { return pset1<Packet2cf>(*from); }
+
 template<> EIGEN_STRONG_INLINE void pstore <std::complex<float> >(std::complex<float> *   to, const Packet2cf& from) { EIGEN_DEBUG_ALIGNED_STORE pstore((float*)to, from.v); }
 template<> EIGEN_STRONG_INLINE void pstoreu<std::complex<float> >(std::complex<float> *   to, const Packet2cf& from) { EIGEN_DEBUG_UNALIGNED_STORE pstoreu((float*)to, from.v); }
 
@@ -144,7 +147,7 @@ template<> EIGEN_STRONG_INLINE Packet2cf preverse(const Packet2cf& a)
   return Packet2cf(a_r128);
 }
 
-EIGEN_STRONG_INLINE Packet2cf pcplxflip(const Packet2cf& a)
+template<> EIGEN_STRONG_INLINE Packet2cf pcplxflip<Packet2cf>(const Packet2cf& a)
 {
   return Packet2cf(vrev64q_f32(a.v));
 }
@@ -220,7 +223,7 @@ template<> struct conj_helper<Packet2cf, Packet2cf, false,true>
 
   EIGEN_STRONG_INLINE Packet2cf pmul(const Packet2cf& a, const Packet2cf& b) const
   {
-    return pmul(a, pconj(b));
+    return internal::pmul(a, pconj(b));
   }
 };
 
@@ -231,7 +234,7 @@ template<> struct conj_helper<Packet2cf, Packet2cf, true,false>
 
   EIGEN_STRONG_INLINE Packet2cf pmul(const Packet2cf& a, const Packet2cf& b) const
   {
-    return pmul(pconj(a), b);
+    return internal::pmul(pconj(a), b);
   }
 };
 
@@ -242,7 +245,7 @@ template<> struct conj_helper<Packet2cf, Packet2cf, true,true>
 
   EIGEN_STRONG_INLINE Packet2cf pmul(const Packet2cf& a, const Packet2cf& b) const
   {
-    return pconj(pmul(a, b));
+    return pconj(internal::pmul(a, b));
   }
 };
 

Eigen/src/Core/arch/NEON/PacketMath.h

@@ -41,7 +41,7 @@ namespace internal {
 
 typedef float32x4_t Packet4f;
 typedef int32x4_t   Packet4i;
-typedef uint32x4_t   Packet4ui;
+typedef uint32x4_t  Packet4ui;
 
 #define _EIGEN_DECLARE_CONST_Packet4f(NAME,X) \
   const Packet4f p4f_##NAME = pset1<Packet4f>(X)
@@ -52,6 +52,16 @@ typedef uint32x4_t   Packet4ui;
 #define _EIGEN_DECLARE_CONST_Packet4i(NAME,X) \
   const Packet4i p4i_##NAME = pset1<Packet4i>(X)
 
+#if defined(__llvm__) && !defined(__clang__)
+  //Special treatment for Apple's llvm-gcc, its NEON packet types are unions
+  #define EIGEN_INIT_NEON_PACKET2(X, Y)       {{X, Y}}
+  #define EIGEN_INIT_NEON_PACKET4(X, Y, Z, W) {{X, Y, Z, W}}
+#else
+  //Default initializer for packets
+  #define EIGEN_INIT_NEON_PACKET2(X, Y)       {X, Y}
+  #define EIGEN_INIT_NEON_PACKET4(X, Y, Z, W) {X, Y, Z, W}
+#endif
+    
 #ifndef __pld
 #define __pld(x) asm volatile ( "   pld [%[addr]]\n" :: [addr] "r" (x) : "cc" );
 #endif
@@ -84,8 +94,8 @@ template<> struct packet_traits<int>    : default_packet_traits
   };
 };
 
-#if (defined __GNUC__) && (!(EIGEN_GNUC_AT_LEAST(4,4)))
-// workaround gcc 4.2 and 4.3 compilatin issue
+#if EIGEN_GNUC_AT_MOST(4,4) && !defined(__llvm__)
+// workaround gcc 4.2, 4.3 and 4.4 compilatin issue
 EIGEN_STRONG_INLINE float32x4_t vld1q_f32(const float* x) { return ::vld1q_f32((const float32_t*)x); }
 EIGEN_STRONG_INLINE float32x2_t vld1_f32 (const float* x) { return ::vld1_f32 ((const float32_t*)x); }
 EIGEN_STRONG_INLINE void        vst1q_f32(float* to, float32x4_t from) { ::vst1q_f32((float32_t*)to,from); }
@@ -100,12 +110,12 @@ template<> EIGEN_STRONG_INLINE Packet4i pset1<Packet4i>(const int&    from)   {
 
 template<> EIGEN_STRONG_INLINE Packet4f plset<float>(const float& a)
 {
-  Packet4f countdown = { 3, 2, 1, 0 };
+  Packet4f countdown = EIGEN_INIT_NEON_PACKET4(0, 1, 2, 3);
   return vaddq_f32(pset1<Packet4f>(a), countdown);
 }
 template<> EIGEN_STRONG_INLINE Packet4i plset<int>(const int& a)
 {
-  Packet4i countdown = { 3, 2, 1, 0 };
+  Packet4i countdown = EIGEN_INIT_NEON_PACKET4(0, 1, 2, 3);
   return vaddq_s32(pset1<Packet4i>(a), countdown);
 }
 
@@ -191,14 +201,14 @@ template<> EIGEN_STRONG_INLINE Packet4f ploaddup<Packet4f>(const float*   from)
 {
   float32x2_t lo, hi;
   lo = vdup_n_f32(*from);
-  hi = vdup_n_f32(*from);
+  hi = vdup_n_f32(*(from+1));
   return vcombine_f32(lo, hi);
 }
 template<> EIGEN_STRONG_INLINE Packet4i ploaddup<Packet4i>(const int*     from)
 {
   int32x2_t lo, hi;
   lo = vdup_n_s32(*from);
-  hi = vdup_n_s32(*from);
+  hi = vdup_n_s32(*(from+1));
   return vcombine_s32(lo, hi);
 }
 
@@ -395,25 +405,29 @@ template<> EIGEN_STRONG_INLINE int predux_max<Packet4i>(const Packet4i& a)
   return s[0];
 }
 
-template<int Offset>
-struct palign_impl<Offset,Packet4f>
-{
-  EIGEN_STRONG_INLINE static void run(Packet4f& first, const Packet4f& second)
-  {
-    if (Offset!=0)
-      first = vextq_f32(first, second, Offset);
-  }
-};
+// this PALIGN_NEON business is to work around a bug in LLVM Clang 3.0 causing incorrect compilation errors,
+// see bug 347 and this LLVM bug: http://llvm.org/bugs/show_bug.cgi?id=11074
+#define PALIGN_NEON(Offset,Type,Command) \
+template<>\
+struct palign_impl<Offset,Type>\
+{\
+    EIGEN_STRONG_INLINE static void run(Type& first, const Type& second)\
+    {\
+        if (Offset!=0)\
+            first = Command(first, second, Offset);\
+    }\
+};\
 
-template<int Offset>
-struct palign_impl<Offset,Packet4i>
-{
-  EIGEN_STRONG_INLINE static void run(Packet4i& first, const Packet4i& second)
-  {
-    if (Offset!=0)
-      first = vextq_s32(first, second, Offset);
-  }
-};
+PALIGN_NEON(0,Packet4f,vextq_f32)
+PALIGN_NEON(1,Packet4f,vextq_f32)
+PALIGN_NEON(2,Packet4f,vextq_f32)
+PALIGN_NEON(3,Packet4f,vextq_f32)
+PALIGN_NEON(0,Packet4i,vextq_s32)
+PALIGN_NEON(1,Packet4i,vextq_s32)
+PALIGN_NEON(2,Packet4i,vextq_s32)
+PALIGN_NEON(3,Packet4i,vextq_s32)
+    
+#undef PALIGN_NEON
 
 } // end namespace internal
 

Eigen/src/Core/arch/SSE/Complex.h

@@ -97,23 +97,30 @@ template<> EIGEN_STRONG_INLINE Packet2cf pandnot<Packet2cf>(const Packet2cf& a,
 template<> EIGEN_STRONG_INLINE Packet2cf pload <Packet2cf>(const std::complex<float>* from) { EIGEN_DEBUG_ALIGNED_LOAD return Packet2cf(pload<Packet4f>(&real_ref(*from))); }
 template<> EIGEN_STRONG_INLINE Packet2cf ploadu<Packet2cf>(const std::complex<float>* from) { EIGEN_DEBUG_UNALIGNED_LOAD return Packet2cf(ploadu<Packet4f>(&real_ref(*from))); }
 
+template<> EIGEN_STRONG_INLINE Packet2cf pset1<Packet2cf>(const std::complex<float>&  from)
+{
+  Packet2cf res;
+  #if EIGEN_GNUC_AT_MOST(4,2)
+  // workaround annoying "may be used uninitialized in this function" warning with gcc 4.2
+  res.v = _mm_loadl_pi(_mm_set1_ps(0.0f), (const __m64*)&from);
+  #else
+  res.v = _mm_loadl_pi(res.v, (const __m64*)&from);
+  #endif
+  return Packet2cf(_mm_movelh_ps(res.v,res.v));
+}
+
+template<> EIGEN_STRONG_INLINE Packet2cf ploaddup<Packet2cf>(const std::complex<float>* from) { return pset1<Packet2cf>(*from); }
+
 template<> EIGEN_STRONG_INLINE void pstore <std::complex<float> >(std::complex<float> *   to, const Packet2cf& from) { EIGEN_DEBUG_ALIGNED_STORE pstore(&real_ref(*to), from.v); }
 template<> EIGEN_STRONG_INLINE void pstoreu<std::complex<float> >(std::complex<float> *   to, const Packet2cf& from) { EIGEN_DEBUG_UNALIGNED_STORE pstoreu(&real_ref(*to), from.v); }
 
 template<> EIGEN_STRONG_INLINE void prefetch<std::complex<float> >(const std::complex<float> *   addr) { _mm_prefetch((const char*)(addr), _MM_HINT_T0); }
 
-template<> EIGEN_STRONG_INLINE Packet2cf pset1<Packet2cf>(const std::complex<float>&  from)
-{
-  Packet2cf res;
-  res.v = _mm_loadl_pi(res.v, (const __m64*)&from);
-  return Packet2cf(_mm_movelh_ps(res.v,res.v));
-}
-
 template<> EIGEN_STRONG_INLINE std::complex<float>  pfirst<Packet2cf>(const Packet2cf& a)
 {
-  #if (defined __GNUC__) && (__GNUC__==4) && (__GNUC_MINOR__==2) && (__GNUC_PATCHLEVEL__<=3)
-  // workaround gcc 4.2.1 ICE (mac's gcc version) - I'm not sure how the 4.2.2 and 4.2.3 deal with it, but 4.2.4 works well.
-  // this is not performance wise ideal, but who cares...
+  #if EIGEN_GNUC_AT_MOST(4,3)
+  // Workaround gcc 4.2 ICE - this is not performance wise ideal, but who cares...
+  // This workaround also fix invalid code generation with gcc 4.3
   EIGEN_ALIGN16 std::complex<float> res[2];
   _mm_store_ps((float*)res, a.v);
   return res[0];
@@ -308,6 +315,8 @@ template<> EIGEN_STRONG_INLINE Packet1cd ploadu<Packet1cd>(const std::complex<do
 template<> EIGEN_STRONG_INLINE Packet1cd pset1<Packet1cd>(const std::complex<double>&  from)
 { /* here we really have to use unaligned loads :( */ return ploadu<Packet1cd>(&from); }
 
+template<> EIGEN_STRONG_INLINE Packet1cd ploaddup<Packet1cd>(const std::complex<double>* from) { return pset1<Packet1cd>(*from); }
+
 // FIXME force unaligned store, this is a temporary fix
 template<> EIGEN_STRONG_INLINE void pstore <std::complex<double> >(std::complex<double> *   to, const Packet1cd& from) { EIGEN_DEBUG_ALIGNED_STORE pstore((double*)to, from.v); }
 template<> EIGEN_STRONG_INLINE void pstoreu<std::complex<double> >(std::complex<double> *   to, const Packet1cd& from) { EIGEN_DEBUG_UNALIGNED_STORE pstoreu((double*)to, from.v); }

Eigen/src/Core/arch/SSE/PacketMath.h

@@ -110,23 +110,18 @@ template<> struct unpacket_traits<Packet4f> { typedef float  type; enum {size=4}
 template<> struct unpacket_traits<Packet2d> { typedef double type; enum {size=2}; };
 template<> struct unpacket_traits<Packet4i> { typedef int    type; enum {size=4}; };
 
-#ifdef __GNUC__
-// Sometimes GCC implements _mm_set1_p* using multiple moves,
-// that is inefficient :( (e.g., see gemm_pack_rhs)
-template<> EIGEN_STRONG_INLINE Packet4f pset1<Packet4f>(const float&  from) {
-  Packet4f res = _mm_set_ss(from);
-  return vec4f_swizzle1(res,0,0,0,0);
-}
-template<> EIGEN_STRONG_INLINE Packet2d pset1<Packet2d>(const double&  from) {
-  // NOTE the SSE3 intrinsic _mm_loaddup_pd is never faster but sometimes much slower
-  Packet2d res = _mm_set_sd(from);
-  return vec2d_swizzle1(res, 0, 0);
-}
+#if defined(_MSC_VER) && (_MSC_VER==1500)
+// Workaround MSVC 9 internal compiler error.
+// TODO: It has been detected with win64 builds (amd64), so let's check whether it also happens in 32bits+SSE mode
+// TODO: let's check whether there does not exist a better fix, like adding a pset0() function. (it crashed on pset1(0)).
+template<> EIGEN_STRONG_INLINE Packet4f pset1<Packet4f>(const float&  from) { return _mm_set_ps(from,from,from,from); }
+template<> EIGEN_STRONG_INLINE Packet2d pset1<Packet2d>(const double& from) { return _mm_set_pd(from,from); }
+template<> EIGEN_STRONG_INLINE Packet4i pset1<Packet4i>(const int&    from) { return _mm_set_epi32(from,from,from,from); }
 #else
 template<> EIGEN_STRONG_INLINE Packet4f pset1<Packet4f>(const float&  from) { return _mm_set1_ps(from); }
 template<> EIGEN_STRONG_INLINE Packet2d pset1<Packet2d>(const double& from) { return _mm_set1_pd(from); }
-#endif
 template<> EIGEN_STRONG_INLINE Packet4i pset1<Packet4i>(const int&    from) { return _mm_set1_epi32(from); }
+#endif
 
 template<> EIGEN_STRONG_INLINE Packet4f plset<float>(const float& a) { return _mm_add_ps(pset1<Packet4f>(a), _mm_set_ps(3,2,1,0)); }
 template<> EIGEN_STRONG_INLINE Packet2d plset<double>(const double& a) { return _mm_add_pd(pset1<Packet2d>(a),_mm_set_pd(1,0)); }
@@ -245,29 +240,52 @@ template<> EIGEN_STRONG_INLINE Packet4i pload<Packet4i>(const int*     from) { E
 // a correct instruction dependency.
 // TODO: do the same for MSVC (ICC is compatible)
 // NOTE: with the code below, MSVC's compiler crashes!
+
+#if defined(__GNUC__) && defined(__i386__)
+  // bug 195: gcc/i386 emits weird x87 fldl/fstpl instructions for _mm_load_sd
+  #define EIGEN_AVOID_CUSTOM_UNALIGNED_LOADS 1
+#elif defined(__clang__)
+  // bug 201: Segfaults in __mm_loadh_pd with clang 2.8
+  #define EIGEN_AVOID_CUSTOM_UNALIGNED_LOADS 1
+#else
+  #define EIGEN_AVOID_CUSTOM_UNALIGNED_LOADS 0
+#endif
+
 template<> EIGEN_STRONG_INLINE Packet4f ploadu<Packet4f>(const float* from)
 {
   EIGEN_DEBUG_UNALIGNED_LOAD
+#if EIGEN_AVOID_CUSTOM_UNALIGNED_LOADS
+  return _mm_loadu_ps(from);
+#else
   __m128d res;
   res =  _mm_load_sd((const double*)(from)) ;
   res =  _mm_loadh_pd(res, (const double*)(from+2)) ;
   return _mm_castpd_ps(res);
+#endif
 }
 template<> EIGEN_STRONG_INLINE Packet2d ploadu<Packet2d>(const double* from)
 {
   EIGEN_DEBUG_UNALIGNED_LOAD
+#if EIGEN_AVOID_CUSTOM_UNALIGNED_LOADS
+  return _mm_loadu_pd(from);
+#else
   __m128d res;
   res = _mm_load_sd(from) ;
   res = _mm_loadh_pd(res,from+1);
   return res;
+#endif
 }
 template<> EIGEN_STRONG_INLINE Packet4i ploadu<Packet4i>(const int* from)
 {
   EIGEN_DEBUG_UNALIGNED_LOAD
+#if EIGEN_AVOID_CUSTOM_UNALIGNED_LOADS
+  return _mm_loadu_si128(reinterpret_cast<const Packet4i*>(from));
+#else
   __m128d res;
   res =  _mm_load_sd((const double*)(from)) ;
   res =  _mm_loadh_pd(res, (const double*)(from+2)) ;
   return _mm_castpd_si128(res);
+#endif
 }
 #endif
 
@@ -296,6 +314,19 @@ template<> EIGEN_STRONG_INLINE void pstoreu<double>(double* to, const Packet2d&
 template<> EIGEN_STRONG_INLINE void pstoreu<float>(float*  to, const Packet4f& from) { EIGEN_DEBUG_UNALIGNED_STORE pstoreu((double*)to, _mm_castps_pd(from)); }
 template<> EIGEN_STRONG_INLINE void pstoreu<int>(int*      to, const Packet4i& from) { EIGEN_DEBUG_UNALIGNED_STORE pstoreu((double*)to, _mm_castsi128_pd(from)); }
 
+// some compilers might be tempted to perform multiple moves instead of using a vector path.
+template<> EIGEN_STRONG_INLINE void pstore1<Packet4f>(float* to, const float& a)
+{
+  Packet4f pa = _mm_set_ss(a);
+  pstore(to, vec4f_swizzle1(pa,0,0,0,0));
+}
+// some compilers might be tempted to perform multiple moves instead of using a vector path.
+template<> EIGEN_STRONG_INLINE void pstore1<Packet2d>(double* to, const double& a)
+{
+  Packet2d pa = _mm_set_sd(a);
+  pstore(to, vec2d_swizzle1(pa,0,0));
+}
+
 template<> EIGEN_STRONG_INLINE void prefetch<float>(const float*   addr) { _mm_prefetch((const char*)(addr), _MM_HINT_T0); }
 template<> EIGEN_STRONG_INLINE void prefetch<double>(const double* addr) { _mm_prefetch((const char*)(addr), _MM_HINT_T0); }
 template<> EIGEN_STRONG_INLINE void prefetch<int>(const int*       addr) { _mm_prefetch((const char*)(addr), _MM_HINT_T0); }

Eigen/src/Core/products/GeneralBlockPanelKernel.h

@@ -81,6 +81,7 @@ inline void manage_caching_sizes(Action action, std::ptrdiff_t* l1=0, std::ptrdi
 template<typename LhsScalar, typename RhsScalar, int KcFactor>
 void computeProductBlockingSizes(std::ptrdiff_t& k, std::ptrdiff_t& m, std::ptrdiff_t& n)
 {
+  EIGEN_UNUSED_VARIABLE(n);
   // Explanations:
   // Let's recall the product algorithms form kc x nc horizontal panels B' on the rhs and
   // mc x kc blocks A' on the lhs. A' has to fit into L2 cache. Moreover, B' is processed
@@ -102,7 +103,6 @@ void computeProductBlockingSizes(std::ptrdiff_t& k, std::ptrdiff_t& m, std::ptrd
   k = std::min<std::ptrdiff_t>(k, l1/kdiv);
   std::ptrdiff_t _m = k>0 ? l2/(4 * sizeof(LhsScalar) * k) : 0;
   if(_m<m) m = _m & mr_mask;
-  n = n;
 }
 
 template<typename LhsScalar, typename RhsScalar>
@@ -118,14 +118,14 @@ inline void computeProductBlockingSizes(std::ptrdiff_t& k, std::ptrdiff_t& m, st
   // FIXME (a bit overkill maybe ?)
 
   template<typename CJ, typename A, typename B, typename C, typename T> struct gebp_madd_selector {
-    EIGEN_STRONG_INLINE EIGEN_ALWAYS_INLINE_ATTRIB static void run(const CJ& cj, A& a, B& b, C& c, T& /*t*/)
+    EIGEN_ALWAYS_INLINE static void run(const CJ& cj, A& a, B& b, C& c, T& /*t*/)
     {
       c = cj.pmadd(a,b,c);
     }
   };
 
   template<typename CJ, typename T> struct gebp_madd_selector<CJ,T,T,T,T> {
-    EIGEN_STRONG_INLINE EIGEN_ALWAYS_INLINE_ATTRIB static void run(const CJ& cj, T& a, T& b, T& c, T& t)
+    EIGEN_ALWAYS_INLINE static void run(const CJ& cj, T& a, T& b, T& c, T& t)
     {
       t = b; t = cj.pmul(a,t); c = padd(c,t);
     }
@@ -199,7 +199,7 @@ public:
   EIGEN_STRONG_INLINE void unpackRhs(DenseIndex n, const RhsScalar* rhs, RhsScalar* b)
   {
     for(DenseIndex k=0; k<n; k++)
-      pstore(&b[k*RhsPacketSize], pset1<RhsPacket>(rhs[k]));
+      pstore1<RhsPacket>(&b[k*RhsPacketSize], rhs[k]);
   }
 
   EIGEN_STRONG_INLINE void loadRhs(const RhsScalar* b, RhsPacket& dest) const
@@ -270,7 +270,7 @@ public:
   EIGEN_STRONG_INLINE void unpackRhs(DenseIndex n, const RhsScalar* rhs, RhsScalar* b)
   {
     for(DenseIndex k=0; k<n; k++)
-      pstore(&b[k*RhsPacketSize], pset1<RhsPacket>(rhs[k]));
+      pstore1<RhsPacket>(&b[k*RhsPacketSize], rhs[k]);
   }
 
   EIGEN_STRONG_INLINE void loadRhs(const RhsScalar* b, RhsPacket& dest) const
@@ -363,8 +363,8 @@ public:
     {
       if(Vectorizable)
       {
-        pstore((RealScalar*)&b[k*ResPacketSize*2+0], pset1<RealPacket>(real(rhs[k])));
-        pstore((RealScalar*)&b[k*ResPacketSize*2+ResPacketSize], pset1<RealPacket>(imag(rhs[k])));
+        pstore1<RealPacket>((RealScalar*)&b[k*ResPacketSize*2+0],             real(rhs[k]));
+        pstore1<RealPacket>((RealScalar*)&b[k*ResPacketSize*2+ResPacketSize], imag(rhs[k]));
       }
       else
         b[k] = rhs[k];
@@ -475,7 +475,7 @@ public:
   EIGEN_STRONG_INLINE void unpackRhs(DenseIndex n, const RhsScalar* rhs, RhsScalar* b)
   {
     for(DenseIndex k=0; k<n; k++)
-      pstore(&b[k*RhsPacketSize], pset1<RhsPacket>(rhs[k]));
+      pstore1<RhsPacket>(&b[k*RhsPacketSize], rhs[k]);
   }
 
   EIGEN_STRONG_INLINE void loadRhs(const RhsScalar* b, RhsPacket& dest) const
@@ -536,7 +536,7 @@ struct gebp_kernel
     ResPacketSize = Traits::ResPacketSize
   };
 
-  EIGEN_FLATTEN_ATTRIB
+  EIGEN_DONT_INLINE EIGEN_FLATTEN_ATTRIB
   void operator()(ResScalar* res, Index resStride, const LhsScalar* blockA, const RhsScalar* blockB, Index rows, Index depth, Index cols, ResScalar alpha,
                   Index strideA=-1, Index strideB=-1, Index offsetA=0, Index offsetB=0, RhsScalar* unpackedB = 0)
   {
@@ -598,64 +598,64 @@ struct gebp_kernel
           if(nr==2)
           {
             LhsPacket A0, A1;
-            RhsPacket B0;
+            RhsPacket B_0;
             RhsPacket T0;
             
 EIGEN_ASM_COMMENT("mybegin2");
             traits.loadLhs(&blA[0*LhsProgress], A0);
             traits.loadLhs(&blA[1*LhsProgress], A1);
-            traits.loadRhs(&blB[0*RhsProgress], B0);
-            traits.madd(A0,B0,C0,T0);
-            traits.madd(A1,B0,C4,B0);
-            traits.loadRhs(&blB[1*RhsProgress], B0);
-            traits.madd(A0,B0,C1,T0);
-            traits.madd(A1,B0,C5,B0);
+            traits.loadRhs(&blB[0*RhsProgress], B_0);
+            traits.madd(A0,B_0,C0,T0);
+            traits.madd(A1,B_0,C4,B_0);
+            traits.loadRhs(&blB[1*RhsProgress], B_0);
+            traits.madd(A0,B_0,C1,T0);
+            traits.madd(A1,B_0,C5,B_0);
 
             traits.loadLhs(&blA[2*LhsProgress], A0);
             traits.loadLhs(&blA[3*LhsProgress], A1);
-            traits.loadRhs(&blB[2*RhsProgress], B0);
-            traits.madd(A0,B0,C0,T0);
-            traits.madd(A1,B0,C4,B0);
-            traits.loadRhs(&blB[3*RhsProgress], B0);
-            traits.madd(A0,B0,C1,T0);
-            traits.madd(A1,B0,C5,B0);
+            traits.loadRhs(&blB[2*RhsProgress], B_0);
+            traits.madd(A0,B_0,C0,T0);
+            traits.madd(A1,B_0,C4,B_0);
+            traits.loadRhs(&blB[3*RhsProgress], B_0);
+            traits.madd(A0,B_0,C1,T0);
+            traits.madd(A1,B_0,C5,B_0);
 
             traits.loadLhs(&blA[4*LhsProgress], A0);
             traits.loadLhs(&blA[5*LhsProgress], A1);
-            traits.loadRhs(&blB[4*RhsProgress], B0);
-            traits.madd(A0,B0,C0,T0);
-            traits.madd(A1,B0,C4,B0);
-            traits.loadRhs(&blB[5*RhsProgress], B0);
-            traits.madd(A0,B0,C1,T0);
-            traits.madd(A1,B0,C5,B0);
+            traits.loadRhs(&blB[4*RhsProgress], B_0);
+            traits.madd(A0,B_0,C0,T0);
+            traits.madd(A1,B_0,C4,B_0);
+            traits.loadRhs(&blB[5*RhsProgress], B_0);
+            traits.madd(A0,B_0,C1,T0);
+            traits.madd(A1,B_0,C5,B_0);
 
             traits.loadLhs(&blA[6*LhsProgress], A0);
             traits.loadLhs(&blA[7*LhsProgress], A1);
-            traits.loadRhs(&blB[6*RhsProgress], B0);
-            traits.madd(A0,B0,C0,T0);
-            traits.madd(A1,B0,C4,B0);
-            traits.loadRhs(&blB[7*RhsProgress], B0);
-            traits.madd(A0,B0,C1,T0);
-            traits.madd(A1,B0,C5,B0);
+            traits.loadRhs(&blB[6*RhsProgress], B_0);
+            traits.madd(A0,B_0,C0,T0);
+            traits.madd(A1,B_0,C4,B_0);
+            traits.loadRhs(&blB[7*RhsProgress], B_0);
+            traits.madd(A0,B_0,C1,T0);
+            traits.madd(A1,B_0,C5,B_0);
 EIGEN_ASM_COMMENT("myend");
           }
           else
           {
 EIGEN_ASM_COMMENT("mybegin4");
             LhsPacket A0, A1;
-            RhsPacket B0, B1, B2, B3;
+            RhsPacket B_0, B1, B2, B3;
             RhsPacket T0;
             
             traits.loadLhs(&blA[0*LhsProgress], A0);
             traits.loadLhs(&blA[1*LhsProgress], A1);
-            traits.loadRhs(&blB[0*RhsProgress], B0);
+            traits.loadRhs(&blB[0*RhsProgress], B_0);
             traits.loadRhs(&blB[1*RhsProgress], B1);
 
-            traits.madd(A0,B0,C0,T0);
+            traits.madd(A0,B_0,C0,T0);
             traits.loadRhs(&blB[2*RhsProgress], B2);
-            traits.madd(A1,B0,C4,B0);
+            traits.madd(A1,B_0,C4,B_0);
             traits.loadRhs(&blB[3*RhsProgress], B3);
-            traits.loadRhs(&blB[4*RhsProgress], B0);
+            traits.loadRhs(&blB[4*RhsProgress], B_0);
             traits.madd(A0,B1,C1,T0);
             traits.madd(A1,B1,C5,B1);
             traits.loadRhs(&blB[5*RhsProgress], B1);
@@ -667,9 +667,9 @@ EIGEN_ASM_COMMENT("mybegin4");
             traits.madd(A1,B3,C7,B3);
             traits.loadLhs(&blA[3*LhsProgress], A1);
             traits.loadRhs(&blB[7*RhsProgress], B3);
-            traits.madd(A0,B0,C0,T0);
-            traits.madd(A1,B0,C4,B0);
-            traits.loadRhs(&blB[8*RhsProgress], B0);
+            traits.madd(A0,B_0,C0,T0);
+            traits.madd(A1,B_0,C4,B_0);
+            traits.loadRhs(&blB[8*RhsProgress], B_0);
             traits.madd(A0,B1,C1,T0);
             traits.madd(A1,B1,C5,B1);
             traits.loadRhs(&blB[9*RhsProgress], B1);
@@ -682,9 +682,9 @@ EIGEN_ASM_COMMENT("mybegin4");
             traits.loadLhs(&blA[5*LhsProgress], A1);
             traits.loadRhs(&blB[11*RhsProgress], B3);
 
-            traits.madd(A0,B0,C0,T0);
-            traits.madd(A1,B0,C4,B0);
-            traits.loadRhs(&blB[12*RhsProgress], B0);
+            traits.madd(A0,B_0,C0,T0);
+            traits.madd(A1,B_0,C4,B_0);
+            traits.loadRhs(&blB[12*RhsProgress], B_0);
             traits.madd(A0,B1,C1,T0);
             traits.madd(A1,B1,C5,B1);
             traits.loadRhs(&blB[13*RhsProgress], B1);
@@ -696,8 +696,8 @@ EIGEN_ASM_COMMENT("mybegin4");
             traits.madd(A1,B3,C7,B3);
             traits.loadLhs(&blA[7*LhsProgress], A1);
             traits.loadRhs(&blB[15*RhsProgress], B3);
-            traits.madd(A0,B0,C0,T0);
-            traits.madd(A1,B0,C4,B0);
+            traits.madd(A0,B_0,C0,T0);
+            traits.madd(A1,B_0,C4,B_0);
             traits.madd(A0,B1,C1,T0);
             traits.madd(A1,B1,C5,B1);
             traits.madd(A0,B2,C2,T0);
@@ -715,32 +715,32 @@ EIGEN_ASM_COMMENT("mybegin4");
           if(nr==2)
           {
             LhsPacket A0, A1;
-            RhsPacket B0;
+            RhsPacket B_0;
             RhsPacket T0;
 
             traits.loadLhs(&blA[0*LhsProgress], A0);
             traits.loadLhs(&blA[1*LhsProgress], A1);
-            traits.loadRhs(&blB[0*RhsProgress], B0);
-            traits.madd(A0,B0,C0,T0);
-            traits.madd(A1,B0,C4,B0);
-            traits.loadRhs(&blB[1*RhsProgress], B0);
-            traits.madd(A0,B0,C1,T0);
-            traits.madd(A1,B0,C5,B0);
+            traits.loadRhs(&blB[0*RhsProgress], B_0);
+            traits.madd(A0,B_0,C0,T0);
+            traits.madd(A1,B_0,C4,B_0);
+            traits.loadRhs(&blB[1*RhsProgress], B_0);
+            traits.madd(A0,B_0,C1,T0);
+            traits.madd(A1,B_0,C5,B_0);
           }
           else
           {
             LhsPacket A0, A1;
-            RhsPacket B0, B1, B2, B3;
+            RhsPacket B_0, B1, B2, B3;
             RhsPacket T0;
 
             traits.loadLhs(&blA[0*LhsProgress], A0);
             traits.loadLhs(&blA[1*LhsProgress], A1);
-            traits.loadRhs(&blB[0*RhsProgress], B0);
+            traits.loadRhs(&blB[0*RhsProgress], B_0);
             traits.loadRhs(&blB[1*RhsProgress], B1);
 
-            traits.madd(A0,B0,C0,T0);
+            traits.madd(A0,B_0,C0,T0);
             traits.loadRhs(&blB[2*RhsProgress], B2);
-            traits.madd(A1,B0,C4,B0);
+            traits.madd(A1,B_0,C4,B_0);
             traits.loadRhs(&blB[3*RhsProgress], B3);
             traits.madd(A0,B1,C1,T0);
             traits.madd(A1,B1,C5,B1);
@@ -827,42 +827,42 @@ EIGEN_ASM_COMMENT("mybegin4");
           if(nr==2)
           {
             LhsPacket A0;
-            RhsPacket B0, B1;
+            RhsPacket B_0, B1;
 
             traits.loadLhs(&blA[0*LhsProgress], A0);
-            traits.loadRhs(&blB[0*RhsProgress], B0);
+            traits.loadRhs(&blB[0*RhsProgress], B_0);
             traits.loadRhs(&blB[1*RhsProgress], B1);
-            traits.madd(A0,B0,C0,B0);
-            traits.loadRhs(&blB[2*RhsProgress], B0);
+            traits.madd(A0,B_0,C0,B_0);
+            traits.loadRhs(&blB[2*RhsProgress], B_0);
             traits.madd(A0,B1,C1,B1);
             traits.loadLhs(&blA[1*LhsProgress], A0);
             traits.loadRhs(&blB[3*RhsProgress], B1);
-            traits.madd(A0,B0,C0,B0);
-            traits.loadRhs(&blB[4*RhsProgress], B0);
+            traits.madd(A0,B_0,C0,B_0);
+            traits.loadRhs(&blB[4*RhsProgress], B_0);
             traits.madd(A0,B1,C1,B1);
             traits.loadLhs(&blA[2*LhsProgress], A0);
             traits.loadRhs(&blB[5*RhsProgress], B1);
-            traits.madd(A0,B0,C0,B0);
-            traits.loadRhs(&blB[6*RhsProgress], B0);
+            traits.madd(A0,B_0,C0,B_0);
+            traits.loadRhs(&blB[6*RhsProgress], B_0);
             traits.madd(A0,B1,C1,B1);
             traits.loadLhs(&blA[3*LhsProgress], A0);
             traits.loadRhs(&blB[7*RhsProgress], B1);
-            traits.madd(A0,B0,C0,B0);
+            traits.madd(A0,B_0,C0,B_0);
             traits.madd(A0,B1,C1,B1);
           }
           else
           {
             LhsPacket A0;
-            RhsPacket B0, B1, B2, B3;
+            RhsPacket B_0, B1, B2, B3;
 
             traits.loadLhs(&blA[0*LhsProgress], A0);
-            traits.loadRhs(&blB[0*RhsProgress], B0);
+            traits.loadRhs(&blB[0*RhsProgress], B_0);
             traits.loadRhs(&blB[1*RhsProgress], B1);
 
-            traits.madd(A0,B0,C0,B0);
+            traits.madd(A0,B_0,C0,B_0);
             traits.loadRhs(&blB[2*RhsProgress], B2);
             traits.loadRhs(&blB[3*RhsProgress], B3);
-            traits.loadRhs(&blB[4*RhsProgress], B0);
+            traits.loadRhs(&blB[4*RhsProgress], B_0);
             traits.madd(A0,B1,C1,B1);
             traits.loadRhs(&blB[5*RhsProgress], B1);
             traits.madd(A0,B2,C2,B2);
@@ -870,8 +870,8 @@ EIGEN_ASM_COMMENT("mybegin4");
             traits.madd(A0,B3,C3,B3);
             traits.loadLhs(&blA[1*LhsProgress], A0);
             traits.loadRhs(&blB[7*RhsProgress], B3);
-            traits.madd(A0,B0,C0,B0);
-            traits.loadRhs(&blB[8*RhsProgress], B0);
+            traits.madd(A0,B_0,C0,B_0);
+            traits.loadRhs(&blB[8*RhsProgress], B_0);
             traits.madd(A0,B1,C1,B1);
             traits.loadRhs(&blB[9*RhsProgress], B1);
             traits.madd(A0,B2,C2,B2);
@@ -880,8 +880,8 @@ EIGEN_ASM_COMMENT("mybegin4");
             traits.loadLhs(&blA[2*LhsProgress], A0);
             traits.loadRhs(&blB[11*RhsProgress], B3);
 
-            traits.madd(A0,B0,C0,B0);
-            traits.loadRhs(&blB[12*RhsProgress], B0);
+            traits.madd(A0,B_0,C0,B_0);
+            traits.loadRhs(&blB[12*RhsProgress], B_0);
             traits.madd(A0,B1,C1,B1);
             traits.loadRhs(&blB[13*RhsProgress], B1);
             traits.madd(A0,B2,C2,B2);
@@ -890,7 +890,7 @@ EIGEN_ASM_COMMENT("mybegin4");
 
             traits.loadLhs(&blA[3*LhsProgress], A0);
             traits.loadRhs(&blB[15*RhsProgress], B3);
-            traits.madd(A0,B0,C0,B0);
+            traits.madd(A0,B_0,C0,B_0);
             traits.madd(A0,B1,C1,B1);
             traits.madd(A0,B2,C2,B2);
             traits.madd(A0,B3,C3,B3);
@@ -905,26 +905,26 @@ EIGEN_ASM_COMMENT("mybegin4");
           if(nr==2)
           {
             LhsPacket A0;
-            RhsPacket B0, B1;
+            RhsPacket B_0, B1;
 
             traits.loadLhs(&blA[0*LhsProgress], A0);
-            traits.loadRhs(&blB[0*RhsProgress], B0);
+            traits.loadRhs(&blB[0*RhsProgress], B_0);
             traits.loadRhs(&blB[1*RhsProgress], B1);
-            traits.madd(A0,B0,C0,B0);
+            traits.madd(A0,B_0,C0,B_0);
             traits.madd(A0,B1,C1,B1);
           }
           else
           {
             LhsPacket A0;
-            RhsPacket B0, B1, B2, B3;
+            RhsPacket B_0, B1, B2, B3;
 
             traits.loadLhs(&blA[0*LhsProgress], A0);
-            traits.loadRhs(&blB[0*RhsProgress], B0);
+            traits.loadRhs(&blB[0*RhsProgress], B_0);
             traits.loadRhs(&blB[1*RhsProgress], B1);
             traits.loadRhs(&blB[2*RhsProgress], B2);
             traits.loadRhs(&blB[3*RhsProgress], B3);
 
-            traits.madd(A0,B0,C0,B0);
+            traits.madd(A0,B_0,C0,B_0);
             traits.madd(A0,B1,C1,B1);
             traits.madd(A0,B2,C2,B2);
             traits.madd(A0,B3,C3,B3);
@@ -971,26 +971,26 @@ EIGEN_ASM_COMMENT("mybegin4");
           if(nr==2)
           {
             LhsScalar A0;
-            RhsScalar B0, B1;
+            RhsScalar B_0, B1;
 
             A0 = blA[k];
-            B0 = blB[0];
+            B_0 = blB[0];
             B1 = blB[1];
-            MADD(cj,A0,B0,C0,B0);
+            MADD(cj,A0,B_0,C0,B_0);
             MADD(cj,A0,B1,C1,B1);
           }
           else
           {
             LhsScalar A0;
-            RhsScalar B0, B1, B2, B3;
+            RhsScalar B_0, B1, B2, B3;
 
             A0 = blA[k];
-            B0 = blB[0];
+            B_0 = blB[0];
             B1 = blB[1];
             B2 = blB[2];
             B3 = blB[3];
 
-            MADD(cj,A0,B0,C0,B0);
+            MADD(cj,A0,B_0,C0,B_0);
             MADD(cj,A0,B1,C1,B1);
             MADD(cj,A0,B2,C2,B2);
             MADD(cj,A0,B3,C3,B3);
@@ -1009,12 +1009,7 @@ EIGEN_ASM_COMMENT("mybegin4");
     for(Index j2=packet_cols; j2<cols; j2++)
     {
       // unpack B
-      {
-        traits.unpackRhs(depth, &blockB[j2*strideB+offsetB], unpackedB);
-//         const RhsScalar* blB = &blockB[j2*strideB+offsetB];
-//         for(Index k=0; k<depth; k++)
-//           pstore(&unpackedB[k*RhsPacketSize], pset1<RhsPacket>(blB[k]));
-      }
+      traits.unpackRhs(depth, &blockB[j2*strideB+offsetB], unpackedB);
 
       for(Index i=0; i<peeled_mc; i+=mr)
       {
@@ -1032,14 +1027,14 @@ EIGEN_ASM_COMMENT("mybegin4");
         for(Index k=0; k<depth; k++)
         {
           LhsPacket A0, A1;
-          RhsPacket B0;
+          RhsPacket B_0;
           RhsPacket T0;
 
           traits.loadLhs(&blA[0*LhsProgress], A0);
           traits.loadLhs(&blA[1*LhsProgress], A1);
-          traits.loadRhs(&blB[0*RhsProgress], B0);
-          traits.madd(A0,B0,C0,T0);
-          traits.madd(A1,B0,C4,B0);
+          traits.loadRhs(&blB[0*RhsProgress], B_0);
+          traits.madd(A0,B_0,C0,T0);
+          traits.madd(A1,B_0,C4,B_0);
 
           blB += RhsProgress;
           blA += 2*LhsProgress;
@@ -1071,10 +1066,10 @@ EIGEN_ASM_COMMENT("mybegin4");
         for(Index k=0; k<depth; k++)
         {
           LhsPacket A0;
-          RhsPacket B0;
+          RhsPacket B_0;
           traits.loadLhs(blA, A0);
-          traits.loadRhs(blB, B0);
-          traits.madd(A0, B0, C0, B0);
+          traits.loadRhs(blB, B_0);
+          traits.madd(A0, B_0, C0, B_0);
           blB += RhsProgress;
           blA += LhsProgress;
         }
@@ -1096,8 +1091,8 @@ EIGEN_ASM_COMMENT("mybegin4");
         for(Index k=0; k<depth; k++)
         {
           LhsScalar A0 = blA[k];
-          RhsScalar B0 = blB[k];
-          MADD(cj, A0, B0, C0, B0);
+          RhsScalar B_0 = blB[k];
+          MADD(cj, A0, B_0, C0, B_0);
         }
         res[(j2+0)*resStride + i] += alpha*C0;
       }

Eigen/src/Core/products/GeneralMatrixMatrix.h

@@ -78,7 +78,7 @@ static void run(Index rows, Index cols, Index depth,
   typedef gebp_traits<LhsScalar,RhsScalar> Traits;
 
   Index kc = blocking.kc();                 // cache block size along the K direction
-  Index mc = std::min(rows,blocking.mc());  // cache block size along the M direction
+  Index mc = (std::min)(rows,blocking.mc());  // cache block size along the M direction
   //Index nc = blocking.nc(); // cache block size along the N direction
 
   gemm_pack_lhs<LhsScalar, Index, Traits::mr, Traits::LhsProgress, LhsStorageOrder> pack_lhs;
@@ -94,15 +94,16 @@ static void run(Index rows, Index cols, Index depth,
     
     std::size_t sizeA = kc*mc;
     std::size_t sizeW = kc*Traits::WorkSpaceFactor;
-    LhsScalar* blockA = ei_aligned_stack_new(LhsScalar, sizeA);
-    RhsScalar* w = ei_aligned_stack_new(RhsScalar, sizeW);
+    ei_declare_aligned_stack_constructed_variable(LhsScalar, blockA, sizeA, 0);
+    ei_declare_aligned_stack_constructed_variable(RhsScalar, w, sizeW, 0);
+    
     RhsScalar* blockB = blocking.blockB();
     eigen_internal_assert(blockB!=0);
 
     // For each horizontal panel of the rhs, and corresponding vertical panel of the lhs...
     for(Index k=0; k<depth; k+=kc)
     {
-      const Index actual_kc = std::min(k+kc,depth)-k; // => rows of B', and cols of the A'
+      const Index actual_kc = (std::min)(k+kc,depth)-k; // => rows of B', and cols of the A'
 
       // In order to reduce the chance that a thread has to wait for the other,
       // let's start by packing A'.
@@ -139,7 +140,7 @@ static void run(Index rows, Index cols, Index depth,
       // Then keep going as usual with the remaining A'
       for(Index i=mc; i<rows; i+=mc)
       {
-        const Index actual_mc = std::min(i+mc,rows)-i;
+        const Index actual_mc = (std::min)(i+mc,rows)-i;
 
         // pack A_i,k to A'
         pack_lhs(blockA, &lhs(i,k), lhsStride, actual_kc, actual_mc);
@@ -154,9 +155,6 @@ static void run(Index rows, Index cols, Index depth,
         #pragma omp atomic
         --(info[j].users);
     }
-
-    ei_aligned_stack_delete(LhsScalar, blockA, kc*mc);
-    ei_aligned_stack_delete(RhsScalar, w, sizeW);
   }
   else
 #endif // EIGEN_HAS_OPENMP
@@ -167,15 +165,16 @@ static void run(Index rows, Index cols, Index depth,
     std::size_t sizeA = kc*mc;
     std::size_t sizeB = kc*cols;
     std::size_t sizeW = kc*Traits::WorkSpaceFactor;
-    LhsScalar *blockA = blocking.blockA()==0 ? ei_aligned_stack_new(LhsScalar, sizeA) : blocking.blockA();
-    RhsScalar *blockB = blocking.blockB()==0 ? ei_aligned_stack_new(RhsScalar, sizeB) : blocking.blockB();
-    RhsScalar *blockW = blocking.blockW()==0 ? ei_aligned_stack_new(RhsScalar, sizeW) : blocking.blockW();
+
+    ei_declare_aligned_stack_constructed_variable(LhsScalar, blockA, sizeA, blocking.blockA());
+    ei_declare_aligned_stack_constructed_variable(RhsScalar, blockB, sizeB, blocking.blockB());
+    ei_declare_aligned_stack_constructed_variable(RhsScalar, blockW, sizeW, blocking.blockW());
 
     // For each horizontal panel of the rhs, and corresponding panel of the lhs...
     // (==GEMM_VAR1)
     for(Index k2=0; k2<depth; k2+=kc)
     {
-      const Index actual_kc = std::min(k2+kc,depth)-k2;
+      const Index actual_kc = (std::min)(k2+kc,depth)-k2;
 
       // OK, here we have selected one horizontal panel of rhs and one vertical panel of lhs.
       // => Pack rhs's panel into a sequential chunk of memory (L2 caching)
@@ -188,7 +187,7 @@ static void run(Index rows, Index cols, Index depth,
       // (==GEPP_VAR1)
       for(Index i2=0; i2<rows; i2+=mc)
       {
-        const Index actual_mc = std::min(i2+mc,rows)-i2;
+        const Index actual_mc = (std::min)(i2+mc,rows)-i2;
 
         // We pack the lhs's block into a sequential chunk of memory (L1 caching)
         // Note that this block will be read a very high number of times, which is equal to the number of
@@ -200,10 +199,6 @@ static void run(Index rows, Index cols, Index depth,
 
       }
     }
-
-    if(blocking.blockA()==0) ei_aligned_stack_delete(LhsScalar, blockA, sizeA);
-    if(blocking.blockB()==0) ei_aligned_stack_delete(RhsScalar, blockB, sizeB);
-    if(blocking.blockW()==0) ei_aligned_stack_delete(RhsScalar, blockW, sizeW);
   }
 }
 

Eigen/src/Core/products/GeneralMatrixMatrixTriangular.h

@@ -83,10 +83,10 @@ struct general_matrix_matrix_triangular_product<Index,LhsScalar,LhsStorageOrder,
     if(mc > Traits::nr)
       mc = (mc/Traits::nr)*Traits::nr;
 
-    LhsScalar* blockA = ei_aligned_stack_new(LhsScalar, kc*mc);
     std::size_t sizeW = kc*Traits::WorkSpaceFactor;
     std::size_t sizeB = sizeW + kc*size;
-    RhsScalar* allocatedBlockB = ei_aligned_stack_new(RhsScalar, sizeB);
+    ei_declare_aligned_stack_constructed_variable(LhsScalar, blockA, kc*mc, 0);
+    ei_declare_aligned_stack_constructed_variable(RhsScalar, allocatedBlockB, sizeB, 0);
     RhsScalar* blockB = allocatedBlockB + sizeW;
     
     gemm_pack_lhs<LhsScalar, Index, Traits::mr, Traits::LhsProgress, LhsStorageOrder> pack_lhs;
@@ -96,14 +96,14 @@ struct general_matrix_matrix_triangular_product<Index,LhsScalar,LhsStorageOrder,
 
     for(Index k2=0; k2<depth; k2+=kc)
     {
-      const Index actual_kc = std::min(k2+kc,depth)-k2;
+      const Index actual_kc = (std::min)(k2+kc,depth)-k2;
 
       // note that the actual rhs is the transpose/adjoint of mat
       pack_rhs(blockB, &rhs(k2,0), rhsStride, actual_kc, size);
 
       for(Index i2=0; i2<size; i2+=mc)
       {
-        const Index actual_mc = std::min(i2+mc,size)-i2;
+        const Index actual_mc = (std::min)(i2+mc,size)-i2;
 
         pack_lhs(blockA, &lhs(i2, k2), lhsStride, actual_kc, actual_mc);
 
@@ -112,7 +112,7 @@ struct general_matrix_matrix_triangular_product<Index,LhsScalar,LhsStorageOrder,
         //  2 - the actual_mc x actual_mc symmetric block => processed with a special kernel
         //  3 - after the diagonal => processed with gebp or skipped
         if (UpLo==Lower)
-          gebp(res+i2, resStride, blockA, blockB, actual_mc, actual_kc, std::min(size,i2), alpha,
+          gebp(res+i2, resStride, blockA, blockB, actual_mc, actual_kc, (std::min)(size,i2), alpha,
                -1, -1, 0, 0, allocatedBlockB);
 
         sybb(res+resStride*i2 + i2, resStride, blockA, blockB + actual_kc*i2, actual_mc, actual_kc, alpha, allocatedBlockB);
@@ -120,13 +120,11 @@ struct general_matrix_matrix_triangular_product<Index,LhsScalar,LhsStorageOrder,
         if (UpLo==Upper)
         {
           Index j2 = i2+actual_mc;
-          gebp(res+resStride*j2+i2, resStride, blockA, blockB+actual_kc*j2, actual_mc, actual_kc, std::max(Index(0), size-j2), alpha,
+          gebp(res+resStride*j2+i2, resStride, blockA, blockB+actual_kc*j2, actual_mc, actual_kc, (std::max)(Index(0), size-j2), alpha,
                -1, -1, 0, 0, allocatedBlockB);
         }
       }
     }
-    ei_aligned_stack_delete(LhsScalar, blockA, kc*mc);
-    ei_aligned_stack_delete(RhsScalar, allocatedBlockB, sizeB);
   }
 };
 

Eigen/src/Core/products/GeneralMatrixVector.h

@@ -134,7 +134,7 @@ EIGEN_DONT_INLINE static void run(
     }
     else
     {
-      skipColumns = std::min(skipColumns,cols);
+      skipColumns = (std::min)(skipColumns,cols);
       // note that the skiped columns are processed later.
     }
 
@@ -386,7 +386,7 @@ EIGEN_DONT_INLINE static void run(
     }
     else
     {
-      skipRows = std::min(skipRows,Index(rows));
+      skipRows = (std::min)(skipRows,Index(rows));
       // note that the skiped columns are processed later.
     }
     eigen_internal_assert(  alignmentPattern==NoneAligned

Eigen/src/Core/products/Parallelizer.h

@@ -30,7 +30,7 @@ namespace internal {
 /** \internal */
 inline void manage_multi_threading(Action action, int* v)
 {
-  static int m_maxThreads = -1;
+  static EIGEN_UNUSED int m_maxThreads = -1;
 
   if(action==SetAction)
   {

Eigen/src/Core/products/SelfadjointMatrixMatrix.h

@@ -114,7 +114,7 @@ struct symm_pack_rhs
     }
 
     // second part: diagonal block
-    for(Index j2=k2; j2<std::min(k2+rows,packet_cols); j2+=nr)
+    for(Index j2=k2; j2<(std::min)(k2+rows,packet_cols); j2+=nr)
     {
       // again we can split vertically in three different parts (transpose, symmetric, normal)
       // transpose
@@ -179,7 +179,7 @@ struct symm_pack_rhs
     for(Index j2=packet_cols; j2<cols; ++j2)
     {
       // transpose
-      Index half = std::min(end_k,j2);
+      Index half = (std::min)(end_k,j2);
       for(Index k=k2; k<half; k++)
       {
         blockB[count] = conj(rhs(j2,k));
@@ -261,12 +261,12 @@ struct product_selfadjoint_matrix<Scalar,Index,LhsStorageOrder,true,ConjugateLhs
     Index nc = cols;  // cache block size along the N direction
     computeProductBlockingSizes<Scalar,Scalar>(kc, mc, nc);
     // kc must smaller than mc
-    kc = std::min(kc,mc);
+    kc = (std::min)(kc,mc);
 
-    Scalar* blockA = ei_aligned_stack_new(Scalar, kc*mc);
     std::size_t sizeW = kc*Traits::WorkSpaceFactor;
     std::size_t sizeB = sizeW + kc*cols;
-    Scalar* allocatedBlockB = ei_aligned_stack_new(Scalar, sizeB);
+    ei_declare_aligned_stack_constructed_variable(Scalar, blockA, kc*mc, 0);
+    ei_declare_aligned_stack_constructed_variable(Scalar, allocatedBlockB, sizeB, 0);
     Scalar* blockB = allocatedBlockB + sizeW;
 
     gebp_kernel<Scalar, Scalar, Index, Traits::mr, Traits::nr, ConjugateLhs, ConjugateRhs> gebp_kernel;
@@ -276,7 +276,7 @@ struct product_selfadjoint_matrix<Scalar,Index,LhsStorageOrder,true,ConjugateLhs
 
     for(Index k2=0; k2<size; k2+=kc)
     {
-      const Index actual_kc = std::min(k2+kc,size)-k2;
+      const Index actual_kc = (std::min)(k2+kc,size)-k2;
 
       // we have selected one row panel of rhs and one column panel of lhs
       // pack rhs's panel into a sequential chunk of memory
@@ -289,7 +289,7 @@ struct product_selfadjoint_matrix<Scalar,Index,LhsStorageOrder,true,ConjugateLhs
       //  3 - the panel below the diagonal block => generic packed copy
       for(Index i2=0; i2<k2; i2+=mc)
       {
-        const Index actual_mc = std::min(i2+mc,k2)-i2;
+        const Index actual_mc = (std::min)(i2+mc,k2)-i2;
         // transposed packed copy
         pack_lhs_transposed(blockA, &lhs(k2, i2), lhsStride, actual_kc, actual_mc);
 
@@ -297,7 +297,7 @@ struct product_selfadjoint_matrix<Scalar,Index,LhsStorageOrder,true,ConjugateLhs
       }
       // the block diagonal
       {
-        const Index actual_mc = std::min(k2+kc,size)-k2;
+        const Index actual_mc = (std::min)(k2+kc,size)-k2;
         // symmetric packed copy
         pack_lhs(blockA, &lhs(k2,k2), lhsStride, actual_kc, actual_mc);
 
@@ -306,16 +306,13 @@ struct product_selfadjoint_matrix<Scalar,Index,LhsStorageOrder,true,ConjugateLhs
 
       for(Index i2=k2+kc; i2<size; i2+=mc)
       {
-        const Index actual_mc = std::min(i2+mc,size)-i2;
+        const Index actual_mc = (std::min)(i2+mc,size)-i2;
         gemm_pack_lhs<Scalar, Index, Traits::mr, Traits::LhsProgress, LhsStorageOrder,false>()
           (blockA, &lhs(i2, k2), lhsStride, actual_kc, actual_mc);
 
         gebp_kernel(res+i2, resStride, blockA, blockB, actual_mc, actual_kc, cols, alpha);
       }
     }
-
-    ei_aligned_stack_delete(Scalar, blockA, kc*mc);
-    ei_aligned_stack_delete(Scalar, allocatedBlockB, sizeB);
   }
 };
 
@@ -343,11 +340,10 @@ struct product_selfadjoint_matrix<Scalar,Index,LhsStorageOrder,false,ConjugateLh
     Index mc = rows;  // cache block size along the M direction
     Index nc = cols;  // cache block size along the N direction
     computeProductBlockingSizes<Scalar,Scalar>(kc, mc, nc);
-
-    Scalar* blockA = ei_aligned_stack_new(Scalar, kc*mc);
     std::size_t sizeW = kc*Traits::WorkSpaceFactor;
     std::size_t sizeB = sizeW + kc*cols;
-    Scalar* allocatedBlockB = ei_aligned_stack_new(Scalar, sizeB);
+    ei_declare_aligned_stack_constructed_variable(Scalar, blockA, kc*mc, 0);
+    ei_declare_aligned_stack_constructed_variable(Scalar, allocatedBlockB, sizeB, 0);
     Scalar* blockB = allocatedBlockB + sizeW;
 
     gebp_kernel<Scalar, Scalar, Index, Traits::mr, Traits::nr, ConjugateLhs, ConjugateRhs> gebp_kernel;
@@ -356,22 +352,19 @@ struct product_selfadjoint_matrix<Scalar,Index,LhsStorageOrder,false,ConjugateLh
 
     for(Index k2=0; k2<size; k2+=kc)
     {
-      const Index actual_kc = std::min(k2+kc,size)-k2;
+      const Index actual_kc = (std::min)(k2+kc,size)-k2;
 
       pack_rhs(blockB, _rhs, rhsStride, actual_kc, cols, k2);
 
       // => GEPP
       for(Index i2=0; i2<rows; i2+=mc)
       {
-        const Index actual_mc = std::min(i2+mc,rows)-i2;
+        const Index actual_mc = (std::min)(i2+mc,rows)-i2;
         pack_lhs(blockA, &lhs(i2, k2), lhsStride, actual_kc, actual_mc);
 
         gebp_kernel(res+i2, resStride, blockA, blockB, actual_mc, actual_kc, cols, alpha);
       }
     }
-
-    ei_aligned_stack_delete(Scalar, blockA, kc*mc);
-    ei_aligned_stack_delete(Scalar, allocatedBlockB, sizeB);
   }
 };
 

Eigen/src/Core/products/SelfadjointMatrixVector.h

@@ -62,17 +62,15 @@ static EIGEN_DONT_INLINE void product_selfadjoint_vector(
   // FIXME this copy is now handled outside product_selfadjoint_vector, so it could probably be removed.
   // if the rhs is not sequentially stored in memory we copy it to a temporary buffer,
   // this is because we need to extract packets
-  const Scalar* EIGEN_RESTRICT rhs = _rhs;
+  ei_declare_aligned_stack_constructed_variable(Scalar,rhs,size,rhsIncr==1 ? const_cast<Scalar*>(_rhs) : 0);  
   if (rhsIncr!=1)
   {
-    Scalar* r = ei_aligned_stack_new(Scalar, size);
     const Scalar* it = _rhs;
     for (Index i=0; i<size; ++i, it+=rhsIncr)
-      r[i] = *it;
-    rhs = r;
+      rhs[i] = *it;
   }
 
-  Index bound = std::max(Index(0),size-8) & 0xfffffffe;
+  Index bound = (std::max)(Index(0),size-8) & 0xfffffffe;
   if (FirstTriangular)
     bound = size - bound;
 
@@ -160,9 +158,6 @@ static EIGEN_DONT_INLINE void product_selfadjoint_vector(
     }
     res[j] += alpha * t2;
   }
-
-  if(rhsIncr!=1)
-    ei_aligned_stack_delete(Scalar, const_cast<Scalar*>(rhs), size);
 }
 
 } // end namespace internal 
@@ -211,40 +206,28 @@ struct SelfadjointProductMatrix<Lhs,LhsMode,false,Rhs,0,true>
     
     internal::gemv_static_vector_if<ResScalar,Dest::SizeAtCompileTime,Dest::MaxSizeAtCompileTime,!EvalToDest> static_dest;
     internal::gemv_static_vector_if<RhsScalar,_ActualRhsType::SizeAtCompileTime,_ActualRhsType::MaxSizeAtCompileTime,!UseRhs> static_rhs;
+
+    ei_declare_aligned_stack_constructed_variable(ResScalar,actualDestPtr,dest.size(),
+                                                  EvalToDest ? dest.data() : static_dest.data());
+                                                  
+    ei_declare_aligned_stack_constructed_variable(RhsScalar,actualRhsPtr,rhs.size(),
+        UseRhs ? const_cast<RhsScalar*>(rhs.data()) : static_rhs.data());
     
-    bool freeDestPtr = false;
-    ResScalar* actualDestPtr;
-    if(EvalToDest)
-      actualDestPtr = dest.data();
-    else
+    if(!EvalToDest)
     {
       #ifdef EIGEN_DENSE_STORAGE_CTOR_PLUGIN
       int size = dest.size();
       EIGEN_DENSE_STORAGE_CTOR_PLUGIN
       #endif
-      if((actualDestPtr=static_dest.data())==0)
-      {
-        freeDestPtr = true;
-        actualDestPtr = ei_aligned_stack_new(ResScalar,dest.size());
-      }
       MappedDest(actualDestPtr, dest.size()) = dest;
     }
       
-    bool freeRhsPtr = false;
-    RhsScalar* actualRhsPtr;
-    if(UseRhs)
-      actualRhsPtr = const_cast<RhsScalar*>(rhs.data());
-    else
+    if(!UseRhs)
     {
       #ifdef EIGEN_DENSE_STORAGE_CTOR_PLUGIN
       int size = rhs.size();
       EIGEN_DENSE_STORAGE_CTOR_PLUGIN
       #endif
-      if((actualRhsPtr=static_rhs.data())==0)
-      {
-        freeRhsPtr = true;
-        actualRhsPtr = ei_aligned_stack_new(RhsScalar,rhs.size());
-      }
       Map<typename _ActualRhsType::PlainObject>(actualRhsPtr, rhs.size()) = rhs;
     }
       
@@ -259,11 +242,7 @@ struct SelfadjointProductMatrix<Lhs,LhsMode,false,Rhs,0,true>
       );
     
     if(!EvalToDest)
-    {
       dest = MappedDest(actualDestPtr, dest.size());
-      if(freeDestPtr) ei_aligned_stack_delete(ResScalar, actualDestPtr, dest.size());
-    }
-    if(freeRhsPtr) ei_aligned_stack_delete(RhsScalar, actualRhsPtr, rhs.size());
   }
 };
 

Eigen/src/Core/products/SelfadjointProduct.h

@@ -81,27 +81,17 @@ struct selfadjoint_product_selector<MatrixType,OtherType,UpLo,true>
       UseOtherDirectly = _ActualOtherType::InnerStrideAtCompileTime==1
     };
     internal::gemv_static_vector_if<Scalar,OtherType::SizeAtCompileTime,OtherType::MaxSizeAtCompileTime,!UseOtherDirectly> static_other;
-    
-    bool freeOtherPtr = false;
-    Scalar* actualOtherPtr;
-    if(UseOtherDirectly)
-      actualOtherPtr = const_cast<Scalar*>(actualOther.data());
-    else
-    {
-      if((actualOtherPtr=static_other.data())==0)
-      {
-        freeOtherPtr = true;
-        actualOtherPtr = ei_aligned_stack_new(Scalar,other.size());
-      }
+
+    ei_declare_aligned_stack_constructed_variable(Scalar, actualOtherPtr, other.size(),
+      (UseOtherDirectly ? const_cast<Scalar*>(actualOther.data()) : static_other.data()));
+      
+    if(!UseOtherDirectly)
       Map<typename _ActualOtherType::PlainObject>(actualOtherPtr, actualOther.size()) = actualOther;
-    }
     
     selfadjoint_rank1_update<Scalar,Index,StorageOrder,UpLo,
                               OtherBlasTraits::NeedToConjugate  && NumTraits<Scalar>::IsComplex,
                             (!OtherBlasTraits::NeedToConjugate) && NumTraits<Scalar>::IsComplex>
           ::run(other.size(), mat.data(), mat.outerStride(), actualOtherPtr, actualAlpha);
-    
-    if((!UseOtherDirectly) && freeOtherPtr) ei_aligned_stack_delete(Scalar, actualOtherPtr, other.size());
   }
 };
 

Eigen/src/Core/products/TriangularMatrixMatrix.h

@@ -96,33 +96,38 @@ struct product_triangular_matrix_matrix<Scalar,Index,Mode,true,
                                            LhsStorageOrder,ConjugateLhs,
                                            RhsStorageOrder,ConjugateRhs,ColMajor>
 {
+  
+  typedef gebp_traits<Scalar,Scalar> Traits;
+  enum {
+    SmallPanelWidth   = 2 * EIGEN_PLAIN_ENUM_MAX(Traits::mr,Traits::nr),
+    IsLower = (Mode&Lower) == Lower,
+    SetDiag = (Mode&(ZeroDiag|UnitDiag)) ? 0 : 1
+  };
 
   static EIGEN_DONT_INLINE void run(
-    Index rows, Index cols, Index depth,
+    Index _rows, Index _cols, Index _depth,
     const Scalar* _lhs, Index lhsStride,
     const Scalar* _rhs, Index rhsStride,
     Scalar* res,        Index resStride,
     Scalar alpha)
   {
+    // strip zeros
+    Index diagSize  = (std::min)(_rows,_depth);
+    Index rows      = IsLower ? _rows : diagSize;
+    Index depth     = IsLower ? diagSize : _depth;
+    Index cols      = _cols;
+    
     const_blas_data_mapper<Scalar, Index, LhsStorageOrder> lhs(_lhs,lhsStride);
     const_blas_data_mapper<Scalar, Index, RhsStorageOrder> rhs(_rhs,rhsStride);
 
-    typedef gebp_traits<Scalar,Scalar> Traits;
-    enum {
-      SmallPanelWidth   = EIGEN_PLAIN_ENUM_MAX(Traits::mr,Traits::nr),
-      IsLower = (Mode&Lower) == Lower,
-      SetDiag = (Mode&(ZeroDiag|UnitDiag)) ? 0 : 1
-    };
-
     Index kc = depth; // cache block size along the K direction
     Index mc = rows;  // cache block size along the M direction
     Index nc = cols;  // cache block size along the N direction
     computeProductBlockingSizes<Scalar,Scalar,4>(kc, mc, nc);
-
-    Scalar* blockA = ei_aligned_stack_new(Scalar, kc*mc);
     std::size_t sizeW = kc*Traits::WorkSpaceFactor;
     std::size_t sizeB = sizeW + kc*cols;
-    Scalar* allocatedBlockB = ei_aligned_stack_new(Scalar, sizeB);
+    ei_declare_aligned_stack_constructed_variable(Scalar, blockA, kc*mc, 0);
+    ei_declare_aligned_stack_constructed_variable(Scalar, allocatedBlockB, sizeB, 0);    
     Scalar* blockB = allocatedBlockB + sizeW;
 
     Matrix<Scalar,SmallPanelWidth,SmallPanelWidth,LhsStorageOrder> triangularBuffer;
@@ -140,7 +145,7 @@ struct product_triangular_matrix_matrix<Scalar,Index,Mode,true,
         IsLower ? k2>0 : k2<depth;
         IsLower ? k2-=kc : k2+=kc)
     {
-      Index actual_kc = std::min(IsLower ? k2 : depth-k2, kc);
+      Index actual_kc = (std::min)(IsLower ? k2 : depth-k2, kc);
       Index actual_k2 = IsLower ? k2-actual_kc : k2;
 
       // align blocks with the end of the triangular part for trapezoidal lhs
@@ -153,10 +158,11 @@ struct product_triangular_matrix_matrix<Scalar,Index,Mode,true,
       pack_rhs(blockB, &rhs(actual_k2,0), rhsStride, actual_kc, cols);
 
       // the selected lhs's panel has to be split in three different parts:
-      //  1 - the part which is above the diagonal block => skip it
+      //  1 - the part which is zero => skip it
       //  2 - the diagonal block => special kernel
-      //  3 - the panel below the diagonal block => GEPP
-      // the block diagonal, if any
+      //  3 - the dense panel below (lower case) or above (upper case) the diagonal block => GEPP
+
+      // the block diagonal, if any:
       if(IsLower || actual_k2<rows)
       {
         // for each small vertical panels of lhs
@@ -194,13 +200,13 @@ struct product_triangular_matrix_matrix<Scalar,Index,Mode,true,
           }
         }
       }
-      // the part below the diagonal => GEPP
+      // the part below (lower case) or above (upper case) the diagonal => GEPP
       {
         Index start = IsLower ? k2 : 0;
-        Index end   = IsLower ? rows : std::min(actual_k2,rows);
+        Index end   = IsLower ? rows : (std::min)(actual_k2,rows);
         for(Index i2=start; i2<end; i2+=mc)
         {
-          const Index actual_mc = std::min(i2+mc,end)-i2;
+          const Index actual_mc = (std::min)(i2+mc,end)-i2;
           gemm_pack_lhs<Scalar, Index, Traits::mr,Traits::LhsProgress, LhsStorageOrder,false>()
             (blockA, &lhs(i2, actual_k2), lhsStride, actual_kc, actual_mc);
 
@@ -208,10 +214,6 @@ struct product_triangular_matrix_matrix<Scalar,Index,Mode,true,
         }
       }
     }
-
-    ei_aligned_stack_delete(Scalar, blockA, kc*mc);
-    ei_aligned_stack_delete(Scalar, allocatedBlockB, sizeB);
-//     delete[] allocatedBlockB;
   }
 };
 
@@ -223,33 +225,38 @@ struct product_triangular_matrix_matrix<Scalar,Index,Mode,false,
                                            LhsStorageOrder,ConjugateLhs,
                                            RhsStorageOrder,ConjugateRhs,ColMajor>
 {
+  typedef gebp_traits<Scalar,Scalar> Traits;
+  enum {
+    SmallPanelWidth   = EIGEN_PLAIN_ENUM_MAX(Traits::mr,Traits::nr),
+    IsLower = (Mode&Lower) == Lower,
+    SetDiag = (Mode&(ZeroDiag|UnitDiag)) ? 0 : 1
+  };
 
   static EIGEN_DONT_INLINE void run(
-    Index rows, Index cols, Index depth,
+    Index _rows, Index _cols, Index _depth,
     const Scalar* _lhs, Index lhsStride,
     const Scalar* _rhs, Index rhsStride,
     Scalar* res,        Index resStride,
     Scalar alpha)
   {
+    // strip zeros
+    Index diagSize  = (std::min)(_cols,_depth);
+    Index rows      = _rows;
+    Index depth     = IsLower ? _depth : diagSize;
+    Index cols      = IsLower ? diagSize : _cols;
+    
     const_blas_data_mapper<Scalar, Index, LhsStorageOrder> lhs(_lhs,lhsStride);
     const_blas_data_mapper<Scalar, Index, RhsStorageOrder> rhs(_rhs,rhsStride);
 
-    typedef gebp_traits<Scalar,Scalar> Traits;
-    enum {
-      SmallPanelWidth   = EIGEN_PLAIN_ENUM_MAX(Traits::mr,Traits::nr),
-      IsLower = (Mode&Lower) == Lower,
-      SetDiag = (Mode&(ZeroDiag|UnitDiag)) ? 0 : 1
-    };
-
     Index kc = depth; // cache block size along the K direction
     Index mc = rows;  // cache block size along the M direction
     Index nc = cols;  // cache block size along the N direction
     computeProductBlockingSizes<Scalar,Scalar,4>(kc, mc, nc);
 
-    Scalar* blockA = ei_aligned_stack_new(Scalar, kc*mc);
     std::size_t sizeW = kc*Traits::WorkSpaceFactor;
     std::size_t sizeB = sizeW + kc*cols;
-    Scalar* allocatedBlockB = ei_aligned_stack_new(Scalar,sizeB);
+    ei_declare_aligned_stack_constructed_variable(Scalar, blockA, kc*mc, 0);
+    ei_declare_aligned_stack_constructed_variable(Scalar, allocatedBlockB, sizeB, 0);
     Scalar* blockB = allocatedBlockB + sizeW;
 
     Matrix<Scalar,SmallPanelWidth,SmallPanelWidth,RhsStorageOrder> triangularBuffer;
@@ -268,7 +275,7 @@ struct product_triangular_matrix_matrix<Scalar,Index,Mode,false,
         IsLower ? k2<depth  : k2>0;
         IsLower ? k2+=kc   : k2-=kc)
     {
-      Index actual_kc = std::min(IsLower ? depth-k2 : k2, kc);
+      Index actual_kc = (std::min)(IsLower ? depth-k2 : k2, kc);
       Index actual_k2 = IsLower ? k2 : k2-actual_kc;
 
       // align blocks with the end of the triangular part for trapezoidal rhs
@@ -279,7 +286,7 @@ struct product_triangular_matrix_matrix<Scalar,Index,Mode,false,
       }
 
       // remaining size
-      Index rs = IsLower ? std::min(cols,actual_k2) : cols - k2;
+      Index rs = IsLower ? (std::min)(cols,actual_k2) : cols - k2;
       // size of the triangular part
       Index ts = (IsLower && actual_k2>=cols) ? 0 : actual_kc;
 
@@ -320,7 +327,7 @@ struct product_triangular_matrix_matrix<Scalar,Index,Mode,false,
 
       for (Index i2=0; i2<rows; i2+=mc)
       {
-        const Index actual_mc = std::min(mc,rows-i2);
+        const Index actual_mc = (std::min)(mc,rows-i2);
         pack_lhs(blockA, &lhs(i2, actual_k2), lhsStride, actual_kc, actual_mc);
 
         // triangular kernel
@@ -347,9 +354,6 @@ struct product_triangular_matrix_matrix<Scalar,Index,Mode,false,
                     -1, -1, 0, 0, allocatedBlockB);
       }
     }
-
-    ei_aligned_stack_delete(Scalar, blockA, kc*mc);
-    ei_aligned_stack_delete(Scalar, allocatedBlockB, sizeB);
   }
 };
 

Eigen/src/Core/products/TriangularMatrixVector.h

@@ -36,15 +36,16 @@ struct product_triangular_matrix_vector<Index,Mode,LhsScalar,ConjLhs,RhsScalar,C
   typedef typename scalar_product_traits<LhsScalar, RhsScalar>::ReturnType ResScalar;
   enum {
     IsLower = ((Mode&Lower)==Lower),
-    HasUnitDiag = (Mode & UnitDiag)==UnitDiag
+    HasUnitDiag = (Mode & UnitDiag)==UnitDiag,
+    HasZeroDiag = (Mode & ZeroDiag)==ZeroDiag
   };
-  static EIGEN_DONT_INLINE  void run(Index rows, Index cols, const LhsScalar* _lhs, Index lhsStride,
+  static EIGEN_DONT_INLINE  void run(Index _rows, Index _cols, const LhsScalar* _lhs, Index lhsStride,
                                      const RhsScalar* _rhs, Index rhsIncr, ResScalar* _res, Index resIncr, ResScalar alpha)
   {
-    EIGEN_UNUSED_VARIABLE(resIncr);
-    eigen_assert(resIncr==1);
-    
     static const Index PanelWidth = EIGEN_TUNE_TRIANGULAR_PANEL_WIDTH;
+    Index size = (std::min)(_rows,_cols);
+    Index rows = IsLower ? _rows : (std::min)(_rows,_cols);
+    Index cols = IsLower ? (std::min)(_rows,_cols) : _cols;
 
     typedef Map<const Matrix<LhsScalar,Dynamic,Dynamic,ColMajor>, 0, OuterStride<> > LhsMap;
     const LhsMap lhs(_lhs,rows,cols,OuterStride<>(lhsStride));
@@ -57,20 +58,20 @@ struct product_triangular_matrix_vector<Index,Mode,LhsScalar,ConjLhs,RhsScalar,C
     typedef Map<Matrix<ResScalar,Dynamic,1> > ResMap;
     ResMap res(_res,rows);
 
-    for (Index pi=0; pi<cols; pi+=PanelWidth)
+    for (Index pi=0; pi<size; pi+=PanelWidth)
     {
-      Index actualPanelWidth = std::min(PanelWidth, cols-pi);
+      Index actualPanelWidth = (std::min)(PanelWidth, size-pi);
       for (Index k=0; k<actualPanelWidth; ++k)
       {
         Index i = pi + k;
-        Index s = IsLower ? (HasUnitDiag ? i+1 : i ) : pi;
+        Index s = IsLower ? ((HasUnitDiag||HasZeroDiag) ? i+1 : i ) : pi;
         Index r = IsLower ? actualPanelWidth-k : k+1;
-        if ((!HasUnitDiag) || (--r)>0)
+        if ((!(HasUnitDiag||HasZeroDiag)) || (--r)>0)
           res.segment(s,r) += (alpha * cjRhs.coeff(i)) * cjLhs.col(i).segment(s,r);
         if (HasUnitDiag)
           res.coeffRef(i) += alpha * cjRhs.coeff(i);
       }
-      Index r = IsLower ? cols - pi - actualPanelWidth : pi;
+      Index r = IsLower ? rows - pi - actualPanelWidth : pi;
       if (r>0)
       {
         Index s = IsLower ? pi+actualPanelWidth : 0;
@@ -81,6 +82,14 @@ struct product_triangular_matrix_vector<Index,Mode,LhsScalar,ConjLhs,RhsScalar,C
             &res.coeffRef(s), resIncr, alpha);
       }
     }
+    if((!IsLower) && cols>size)
+    {
+      general_matrix_vector_product<Index,LhsScalar,ColMajor,ConjLhs,RhsScalar,ConjRhs>::run(
+          rows, cols-size,
+          &lhs.coeffRef(0,size), lhsStride,
+          &rhs.coeffRef(size), rhsIncr,
+          _res, resIncr, alpha);
+    }
   }
 };
 
@@ -90,15 +99,16 @@ struct product_triangular_matrix_vector<Index,Mode,LhsScalar,ConjLhs,RhsScalar,C
   typedef typename scalar_product_traits<LhsScalar, RhsScalar>::ReturnType ResScalar;
   enum {
     IsLower = ((Mode&Lower)==Lower),
-    HasUnitDiag = (Mode & UnitDiag)==UnitDiag
+    HasUnitDiag = (Mode & UnitDiag)==UnitDiag,
+    HasZeroDiag = (Mode & ZeroDiag)==ZeroDiag
   };
-  static void run(Index rows, Index cols, const LhsScalar* _lhs, Index lhsStride,
+  static void run(Index _rows, Index _cols, const LhsScalar* _lhs, Index lhsStride,
                   const RhsScalar* _rhs, Index rhsIncr, ResScalar* _res, Index resIncr, ResScalar alpha)
   {
-    eigen_assert(rhsIncr==1);
-    EIGEN_UNUSED_VARIABLE(rhsIncr);
-    
     static const Index PanelWidth = EIGEN_TUNE_TRIANGULAR_PANEL_WIDTH;
+    Index diagSize = (std::min)(_rows,_cols);
+    Index rows = IsLower ? _rows : diagSize;
+    Index cols = IsLower ? diagSize : _cols;
 
     typedef Map<const Matrix<LhsScalar,Dynamic,Dynamic,RowMajor>, 0, OuterStride<> > LhsMap;
     const LhsMap lhs(_lhs,rows,cols,OuterStride<>(lhsStride));
@@ -111,15 +121,15 @@ struct product_triangular_matrix_vector<Index,Mode,LhsScalar,ConjLhs,RhsScalar,C
     typedef Map<Matrix<ResScalar,Dynamic,1>, 0, InnerStride<> > ResMap;
     ResMap res(_res,rows,InnerStride<>(resIncr));
     
-    for (Index pi=0; pi<cols; pi+=PanelWidth)
+    for (Index pi=0; pi<diagSize; pi+=PanelWidth)
     {
-      Index actualPanelWidth = std::min(PanelWidth, cols-pi);
+      Index actualPanelWidth = (std::min)(PanelWidth, diagSize-pi);
       for (Index k=0; k<actualPanelWidth; ++k)
       {
         Index i = pi + k;
-        Index s = IsLower ? pi  : (HasUnitDiag ? i+1 : i);
+        Index s = IsLower ? pi  : ((HasUnitDiag||HasZeroDiag) ? i+1 : i);
         Index r = IsLower ? k+1 : actualPanelWidth-k;
-        if ((!HasUnitDiag) || (--r)>0)
+        if ((!(HasUnitDiag||HasZeroDiag)) || (--r)>0)
           res.coeffRef(i) += alpha * (cjLhs.row(i).segment(s,r).cwiseProduct(cjRhs.segment(s,r).transpose())).sum();
         if (HasUnitDiag)
           res.coeffRef(i) += alpha * cjRhs.coeff(i);
@@ -135,6 +145,14 @@ struct product_triangular_matrix_vector<Index,Mode,LhsScalar,ConjLhs,RhsScalar,C
             &res.coeffRef(pi), resIncr, alpha);
       }
     }
+    if(IsLower && rows>diagSize)
+    {
+      general_matrix_vector_product<Index,LhsScalar,RowMajor,ConjLhs,RhsScalar,ConjRhs>::run(
+            rows-diagSize, cols,
+            &lhs.coeffRef(diagSize,0), lhsStride,
+            &rhs.coeffRef(0), rhsIncr,
+            &res.coeffRef(diagSize), resIncr, alpha);
+    }
   }
 };
 
@@ -185,8 +203,8 @@ struct TriangularProduct<Mode,false,Lhs,true,Rhs,false>
   template<typename Dest> void scaleAndAddTo(Dest& dst, Scalar alpha) const
   {
     eigen_assert(dst.rows()==m_lhs.rows() && dst.cols()==m_rhs.cols());
-    
-    typedef TriangularProduct<(Mode & UnitDiag) | ((Mode & Lower) ? Upper : Lower),true,Transpose<const Rhs>,false,Transpose<const Lhs>,true> TriangularProductTranspose;
+
+    typedef TriangularProduct<(Mode & (UnitDiag|ZeroDiag)) | ((Mode & Lower) ? Upper : Lower),true,Transpose<const Rhs>,false,Transpose<const Lhs>,true> TriangularProductTranspose;
     Transpose<Dest> dstT(dst);
     internal::trmv_selector<(int(internal::traits<Rhs>::Flags)&RowMajorBit) ? ColMajor : RowMajor>::run(
       TriangularProductTranspose(m_rhs.transpose(),m_lhs.transpose()), dstT, alpha);
@@ -235,23 +253,15 @@ template<> struct trmv_selector<ColMajor>
     
     RhsScalar compatibleAlpha = get_factor<ResScalar,RhsScalar>::run(actualAlpha);
 
-    ResScalar* actualDestPtr;
-    bool freeDestPtr = false;
-    if (evalToDest)
-    {
-      actualDestPtr = dest.data();
-    }
-    else
+    ei_declare_aligned_stack_constructed_variable(ResScalar,actualDestPtr,dest.size(),
+                                                  evalToDest ? dest.data() : static_dest.data());
+
+    if(!evalToDest)
     {
       #ifdef EIGEN_DENSE_STORAGE_CTOR_PLUGIN
       int size = dest.size();
       EIGEN_DENSE_STORAGE_CTOR_PLUGIN
       #endif
-      if((actualDestPtr = static_dest.data())==0)
-      {
-        freeDestPtr = true;
-        actualDestPtr = ei_aligned_stack_new(ResScalar,dest.size());
-      }
       if(!alphaIsCompatible)
       {
         MappedDest(actualDestPtr, dest.size()).setZero();
@@ -277,7 +287,6 @@ template<> struct trmv_selector<ColMajor>
         dest += actualAlpha * MappedDest(actualDestPtr, dest.size());
       else
         dest = MappedDest(actualDestPtr, dest.size());
-      if(freeDestPtr) ei_aligned_stack_delete(ResScalar, actualDestPtr, dest.size());
     }
   }
 };
@@ -310,23 +319,15 @@ template<> struct trmv_selector<RowMajor>
 
     gemv_static_vector_if<RhsScalar,_ActualRhsType::SizeAtCompileTime,_ActualRhsType::MaxSizeAtCompileTime,!DirectlyUseRhs> static_rhs;
 
-    RhsScalar* actualRhsPtr;
-    bool freeRhsPtr = false;
-    if (DirectlyUseRhs)
-    {
-      actualRhsPtr = const_cast<RhsScalar*>(actualRhs.data());
-    }
-    else
+    ei_declare_aligned_stack_constructed_variable(RhsScalar,actualRhsPtr,actualRhs.size(),
+        DirectlyUseRhs ? const_cast<RhsScalar*>(actualRhs.data()) : static_rhs.data());
+
+    if(!DirectlyUseRhs)
     {
       #ifdef EIGEN_DENSE_STORAGE_CTOR_PLUGIN
       int size = actualRhs.size();
       EIGEN_DENSE_STORAGE_CTOR_PLUGIN
       #endif
-      if((actualRhsPtr = static_rhs.data())==0)
-      {
-        freeRhsPtr = true;
-        actualRhsPtr = ei_aligned_stack_new(RhsScalar, actualRhs.size());
-      }
       Map<typename _ActualRhsType::PlainObject>(actualRhsPtr, actualRhs.size()) = actualRhs;
     }
     
@@ -340,8 +341,6 @@ template<> struct trmv_selector<RowMajor>
             actualRhsPtr,1,
             dest.data(),dest.innerStride(),
             actualAlpha);
-
-    if((!DirectlyUseRhs) && freeRhsPtr) ei_aligned_stack_delete(RhsScalar, actualRhsPtr, prod.rhs().size());
   }
 };
 

Eigen/src/Core/products/TriangularSolverMatrix.h

@@ -70,38 +70,48 @@ struct triangular_solve_matrix<Scalar,Index,OnTheLeft,Mode,Conjugate,TriStorageO
     Index nc = cols;  // cache block size along the N direction
     computeProductBlockingSizes<Scalar,Scalar,4>(kc, mc, nc);
 
-    Scalar* blockA = ei_aligned_stack_new(Scalar, kc*mc);
     std::size_t sizeW = kc*Traits::WorkSpaceFactor;
     std::size_t sizeB = sizeW + kc*cols;
-    Scalar* allocatedBlockB = ei_aligned_stack_new(Scalar, sizeB);
+    ei_declare_aligned_stack_constructed_variable(Scalar, blockA, kc*mc, 0);
+    ei_declare_aligned_stack_constructed_variable(Scalar, allocatedBlockB, sizeB, 0);
     Scalar* blockB = allocatedBlockB + sizeW;
+    Scalar* blockW = allocatedBlockB;
 
     conj_if<Conjugate> conj;
     gebp_kernel<Scalar, Scalar, Index, Traits::mr, Traits::nr, Conjugate, false> gebp_kernel;
     gemm_pack_lhs<Scalar, Index, Traits::mr, Traits::LhsProgress, TriStorageOrder> pack_lhs;
     gemm_pack_rhs<Scalar, Index, Traits::nr, ColMajor, false, true> pack_rhs;
 
+    // the goal here is to subdivise the Rhs panels such that we keep some cache
+    // coherence when accessing the rhs elements
+    std::ptrdiff_t l1, l2;
+    manage_caching_sizes(GetAction, &l1, &l2);
+    Index subcols = cols>0 ? l2/(4 * sizeof(Scalar) * otherStride) : 0;
+    subcols = std::max<Index>((subcols/Traits::nr)*Traits::nr, Traits::nr);
+
     for(Index k2=IsLower ? 0 : size;
         IsLower ? k2<size : k2>0;
         IsLower ? k2+=kc : k2-=kc)
     {
-      const Index actual_kc = std::min(IsLower ? size-k2 : k2, kc);
+      const Index actual_kc = (std::min)(IsLower ? size-k2 : k2, kc);
 
       // We have selected and packed a big horizontal panel R1 of rhs. Let B be the packed copy of this panel,
       // and R2 the remaining part of rhs. The corresponding vertical panel of lhs is split into
       // A11 (the triangular part) and A21 the remaining rectangular part.
       // Then the high level algorithm is:
       //  - B = R1                    => general block copy (done during the next step)
-      //  - R1 = L1^-1 B              => tricky part
+      //  - R1 = A11^-1 B             => tricky part
       //  - update B from the new R1  => actually this has to be performed continuously during the above step
-      //  - R2 = L2 * B               => GEPP
+      //  - R2 -= A21 * B             => GEPP
 
-      // The tricky part: compute R1 = L1^-1 B while updating B from R1
-      // The idea is to split L1 into multiple small vertical panels.
-      // Each panel can be split into a small triangular part A1 which is processed without optimization,
-      // and the remaining small part A2 which is processed using gebp with appropriate block strides
+      // The tricky part: compute R1 = A11^-1 B while updating B from R1
+      // The idea is to split A11 into multiple small vertical panels.
+      // Each panel can be split into a small triangular part T1k which is processed without optimization,
+      // and the remaining small part T2k which is processed using gebp with appropriate block strides
+      for(Index j2=0; j2<cols; j2+=subcols)
       {
-        // for each small vertical panels of lhs
+        Index actual_cols = (std::min)(cols-j2,subcols);
+        // for each small vertical panels [T1k^T, T2k^T]^T of lhs
         for (Index k1=0; k1<actual_kc; k1+=SmallPanelWidth)
         {
           Index actualPanelWidth = std::min<Index>(actual_kc-k1, SmallPanelWidth);
@@ -114,7 +124,7 @@ struct triangular_solve_matrix<Scalar,Index,OnTheLeft,Mode,Conjugate,TriStorageO
             Index rs = actualPanelWidth - k - 1; // remaining size
 
             Scalar a = (Mode & UnitDiag) ? Scalar(1) : Scalar(1)/conj(tri(i,i));
-            for (Index j=0; j<cols; ++j)
+            for (Index j=j2; j<j2+actual_cols; ++j)
             {
               if (TriStorageOrder==RowMajor)
               {
@@ -143,7 +153,7 @@ struct triangular_solve_matrix<Scalar,Index,OnTheLeft,Mode,Conjugate,TriStorageO
           Index blockBOffset = IsLower ? k1 : lengthTarget;
 
           // update the respective rows of B from other
-          pack_rhs(blockB, _other+startBlock, otherStride, actualPanelWidth, cols, actual_kc, blockBOffset);
+          pack_rhs(blockB+actual_kc*j2, &other(startBlock,j2), otherStride, actualPanelWidth, actual_cols, actual_kc, blockBOffset);
 
           // GEBP
           if (lengthTarget>0)
@@ -152,19 +162,19 @@ struct triangular_solve_matrix<Scalar,Index,OnTheLeft,Mode,Conjugate,TriStorageO
 
             pack_lhs(blockA, &tri(startTarget,startBlock), triStride, actualPanelWidth, lengthTarget);
 
-            gebp_kernel(_other+startTarget, otherStride, blockA, blockB, lengthTarget, actualPanelWidth, cols, Scalar(-1),
-                        actualPanelWidth, actual_kc, 0, blockBOffset);
+            gebp_kernel(&other(startTarget,j2), otherStride, blockA, blockB+actual_kc*j2, lengthTarget, actualPanelWidth, actual_cols, Scalar(-1),
+                        actualPanelWidth, actual_kc, 0, blockBOffset, blockW);
           }
         }
       }
-
-      // R2 = A2 * B => GEPP
+      
+      // R2 -= A21 * B => GEPP
       {
         Index start = IsLower ? k2+kc : 0;
         Index end   = IsLower ? size : k2-kc;
         for(Index i2=start; i2<end; i2+=mc)
         {
-          const Index actual_mc = std::min(mc,end-i2);
+          const Index actual_mc = (std::min)(mc,end-i2);
           if (actual_mc>0)
           {
             pack_lhs(blockA, &tri(i2, IsLower ? k2 : k2-kc), triStride, actual_kc, actual_mc);
@@ -174,9 +184,6 @@ struct triangular_solve_matrix<Scalar,Index,OnTheLeft,Mode,Conjugate,TriStorageO
         }
       }
     }
-
-    ei_aligned_stack_delete(Scalar, blockA, kc*mc);
-    ei_aligned_stack_delete(Scalar, allocatedBlockB, sizeB);
   }
 };
 
@@ -209,10 +216,10 @@ struct triangular_solve_matrix<Scalar,Index,OnTheRight,Mode,Conjugate,TriStorage
     Index nc = rows;  // cache block size along the N direction
     computeProductBlockingSizes<Scalar,Scalar,4>(kc, mc, nc);
 
-    Scalar* blockA = ei_aligned_stack_new(Scalar, kc*mc);
     std::size_t sizeW = kc*Traits::WorkSpaceFactor;
     std::size_t sizeB = sizeW + kc*size;
-    Scalar* allocatedBlockB = ei_aligned_stack_new(Scalar, sizeB);
+    ei_declare_aligned_stack_constructed_variable(Scalar, blockA, kc*mc, 0);
+    ei_declare_aligned_stack_constructed_variable(Scalar, allocatedBlockB, sizeB, 0);
     Scalar* blockB = allocatedBlockB + sizeW;
 
     conj_if<Conjugate> conj;
@@ -225,7 +232,7 @@ struct triangular_solve_matrix<Scalar,Index,OnTheRight,Mode,Conjugate,TriStorage
         IsLower ? k2>0 : k2<size;
         IsLower ? k2-=kc : k2+=kc)
     {
-      const Index actual_kc = std::min(IsLower ? k2 : size-k2, kc);
+      const Index actual_kc = (std::min)(IsLower ? k2 : size-k2, kc);
       Index actual_k2 = IsLower ? k2-actual_kc : k2 ;
 
       Index startPanel = IsLower ? 0 : k2+actual_kc;
@@ -254,7 +261,7 @@ struct triangular_solve_matrix<Scalar,Index,OnTheRight,Mode,Conjugate,TriStorage
 
       for(Index i2=0; i2<rows; i2+=mc)
       {
-        const Index actual_mc = std::min(mc,rows-i2);
+        const Index actual_mc = (std::min)(mc,rows-i2);
 
         // triangular solver kernel
         {
@@ -314,9 +321,6 @@ struct triangular_solve_matrix<Scalar,Index,OnTheRight,Mode,Conjugate,TriStorage
                       -1, -1, 0, 0, allocatedBlockB);
       }
     }
-
-    ei_aligned_stack_delete(Scalar, blockA, kc*mc);
-    ei_aligned_stack_delete(Scalar, allocatedBlockB, sizeB);
   }
 };
 

Eigen/src/Core/products/TriangularSolverVector.h

@@ -60,7 +60,7 @@ struct triangular_solve_vector<LhsScalar, RhsScalar, Index, OnTheLeft, Mode, Con
         IsLower ? pi<size : pi>0;
         IsLower ? pi+=PanelWidth : pi-=PanelWidth)
     {
-      Index actualPanelWidth = std::min(IsLower ? size - pi : pi, PanelWidth);
+      Index actualPanelWidth = (std::min)(IsLower ? size - pi : pi, PanelWidth);
 
       Index r = IsLower ? pi : size - pi; // remaining size
       if (r > 0)
@@ -114,7 +114,7 @@ struct triangular_solve_vector<LhsScalar, RhsScalar, Index, OnTheLeft, Mode, Con
         IsLower ? pi<size : pi>0;
         IsLower ? pi+=PanelWidth : pi-=PanelWidth)
     {
-      Index actualPanelWidth = std::min(IsLower ? size - pi : pi, PanelWidth);
+      Index actualPanelWidth = (std::min)(IsLower ? size - pi : pi, PanelWidth);
       Index startBlock = IsLower ? pi : pi-actualPanelWidth;
       Index endBlock = IsLower ? pi + actualPanelWidth : 0;
 

Eigen/src/Core/util/Constants.h

@@ -161,23 +161,68 @@ const unsigned int HereditaryBits = RowMajorBit
                                   | EvalBeforeNestingBit
                                   | EvalBeforeAssigningBit;
 
-// Possible values for the Mode parameter of triangularView()
+/** \defgroup enums Enumerations
+  * \ingroup Core_Module
+  *
+  * Various enumerations used in %Eigen. Many of these are used as template parameters.
+  */
+
+/** \ingroup enums
+  * Enum containing possible values for the \p Mode parameter of 
+  * MatrixBase::selfadjointView() and MatrixBase::triangularView(). */
 enum {
-  Lower=0x1, Upper=0x2, UnitDiag=0x4, ZeroDiag=0x8,
-  UnitLower=UnitDiag|Lower, UnitUpper=UnitDiag|Upper,
-  StrictlyLower=ZeroDiag|Lower, StrictlyUpper=ZeroDiag|Upper,
-  SelfAdjoint=0x10};
+  /** View matrix as a lower triangular matrix. */
+  Lower=0x1,                      
+  /** View matrix as an upper triangular matrix. */
+  Upper=0x2,                      
+  /** %Matrix has ones on the diagonal; to be used in combination with #Lower or #Upper. */
+  UnitDiag=0x4, 
+  /** %Matrix has zeros on the diagonal; to be used in combination with #Lower or #Upper. */
+  ZeroDiag=0x8,
+  /** View matrix as a lower triangular matrix with ones on the diagonal. */
+  UnitLower=UnitDiag|Lower, 
+  /** View matrix as an upper triangular matrix with ones on the diagonal. */
+  UnitUpper=UnitDiag|Upper,
+  /** View matrix as a lower triangular matrix with zeros on the diagonal. */
+  StrictlyLower=ZeroDiag|Lower, 
+  /** View matrix as an upper triangular matrix with zeros on the diagonal. */
+  StrictlyUpper=ZeroDiag|Upper,
+  /** Used in BandMatrix and SelfAdjointView to indicate that the matrix is self-adjoint. */
+  SelfAdjoint=0x10
+};
 
-enum { Unaligned=0, Aligned=1 };
-enum { ConditionalJumpCost = 5 };
+/** \ingroup enums
+  * Enum for indicating whether an object is aligned or not. */
+enum { 
+  /** Object is not correctly aligned for vectorization. */
+  Unaligned=0, 
+  /** Object is aligned for vectorization. */
+  Aligned=1 
+};
 
+/** \ingroup enums
+ * Enum used by DenseBase::corner() in Eigen2 compatibility mode. */
 // FIXME after the corner() API change, this was not needed anymore, except by AlignedBox
 // TODO: find out what to do with that. Adapt the AlignedBox API ?
 enum CornerType { TopLeft, TopRight, BottomLeft, BottomRight };
 
-enum DirectionType { Vertical, Horizontal, BothDirections };
-enum ProductEvaluationMode { NormalProduct, CacheFriendlyProduct };
+/** \ingroup enums
+  * Enum containing possible values for the \p Direction parameter of
+  * Reverse, PartialReduxExpr and VectorwiseOp. */
+enum DirectionType { 
+  /** For Reverse, all columns are reversed; 
+    * for PartialReduxExpr and VectorwiseOp, act on columns. */
+  Vertical, 
+  /** For Reverse, all rows are reversed; 
+    * for PartialReduxExpr and VectorwiseOp, act on rows. */
+  Horizontal, 
+  /** For Reverse, both rows and columns are reversed; 
+    * not used for PartialReduxExpr and VectorwiseOp. */
+  BothDirections 
+};
 
+/** \internal \ingroup enums
+  * Enum to specify how to traverse the entries of a matrix. */
 enum {
   /** \internal Default traversal, no vectorization, no index-based access */
   DefaultTraversal,
@@ -196,14 +241,32 @@ enum {
   InvalidTraversal
 };
 
+/** \internal \ingroup enums
+  * Enum to specify whether to unroll loops when traversing over the entries of a matrix. */
 enum {
+  /** \internal Do not unroll loops. */
   NoUnrolling,
+  /** \internal Unroll only the inner loop, but not the outer loop. */
   InnerUnrolling,
+  /** \internal Unroll both the inner and the outer loop. If there is only one loop, 
+    * because linear traversal is used, then unroll that loop. */
   CompleteUnrolling
 };
 
+/** \internal \ingroup enums
+  * Enum to specify whether to use the default (built-in) implementation or the specialization. */
 enum {
+  Specialized,
+  BuiltIn
+};
+
+/** \ingroup enums
+  * Enum containing possible values for the \p _Options template parameter of
+  * Matrix, Array and BandMatrix. */
+enum {
+  /** Storage order is column major (see \ref TopicStorageOrders). */
   ColMajor = 0,
+  /** Storage order is row major (see \ref TopicStorageOrders). */
   RowMajor = 0x1,  // it is only a coincidence that this is equal to RowMajorBit -- don't rely on that
   /** \internal Align the matrix itself if it is vectorizable fixed-size */
   AutoAlign = 0,
@@ -211,11 +274,13 @@ enum {
   DontAlign = 0x2
 };
 
-/** \brief Enum for specifying whether to apply or solve on the left or right. 
-  */
+/** \ingroup enums
+  * Enum for specifying whether to apply or solve on the left or right. */
 enum {
-  OnTheLeft = 1,  /**< \brief Apply transformation on the left. */
-  OnTheRight = 2  /**< \brief Apply transformation on the right. */
+  /** Apply transformation on the left. */
+  OnTheLeft = 1,  
+  /** Apply transformation on the right. */
+  OnTheRight = 2  
 };
 
 /* the following could as well be written:
@@ -239,53 +304,111 @@ namespace {
   EIGEN_UNUSED Default_t Default;
 }
 
+/** \internal \ingroup enums
+  * Used in AmbiVector. */
 enum {
   IsDense         = 0,
   IsSparse
 };
 
+/** \ingroup enums
+  * Used as template parameter in DenseCoeffBase and MapBase to indicate 
+  * which accessors should be provided. */
 enum AccessorLevels {
-  ReadOnlyAccessors, WriteAccessors, DirectAccessors, DirectWriteAccessors
+  /** Read-only access via a member function. */
+  ReadOnlyAccessors, 
+  /** Read/write access via member functions. */
+  WriteAccessors, 
+  /** Direct read-only access to the coefficients. */
+  DirectAccessors, 
+  /** Direct read/write access to the coefficients. */
+  DirectWriteAccessors
 };
 
+/** \ingroup enums
+  * Enum with options to give to various decompositions. */
 enum DecompositionOptions {
-  Pivoting            = 0x01, // LDLT,
-  NoPivoting          = 0x02, // LDLT,
-  ComputeFullU        = 0x04, // SVD,
-  ComputeThinU        = 0x08, // SVD,
-  ComputeFullV        = 0x10, // SVD,
-  ComputeThinV        = 0x20, // SVD,
-  EigenvaluesOnly     = 0x40, // all eigen solvers
-  ComputeEigenvectors = 0x80, // all eigen solvers
+  /** \internal Not used (meant for LDLT?). */
+  Pivoting            = 0x01, 
+  /** \internal Not used (meant for LDLT?). */
+  NoPivoting          = 0x02, 
+  /** Used in JacobiSVD to indicate that the square matrix U is to be computed. */
+  ComputeFullU        = 0x04,
+  /** Used in JacobiSVD to indicate that the thin matrix U is to be computed. */
+  ComputeThinU        = 0x08,
+  /** Used in JacobiSVD to indicate that the square matrix V is to be computed. */
+  ComputeFullV        = 0x10,
+  /** Used in JacobiSVD to indicate that the thin matrix V is to be computed. */
+  ComputeThinV        = 0x20,
+  /** Used in SelfAdjointEigenSolver and GeneralizedSelfAdjointEigenSolver to specify
+    * that only the eigenvalues are to be computed and not the eigenvectors. */
+  EigenvaluesOnly     = 0x40,
+  /** Used in SelfAdjointEigenSolver and GeneralizedSelfAdjointEigenSolver to specify
+    * that both the eigenvalues and the eigenvectors are to be computed. */
+  ComputeEigenvectors = 0x80,
+  /** \internal */
   EigVecMask = EigenvaluesOnly | ComputeEigenvectors,
+  /** Used in GeneralizedSelfAdjointEigenSolver to indicate that it should
+    * solve the generalized eigenproblem \f$ Ax = \lambda B x \f$. */
   Ax_lBx              = 0x100,
+  /** Used in GeneralizedSelfAdjointEigenSolver to indicate that it should
+    * solve the generalized eigenproblem \f$ ABx = \lambda x \f$. */
   ABx_lx              = 0x200,
+  /** Used in GeneralizedSelfAdjointEigenSolver to indicate that it should
+    * solve the generalized eigenproblem \f$ BAx = \lambda x \f$. */
   BAx_lx              = 0x400,
+  /** \internal */
   GenEigMask = Ax_lBx | ABx_lx | BAx_lx
 };
 
+/** \ingroup enums
+  * Possible values for the \p QRPreconditioner template parameter of JacobiSVD. */
 enum QRPreconditioners {
+  /** Do not specify what is to be done if the SVD of a non-square matrix is asked for. */
   NoQRPreconditioner,
+  /** Use a QR decomposition without pivoting as the first step. */
   HouseholderQRPreconditioner,
+  /** Use a QR decomposition with column pivoting as the first step. */
   ColPivHouseholderQRPreconditioner,
+  /** Use a QR decomposition with full pivoting as the first step. */
   FullPivHouseholderQRPreconditioner
 };
 
-/** \brief Enum for reporting the status of a computation.
-  */
+#ifdef Success
+#error The preprocessor symbol 'Success' is defined, possibly by the X11 header file X.h
+#endif
+
+/** \ingroups enums
+  * Enum for reporting the status of a computation. */
 enum ComputationInfo {
-  Success = 0,        /**< \brief Computation was successful. */
-  NumericalIssue = 1, /**< \brief The provided data did not satisfy the prerequisites. */
-  NoConvergence = 2   /**< \brief Iterative procedure did not converge. */
+  /** Computation was successful. */
+  Success = 0,        
+  /** The provided data did not satisfy the prerequisites. */
+  NumericalIssue = 1, 
+  /** Iterative procedure did not converge. */
+  NoConvergence = 2,
+  /** The inputs are invalid, or the algorithm has been properly called.
+    * When assertions are enabled, such errors trigger an assert. */
+  InvalidInput = 3
 };
 
+/** \ingroup enums
+  * Enum used to specify how a particular transformation is stored in a matrix.
+  * \sa Transform, Hyperplane::transform(). */
 enum TransformTraits {
+  /** Transformation is an isometry. */
   Isometry      = 0x1,
+  /** Transformation is an affine transformation stored as a (Dim+1)^2 matrix whose last row is 
+    * assumed to be [0 ... 0 1]. */
   Affine        = 0x2,
+  /** Transformation is an affine transformation stored as a (Dim) x (Dim+1) matrix. */
   AffineCompact = 0x10 | Affine,
+  /** Transformation is a general projective transformation stored as a (Dim+1)^2 matrix. */
   Projective    = 0x20
 };
 
+/** \internal \ingroup enums
+  * Enum used to choose between implementation depending on the computer architecture. */
 namespace Architecture
 {
   enum Type {
@@ -302,8 +425,12 @@ namespace Architecture
   };
 }
 
+/** \internal \ingroup enums
+  * Enum used as template parameter in GeneralProduct. */
 enum { CoeffBasedProductMode, LazyCoeffBasedProductMode, OuterProduct, InnerProduct, GemvProduct, GemmProduct };
 
+/** \internal \ingroup enums
+  * Enum used in experimental parallel implementation. */
 enum Action {GetAction, SetAction};
 
 /** The type used to identify a dense storage. */

Eigen/src/Core/util/DisableMSVCWarnings.h

@@ -1,17 +0,0 @@
-
-#ifdef _MSC_VER
-  // 4100 - unreferenced formal parameter (occurred e.g. in aligned_allocator::destroy(pointer p))
-  // 4101 - unreferenced local variable
-  // 4127 - conditional expression is constant
-  // 4181 - qualifier applied to reference type ignored
-  // 4211 - nonstandard extension used : redefined extern to static
-  // 4244 - 'argument' : conversion from 'type1' to 'type2', possible loss of data
-  // 4273 - QtAlignedMalloc, inconsistent DLL linkage
-  // 4324 - structure was padded due to declspec(align())
-  // 4512 - assignment operator could not be generated
-  // 4522 - 'class' : multiple assignment operators specified
-  // 4700 - uninitialized local variable 'xyz' used
-  // 4717 - 'function' : recursive on all control paths, function will cause runtime stack overflow
-  #pragma warning( push )
-  #pragma warning( disable : 4100 4101 4127 4181 4211 4244 4273 4324 4512 4522 4700 4717 )
-#endif

Eigen/src/Core/util/DisableStupidWarnings.h

@@ -0,0 +1,42 @@
+#ifndef EIGEN_WARNINGS_DISABLED
+#define EIGEN_WARNINGS_DISABLED
+
+#ifdef _MSC_VER
+  // 4100 - unreferenced formal parameter (occurred e.g. in aligned_allocator::destroy(pointer p))
+  // 4101 - unreferenced local variable
+  // 4127 - conditional expression is constant
+  // 4181 - qualifier applied to reference type ignored
+  // 4211 - nonstandard extension used : redefined extern to static
+  // 4244 - 'argument' : conversion from 'type1' to 'type2', possible loss of data
+  // 4273 - QtAlignedMalloc, inconsistent DLL linkage
+  // 4324 - structure was padded due to declspec(align())
+  // 4512 - assignment operator could not be generated
+  // 4522 - 'class' : multiple assignment operators specified
+  // 4700 - uninitialized local variable 'xyz' used
+  // 4717 - 'function' : recursive on all control paths, function will cause runtime stack overflow
+  #ifndef EIGEN_PERMANENTLY_DISABLE_STUPID_WARNINGS
+    #pragma warning( push )
+  #endif
+  #pragma warning( disable : 4100 4101 4127 4181 4211 4244 4273 4324 4512 4522 4700 4717 )
+#elif defined __INTEL_COMPILER
+  // 2196 - routine is both "inline" and "noinline" ("noinline" assumed)
+  //        ICC 12 generates this warning even without any inline keyword, when defining class methods 'inline' i.e. inside of class body
+  // 2536 - type qualifiers are meaningless here
+  //        ICC 12 generates this warning when a function return type is const qualified, even if that type is a template-parameter-dependent
+  //        typedef that may be a reference type.
+  // 279  - controlling expression is constant
+  //        ICC 12 generates this warning on assert(constant_expression_depending_on_template_params) and frankly this is a legitimate use case.
+  #ifndef EIGEN_PERMANENTLY_DISABLE_STUPID_WARNINGS
+    #pragma warning push
+  #endif
+  #pragma warning disable 2196 2536 279
+#elif defined __clang__
+  // -Wconstant-logical-operand - warning: use of logical && with constant operand; switch to bitwise & or remove constant
+  //     this is really a stupid warning as it warns on compile-time expressions involving enums
+  #ifndef EIGEN_PERMANENTLY_DISABLE_STUPID_WARNINGS
+    #pragma clang diagnostic push
+  #endif
+  #pragma clang diagnostic ignored "-Wconstant-logical-operand"
+#endif
+
+#endif // not EIGEN_WARNINGS_DISABLED

Eigen/src/Core/util/EnableMSVCWarnings.h

@@ -1,4 +0,0 @@
-
-#ifdef _MSC_VER
-  #pragma warning( pop )
-#endif

Eigen/src/Core/util/ForwardDeclarations.h

@@ -133,6 +133,7 @@ template<typename ExpressionType> class WithFormat;
 template<typename MatrixType> struct CommaInitializer;
 template<typename Derived> class ReturnByValue;
 template<typename ExpressionType> class ArrayWrapper;
+template<typename ExpressionType> class MatrixWrapper;
 
 namespace internal {
 template<typename DecompositionType, typename Rhs> struct solve_retval_base;
@@ -179,6 +180,9 @@ template<typename Scalar> struct scalar_exp_op;
 template<typename Scalar> struct scalar_log_op;
 template<typename Scalar> struct scalar_cos_op;
 template<typename Scalar> struct scalar_sin_op;
+template<typename Scalar> struct scalar_acos_op;
+template<typename Scalar> struct scalar_asin_op;
+template<typename Scalar> struct scalar_tan_op;
 template<typename Scalar> struct scalar_pow_op;
 template<typename Scalar> struct scalar_inverse_op;
 template<typename Scalar> struct scalar_square_op;
@@ -279,6 +283,8 @@ template<typename MatrixType,int Direction> class Homogeneous;
 // MatrixFunctions module
 template<typename Derived> struct MatrixExponentialReturnValue;
 template<typename Derived> class MatrixFunctionReturnValue;
+template<typename Derived> class MatrixSquareRootReturnValue;
+template<typename Derived> class MatrixLogarithmReturnValue;
 
 namespace internal {
 template <typename Scalar>

Eigen/src/Core/util/Macros.h

@@ -26,18 +26,31 @@
 #ifndef EIGEN_MACROS_H
 #define EIGEN_MACROS_H
 
-#define EIGEN_WORLD_VERSION 2
-#define EIGEN_MAJOR_VERSION 93
-#define EIGEN_MINOR_VERSION 0
+#define EIGEN_WORLD_VERSION 3
+#define EIGEN_MAJOR_VERSION 0
+#define EIGEN_MINOR_VERSION 91
 
 #define EIGEN_VERSION_AT_LEAST(x,y,z) (EIGEN_WORLD_VERSION>x || (EIGEN_WORLD_VERSION>=x && \
                                       (EIGEN_MAJOR_VERSION>y || (EIGEN_MAJOR_VERSION>=y && \
                                                                  EIGEN_MINOR_VERSION>=z))))
 #ifdef __GNUC__
-  #define EIGEN_GNUC_AT_LEAST(x,y) ((__GNUC__>=x && __GNUC_MINOR__>=y) || __GNUC__>x)
+  #define EIGEN_GNUC_AT_LEAST(x,y) ((__GNUC__==x && __GNUC_MINOR__>=y) || __GNUC__>x)
 #else
   #define EIGEN_GNUC_AT_LEAST(x,y) 0
 #endif
+ 
+#ifdef __GNUC__
+  #define EIGEN_GNUC_AT_MOST(x,y) ((__GNUC__==x && __GNUC_MINOR__<=y) || __GNUC__<x)
+#else
+  #define EIGEN_GNUC_AT_MOST(x,y) 0
+#endif
+
+#if EIGEN_GNUC_AT_MOST(4,3) && !defined(__clang__)
+  // see bug 89
+  #define EIGEN_SAFE_TO_USE_STANDARD_ASSERT_MACRO 0
+#else
+  #define EIGEN_SAFE_TO_USE_STANDARD_ASSERT_MACRO 1
+#endif
 
 #if defined(__GNUC__) && (__GNUC__ <= 3)
 #define EIGEN_GCC3_OR_OLDER 1
@@ -109,57 +122,42 @@
 
 #define EIGEN_DEBUG_VAR(x) std::cerr << #x << " = " << x << std::endl;
 
-#ifdef NDEBUG
-# ifndef EIGEN_NO_DEBUG
-#  define EIGEN_NO_DEBUG
-# endif
-#endif
+// concatenate two tokens
+#define EIGEN_CAT2(a,b) a ## b
+#define EIGEN_CAT(a,b) EIGEN_CAT2(a,b)
 
-#ifndef eigen_assert
-#ifdef EIGEN_NO_DEBUG
-#define eigen_assert(x)
-#else
-#define eigen_assert(x) assert(x)
-#endif
-#endif
+// convert a token to a string
+#define EIGEN_MAKESTRING2(a) #a
+#define EIGEN_MAKESTRING(a) EIGEN_MAKESTRING2(a)
 
-#ifdef EIGEN_INTERNAL_DEBUGGING
-#define eigen_internal_assert(x) eigen_assert(x)
-#else
-#define eigen_internal_assert(x)
-#endif
-
-#ifdef EIGEN_NO_DEBUG
-#define EIGEN_ONLY_USED_FOR_DEBUG(x) (void)x
-#else
-#define EIGEN_ONLY_USED_FOR_DEBUG(x)
-#endif
-
-// EIGEN_ALWAYS_INLINE_ATTRIB should be use in the declaration of function
-// which should be inlined even in debug mode.
-// FIXME with the always_inline attribute,
-// gcc 3.4.x reports the following compilation error:
-//   Eval.h:91: sorry, unimplemented: inlining failed in call to 'const Eigen::Eval<Derived> Eigen::MatrixBase<Scalar, Derived>::eval() const'
-//    : function body not available
-#if EIGEN_GNUC_AT_LEAST(4,0)
-#define EIGEN_ALWAYS_INLINE_ATTRIB __attribute__((always_inline))
-#else
-#define EIGEN_ALWAYS_INLINE_ATTRIB
-#endif
-
-#if EIGEN_GNUC_AT_LEAST(4,1)
+#if EIGEN_GNUC_AT_LEAST(4,1) && !defined(__clang__) && !defined(__INTEL_COMPILER)
 #define EIGEN_FLATTEN_ATTRIB __attribute__((flatten))
 #else
 #define EIGEN_FLATTEN_ATTRIB
 #endif
 
-// EIGEN_FORCE_INLINE means "inline as much as possible"
-#if (defined _MSC_VER) || (defined __intel_compiler)
+// EIGEN_STRONG_INLINE is a stronger version of the inline, using __forceinline on MSVC,
+// but it still doesn't use GCC's always_inline. This is useful in (common) situations where MSVC needs forceinline
+// but GCC is still doing fine with just inline.
+#if (defined _MSC_VER) || (defined __INTEL_COMPILER)
 #define EIGEN_STRONG_INLINE __forceinline
 #else
 #define EIGEN_STRONG_INLINE inline
 #endif
 
+// EIGEN_ALWAYS_INLINE is the stronget, it has the effect of making the function inline and adding every possible
+// attribute to maximize inlining. This should only be used when really necessary: in particular,
+// it uses __attribute__((always_inline)) on GCC, which most of the time is useless and can severely harm compile times.
+// FIXME with the always_inline attribute,
+// gcc 3.4.x reports the following compilation error:
+//   Eval.h:91: sorry, unimplemented: inlining failed in call to 'const Eigen::Eval<Derived> Eigen::MatrixBase<Scalar, Derived>::eval() const'
+//    : function body not available
+#if EIGEN_GNUC_AT_LEAST(4,0)
+#define EIGEN_ALWAYS_INLINE __attribute__((always_inline)) inline
+#else
+#define EIGEN_ALWAYS_INLINE EIGEN_STRONG_INLINE
+#endif
+
 #if (defined __GNUC__)
 #define EIGEN_DONT_INLINE __attribute__((noinline))
 #elif (defined _MSC_VER)
@@ -175,12 +173,77 @@
 #define EIGEN_DECLARE_FUNCTION_ALLOWING_MULTIPLE_DEFINITIONS
 #define EIGEN_DEFINE_FUNCTION_ALLOWING_MULTIPLE_DEFINITIONS inline
 
-#if (defined __GNUC__)
-#define EIGEN_DEPRECATED __attribute__((deprecated))
-#elif (defined _MSC_VER)
-#define EIGEN_DEPRECATED __declspec(deprecated)
+#ifdef NDEBUG
+# ifndef EIGEN_NO_DEBUG
+#  define EIGEN_NO_DEBUG
+# endif
+#endif
+
+// eigen_plain_assert is where we implement the workaround for the assert() bug in GCC <= 4.3, see bug 89
+#ifdef EIGEN_NO_DEBUG
+  #define eigen_plain_assert(x)
 #else
-#define EIGEN_DEPRECATED
+  #if EIGEN_SAFE_TO_USE_STANDARD_ASSERT_MACRO
+    namespace Eigen {
+    namespace internal {
+    inline bool copy_bool(bool b) { return b; }
+    }
+    }
+    #define eigen_plain_assert(x) assert(x)
+  #else
+    // work around bug 89
+    #include <cstdlib>   // for abort
+    #include <iostream>  // for std::cerr
+
+    namespace Eigen {
+    namespace internal {
+    // trivial function copying a bool. Must be EIGEN_DONT_INLINE, so we implement it after including Eigen headers.
+    // see bug 89.
+    namespace {
+    EIGEN_DONT_INLINE bool copy_bool(bool b) { return b; }
+    }
+    inline void assert_fail(const char *condition, const char *function, const char *file, int line)
+    {
+      std::cerr << "assertion failed: " << condition << " in function " << function << " at " << file << ":" << line << std::endl;
+      abort();
+    }
+    }
+    }
+    #define eigen_plain_assert(x) \
+      do { \
+        if(!Eigen::internal::copy_bool(x)) \
+          Eigen::internal::assert_fail(EIGEN_MAKESTRING(x), __PRETTY_FUNCTION__, __FILE__, __LINE__); \
+      } while(false)
+  #endif
+#endif
+
+// eigen_assert can be overridden
+#ifndef eigen_assert
+#define eigen_assert(x) eigen_plain_assert(x)
+#endif
+
+#ifdef EIGEN_INTERNAL_DEBUGGING
+#define eigen_internal_assert(x) eigen_assert(x)
+#else
+#define eigen_internal_assert(x)
+#endif
+
+#ifdef EIGEN_NO_DEBUG
+#define EIGEN_ONLY_USED_FOR_DEBUG(x) (void)x
+#else
+#define EIGEN_ONLY_USED_FOR_DEBUG(x)
+#endif
+
+#ifndef EIGEN_NO_DEPRECATED_WARNING
+  #if (defined __GNUC__)
+    #define EIGEN_DEPRECATED __attribute__((deprecated))
+  #elif (defined _MSC_VER)
+    #define EIGEN_DEPRECATED __declspec(deprecated)
+  #else
+    #define EIGEN_DEPRECATED
+  #endif
+#else
+  #define EIGEN_DEPRECATED
 #endif
 
 #if (defined __GNUC__)
@@ -193,7 +256,7 @@
 #define EIGEN_UNUSED_VARIABLE(var) (void)var;
 
 #if (defined __GNUC__)
-#define EIGEN_ASM_COMMENT(X)  asm("#"X)
+#define EIGEN_ASM_COMMENT(X)  asm("#" X)
 #else
 #define EIGEN_ASM_COMMENT(X)
 #endif
@@ -205,9 +268,7 @@
  * If we made alignment depend on whether or not EIGEN_VECTORIZE is defined, it would be impossible to link
  * vectorized and non-vectorized code.
  */
-#if !EIGEN_ALIGN_STATICALLY
-  #define EIGEN_ALIGN_TO_BOUNDARY(n)
-#elif (defined __GNUC__) || (defined __PGI) || (defined __IBMCPP__)
+#if (defined __GNUC__) || (defined __PGI) || (defined __IBMCPP__)
   #define EIGEN_ALIGN_TO_BOUNDARY(n) __attribute__((aligned(n)))
 #elif (defined _MSC_VER)
   #define EIGEN_ALIGN_TO_BOUNDARY(n) __declspec(align(n))
@@ -220,6 +281,14 @@
 
 #define EIGEN_ALIGN16 EIGEN_ALIGN_TO_BOUNDARY(16)
 
+#if EIGEN_ALIGN_STATICALLY
+#define EIGEN_USER_ALIGN_TO_BOUNDARY(n) EIGEN_ALIGN_TO_BOUNDARY(n)
+#define EIGEN_USER_ALIGN16 EIGEN_ALIGN16
+#else
+#define EIGEN_USER_ALIGN_TO_BOUNDARY(n)
+#define EIGEN_USER_ALIGN16
+#endif
+
 #ifdef EIGEN_DONT_USE_RESTRICT_KEYWORD
   #define EIGEN_RESTRICT
 #endif
@@ -244,14 +313,6 @@
 // just an empty macro !
 #define EIGEN_EMPTY
 
-// concatenate two tokens
-#define EIGEN_CAT2(a,b) a ## b
-#define EIGEN_CAT(a,b) EIGEN_CAT2(a,b)
-
-// convert a token to a string
-#define EIGEN_MAKESTRING2(a) #a
-#define EIGEN_MAKESTRING(a) EIGEN_MAKESTRING2(a)
-
 #if defined(_MSC_VER) && (!defined(__INTEL_COMPILER))
 #define EIGEN_INHERIT_ASSIGNMENT_EQUAL_OPERATOR(Derived) \
   using Base::operator =;
@@ -345,7 +406,7 @@
 #define EIGEN_MAKE_CWISE_BINARY_OP(METHOD,FUNCTOR) \
   template<typename OtherDerived> \
   EIGEN_STRONG_INLINE const CwiseBinaryOp<FUNCTOR<Scalar>, const Derived, const OtherDerived> \
-  METHOD(const EIGEN_CURRENT_STORAGE_BASE_CLASS<OtherDerived> &other) const \
+  (METHOD)(const EIGEN_CURRENT_STORAGE_BASE_CLASS<OtherDerived> &other) const \
   { \
     return CwiseBinaryOp<FUNCTOR<Scalar>, const Derived, const OtherDerived>(derived(), other.derived()); \
   }

Eigen/src/Core/util/Memory.h

@@ -82,6 +82,16 @@
 
 namespace internal {
 
+inline void throw_std_bad_alloc()
+{
+  #ifdef EIGEN_EXCEPTIONS
+    throw std::bad_alloc();
+  #else
+    std::size_t huge = -1;
+    new int[huge];
+  #endif
+}
+
 /*****************************************************************************
 *** Implementation of handmade aligned functions                           ***
 *****************************************************************************/
@@ -156,7 +166,7 @@ inline void* generic_aligned_realloc(void* ptr, size_t size, size_t old_size)
 
   if (ptr != 0)
   {
-    std::memcpy(newptr, ptr, std::min(size,old_size));
+    std::memcpy(newptr, ptr, (std::min)(size,old_size));
     aligned_free(ptr);
   }
 
@@ -167,14 +177,36 @@ inline void* generic_aligned_realloc(void* ptr, size_t size, size_t old_size)
 *** Implementation of portable aligned versions of malloc/free/realloc     ***
 *****************************************************************************/
 
+#ifdef EIGEN_NO_MALLOC
+inline void check_that_malloc_is_allowed()
+{
+  eigen_assert(false && "heap allocation is forbidden (EIGEN_NO_MALLOC is defined)");
+}
+#elif defined EIGEN_RUNTIME_NO_MALLOC
+inline bool is_malloc_allowed_impl(bool update, bool new_value = false)
+{
+  static bool value = true;
+  if (update == 1)
+    value = new_value;
+  return value;
+}
+inline bool is_malloc_allowed() { return is_malloc_allowed_impl(false); }
+inline bool set_is_malloc_allowed(bool new_value) { return is_malloc_allowed_impl(true, new_value); }
+inline void check_that_malloc_is_allowed()
+{
+  eigen_assert(is_malloc_allowed() && "heap allocation is forbidden (EIGEN_RUNTIME_NO_MALLOC is defined and g_is_malloc_allowed is false)");
+}
+#else 
+inline void check_that_malloc_is_allowed()
+{}
+#endif
+
 /** \internal Allocates \a size bytes. The returned pointer is guaranteed to have 16 bytes alignment.
-  * On allocation error, the returned pointer is null, and if exceptions are enabled then a std::bad_alloc is thrown.
+  * On allocation error, the returned pointer is null, and std::bad_alloc is thrown.
   */
 inline void* aligned_malloc(size_t size)
 {
-  #ifdef EIGEN_NO_MALLOC
-    eigen_assert(false && "heap allocation is forbidden (EIGEN_NO_MALLOC is defined)");
-  #endif
+  check_that_malloc_is_allowed();
 
   void *result;
   #if !EIGEN_ALIGN
@@ -191,10 +223,9 @@ inline void* aligned_malloc(size_t size)
     result = handmade_aligned_malloc(size);
   #endif
 
-  #ifdef EIGEN_EXCEPTIONS
-    if(result == 0)
-      throw std::bad_alloc();
-  #endif
+  if(!result && size)
+    throw_std_bad_alloc();
+
   return result;
 }
 
@@ -219,7 +250,7 @@ inline void aligned_free(void *ptr)
 /**
 * \internal
 * \brief Reallocates an aligned block of memory.
-* \throws std::bad_alloc if EIGEN_EXCEPTIONS are defined.
+* \throws std::bad_alloc on allocation failure
 **/
 inline void* aligned_realloc(void *ptr, size_t new_size, size_t old_size)
 {
@@ -247,10 +278,9 @@ inline void* aligned_realloc(void *ptr, size_t new_size, size_t old_size)
   result = handmade_aligned_realloc(ptr,new_size,old_size);
 #endif
 
-#ifdef EIGEN_EXCEPTIONS
-  if (result==0 && new_size!=0)
-    throw std::bad_alloc();
-#endif
+  if (!result && new_size)
+    throw_std_bad_alloc();
+
   return result;
 }
 
@@ -259,7 +289,7 @@ inline void* aligned_realloc(void *ptr, size_t new_size, size_t old_size)
 *****************************************************************************/
 
 /** \internal Allocates \a size bytes. If Align is true, then the returned ptr is 16-byte-aligned.
-  * On allocation error, the returned pointer is null, and if exceptions are enabled then a std::bad_alloc is thrown.
+  * On allocation error, the returned pointer is null, and a std::bad_alloc is thrown.
   */
 template<bool Align> inline void* conditional_aligned_malloc(size_t size)
 {
@@ -268,14 +298,11 @@ template<bool Align> inline void* conditional_aligned_malloc(size_t size)
 
 template<> inline void* conditional_aligned_malloc<false>(size_t size)
 {
-  #ifdef EIGEN_NO_MALLOC
-    eigen_assert(false && "heap allocation is forbidden (EIGEN_NO_MALLOC is defined)");
-  #endif
+  check_that_malloc_is_allowed();
 
   void *result = std::malloc(size);
-  #ifdef EIGEN_EXCEPTIONS
-    if(!result) throw std::bad_alloc();
-  #endif
+  if(!result && size)
+    throw_std_bad_alloc();
   return result;
 }
 
@@ -327,18 +354,27 @@ template<typename T> inline void destruct_elements_of_array(T *ptr, size_t size)
 *** Implementation of aligned new/delete-like functions                    ***
 *****************************************************************************/
 
+template<typename T>
+EIGEN_ALWAYS_INLINE void check_size_for_overflow(size_t size)
+{
+  if(size > size_t(-1) / sizeof(T))
+    throw_std_bad_alloc();
+}
+
 /** \internal Allocates \a size objects of type T. The returned pointer is guaranteed to have 16 bytes alignment.
-  * On allocation error, the returned pointer is undefined, but if exceptions are enabled then a std::bad_alloc is thrown.
+  * On allocation error, the returned pointer is undefined, but a std::bad_alloc is thrown.
   * The default constructor of T is called.
   */
 template<typename T> inline T* aligned_new(size_t size)
 {
+  check_size_for_overflow<T>(size);
   T *result = reinterpret_cast<T*>(aligned_malloc(sizeof(T)*size));
   return construct_elements_of_array(result, size);
 }
 
 template<typename T, bool Align> inline T* conditional_aligned_new(size_t size)
 {
+  check_size_for_overflow<T>(size);
   T *result = reinterpret_cast<T*>(conditional_aligned_malloc<Align>(sizeof(T)*size));
   return construct_elements_of_array(result, size);
 }
@@ -363,6 +399,8 @@ template<typename T, bool Align> inline void conditional_aligned_delete(T *ptr,
 
 template<typename T, bool Align> inline T* conditional_aligned_realloc_new(T* pts, size_t new_size, size_t old_size)
 {
+  check_size_for_overflow<T>(new_size);
+  check_size_for_overflow<T>(old_size);
   if(new_size < old_size)
     destruct_elements_of_array(pts+new_size, old_size-new_size);
   T *result = reinterpret_cast<T*>(conditional_aligned_realloc<Align>(reinterpret_cast<void*>(pts), sizeof(T)*new_size, sizeof(T)*old_size));
@@ -374,6 +412,7 @@ template<typename T, bool Align> inline T* conditional_aligned_realloc_new(T* pt
 
 template<typename T, bool Align> inline T* conditional_aligned_new_auto(size_t size)
 {
+  check_size_for_overflow<T>(size);
   T *result = reinterpret_cast<T*>(conditional_aligned_malloc<Align>(sizeof(T)*size));
   if(NumTraits<T>::RequireInitialization)
     construct_elements_of_array(result, size);
@@ -382,6 +421,8 @@ template<typename T, bool Align> inline T* conditional_aligned_new_auto(size_t s
 
 template<typename T, bool Align> inline T* conditional_aligned_realloc_new_auto(T* pts, size_t new_size, size_t old_size)
 {
+  check_size_for_overflow<T>(new_size);
+  check_size_for_overflow<T>(old_size);
   if(NumTraits<T>::RequireInitialization && (new_size < old_size))
     destruct_elements_of_array(pts+new_size, old_size-new_size);
   T *result = reinterpret_cast<T*>(conditional_aligned_realloc<Align>(reinterpret_cast<void*>(pts), sizeof(T)*new_size, sizeof(T)*old_size));
@@ -442,42 +483,116 @@ inline static Index first_aligned(const Scalar* array, Index size)
   }
 }
 
+
+// std::copy is much slower than memcpy, so let's introduce a smart_copy which
+// use memcpy on trivial types, i.e., on types that does not require an initialization ctor.
+template<typename T, bool UseMemcpy> struct smart_copy_helper;
+
+template<typename T> void smart_copy(const T* start, const T* end, T* target)
+{
+  smart_copy_helper<T,!NumTraits<T>::RequireInitialization>::run(start, end, target);
+}
+
+template<typename T> struct smart_copy_helper<T,true> {
+  inline static void run(const T* start, const T* end, T* target)
+  { memcpy(target, start, std::ptrdiff_t(end)-std::ptrdiff_t(start)); }
+};
+
+template<typename T> struct smart_copy_helper<T,false> {
+  inline static void run(const T* start, const T* end, T* target)
+  { std::copy(start, end, target); }
+};
+
+
 } // end namespace internal
 
 /*****************************************************************************
 *** Implementation of runtime stack allocation (falling back to malloc)    ***
 *****************************************************************************/
 
-/** \internal
-  * Allocates an aligned buffer of SIZE bytes on the stack if SIZE is smaller than
-  * EIGEN_STACK_ALLOCATION_LIMIT, and if stack allocation is supported by the platform
-  * (currently, this is Linux only). Otherwise the memory is allocated on the heap.
-  * Data allocated with ei_aligned_stack_alloc \b must be freed by calling
-  * ei_aligned_stack_free(PTR,SIZE).
-  * \code
-  * float * data = ei_aligned_stack_alloc(float,array.size());
-  * // ...
-  * ei_aligned_stack_free(data,float,array.size());
-  * \endcode
-  */
-#if (defined __linux__)
-  #define ei_aligned_stack_alloc(SIZE) (SIZE<=EIGEN_STACK_ALLOCATION_LIMIT) \
-                                    ? alloca(SIZE) \
-                                    : Eigen::internal::aligned_malloc(SIZE)
-  #define ei_aligned_stack_free(PTR,SIZE) if(SIZE>EIGEN_STACK_ALLOCATION_LIMIT) Eigen::internal::aligned_free(PTR)
-#elif defined(_MSC_VER)
-  #define ei_aligned_stack_alloc(SIZE) (SIZE<=EIGEN_STACK_ALLOCATION_LIMIT) \
-                                    ? _alloca(SIZE) \
-                                    : Eigen::internal::aligned_malloc(SIZE)
-  #define ei_aligned_stack_free(PTR,SIZE) if(SIZE>EIGEN_STACK_ALLOCATION_LIMIT) Eigen::internal::aligned_free(PTR)
-#else
-  #define ei_aligned_stack_alloc(SIZE) Eigen::internal::aligned_malloc(SIZE)
-  #define ei_aligned_stack_free(PTR,SIZE) Eigen::internal::aligned_free(PTR)
+// you can overwrite Eigen's default behavior regarding alloca by defining EIGEN_ALLOCA
+// to the appropriate stack allocation function
+#ifndef EIGEN_ALLOCA
+  #if (defined __linux__)
+    #define EIGEN_ALLOCA alloca
+  #elif defined(_MSC_VER)
+    #define EIGEN_ALLOCA _alloca
+  #endif
 #endif
 
-#define ei_aligned_stack_new(TYPE,SIZE) Eigen::internal::construct_elements_of_array(reinterpret_cast<TYPE*>(ei_aligned_stack_alloc(sizeof(TYPE)*SIZE)), SIZE)
-#define ei_aligned_stack_delete(TYPE,PTR,SIZE) do {Eigen::internal::destruct_elements_of_array<TYPE>(PTR, SIZE); \
-                                                   ei_aligned_stack_free(PTR,sizeof(TYPE)*SIZE);} while(0)
+namespace internal {
+
+// This helper class construct the allocated memory, and takes care of destructing and freeing the handled data
+// at destruction time. In practice this helper class is mainly useful to avoid memory leak in case of exceptions.
+template<typename T> class aligned_stack_memory_handler
+{
+  public:
+    /* Creates a stack_memory_handler responsible for the buffer \a ptr of size \a size.
+     * Note that \a ptr can be 0 regardless of the other parameters.
+     * This constructor takes care of constructing/initializing the elements of the buffer if required by the scalar type T (see NumTraits<T>::RequireInitialization).
+     * In this case, the buffer elements will also be destructed when this handler will be destructed.
+     * Finally, if \a dealloc is true, then the pointer \a ptr is freed.
+     **/
+    aligned_stack_memory_handler(T* ptr, size_t size, bool dealloc)
+      : m_ptr(ptr), m_size(size), m_deallocate(dealloc)
+    {
+      if(NumTraits<T>::RequireInitialization && m_ptr)
+        Eigen::internal::construct_elements_of_array(m_ptr, size);
+    }
+    ~aligned_stack_memory_handler()
+    {
+      if(NumTraits<T>::RequireInitialization && m_ptr)
+        Eigen::internal::destruct_elements_of_array<T>(m_ptr, m_size);
+      if(m_deallocate)
+        Eigen::internal::aligned_free(m_ptr);
+    }
+  protected:
+    T* m_ptr;
+    size_t m_size;
+    bool m_deallocate;
+};
+
+}
+
+/** \internal
+  * Declares, allocates and construct an aligned buffer named NAME of SIZE elements of type TYPE on the stack
+  * if SIZE is smaller than EIGEN_STACK_ALLOCATION_LIMIT, and if stack allocation is supported by the platform
+  * (currently, this is Linux and Visual Studio only). Otherwise the memory is allocated on the heap.
+  * The allocated buffer is automatically deleted when exiting the scope of this declaration.
+  * If BUFFER is non null, then the declared variable is simply an alias for BUFFER, and no allocation/deletion occurs.
+  * Here is an example:
+  * \code
+  * {
+  *   ei_declare_aligned_stack_constructed_variable(float,data,size,0);
+  *   // use data[0] to data[size-1]
+  * }
+  * \endcode
+  * The underlying stack allocation function can controlled with the EIGEN_ALLOCA preprocessor token.
+  */
+#ifdef EIGEN_ALLOCA
+
+  #ifdef __arm__
+    #define EIGEN_ALIGNED_ALLOCA(SIZE) reinterpret_cast<void*>((reinterpret_cast<size_t>(EIGEN_ALLOCA(SIZE+16)) & ~(size_t(15))) + 16)
+  #else
+    #define EIGEN_ALIGNED_ALLOCA EIGEN_ALLOCA
+  #endif
+
+  #define ei_declare_aligned_stack_constructed_variable(TYPE,NAME,SIZE,BUFFER) \
+    Eigen::internal::check_size_for_overflow<TYPE>(SIZE); \
+    TYPE* NAME = (BUFFER)!=0 ? (BUFFER) \
+               : reinterpret_cast<TYPE*>( \
+                      (sizeof(TYPE)*SIZE<=EIGEN_STACK_ALLOCATION_LIMIT) ? EIGEN_ALIGNED_ALLOCA(sizeof(TYPE)*SIZE) \
+                    : Eigen::internal::aligned_malloc(sizeof(TYPE)*SIZE) );  \
+    Eigen::internal::aligned_stack_memory_handler<TYPE> EIGEN_CAT(NAME,_stack_memory_destructor)((BUFFER)==0 ? NAME : 0,SIZE,sizeof(TYPE)*SIZE>EIGEN_STACK_ALLOCATION_LIMIT)
+
+#else
+
+  #define ei_declare_aligned_stack_constructed_variable(TYPE,NAME,SIZE,BUFFER) \
+    Eigen::internal::check_size_for_overflow<TYPE>(SIZE); \
+    TYPE* NAME = (BUFFER)!=0 ? BUFFER : reinterpret_cast<TYPE*>(Eigen::internal::aligned_malloc(sizeof(TYPE)*SIZE));    \
+    Eigen::internal::aligned_stack_memory_handler<TYPE> EIGEN_CAT(NAME,_stack_memory_destructor)((BUFFER)==0 ? NAME : 0,SIZE,true)
+    
+#endif
 
 
 /*****************************************************************************
@@ -592,12 +707,13 @@ public:
 
     size_type max_size() const throw()
     {
-        return std::numeric_limits<size_type>::max();
+        return (std::numeric_limits<size_type>::max)();
     }
 
-    pointer allocate( size_type num, const_pointer* hint = 0 )
+    pointer allocate( size_type num, const void* hint = 0 )
     {
-        static_cast<void>( hint ); // suppress unused variable warning
+        EIGEN_UNUSED_VARIABLE(hint);
+        internal::check_size_for_overflow<T>(num);
         return static_cast<pointer>( internal::aligned_malloc( num * sizeof(T) ) );
     }
 
@@ -832,7 +948,7 @@ inline int queryTopLevelCacheSize()
 {
   int l1, l2(-1), l3(-1);
   queryCacheSizes(l1,l2,l3);
-  return std::max(l2,l3);
+  return (std::max)(l2,l3);
 }
 
 } // end namespace internal

Eigen/src/Core/util/ReenableStupidWarnings.h

@@ -0,0 +1,14 @@
+#ifdef EIGEN_WARNINGS_DISABLED
+#undef EIGEN_WARNINGS_DISABLED
+
+#ifndef EIGEN_PERMANENTLY_DISABLE_STUPID_WARNINGS
+  #ifdef _MSC_VER
+    #pragma warning( pop )
+  #elif defined __INTEL_COMPILER
+    #pragma warning pop
+  #elif defined __clang__
+    #pragma clang diagnostic pop
+  #endif
+#endif
+
+#endif // EIGEN_WARNINGS_DISABLED

Eigen/src/Core/util/StaticAssert.h

@@ -70,6 +70,7 @@
         YOU_CALLED_A_DYNAMIC_SIZE_METHOD_ON_A_FIXED_SIZE_MATRIX_OR_VECTOR,
         UNALIGNED_LOAD_AND_STORE_OPERATIONS_UNIMPLEMENTED_ON_ALTIVEC,
         THIS_FUNCTION_IS_NOT_FOR_INTEGER_NUMERIC_TYPES,
+        FLOATING_POINT_ARGUMENT_PASSED__INTEGER_WAS_EXPECTED,
         NUMERIC_TYPE_MUST_BE_REAL,
         COEFFICIENT_WRITE_ACCESS_TO_SELFADJOINT_NOT_SUPPORTED,
         WRITING_TO_TRIANGULAR_PART_WITH_UNIT_DIAGONAL_IS_NOT_SUPPORTED,
@@ -95,7 +96,12 @@
         YOU_PERFORMED_AN_INVALID_TRANSFORMATION_CONVERSION,
         THIS_EXPRESSION_IS_NOT_A_LVALUE__IT_IS_READ_ONLY,
         YOU_ARE_TRYING_TO_USE_AN_INDEX_BASED_ACCESSOR_ON_AN_EXPRESSION_THAT_DOES_NOT_SUPPORT_THAT,
-        THIS_METHOD_IS_ONLY_FOR_1x1_EXPRESSIONS
+        THIS_METHOD_IS_ONLY_FOR_1x1_EXPRESSIONS,
+        THIS_METHOD_IS_ONLY_FOR_EXPRESSIONS_OF_BOOL,
+        THIS_METHOD_IS_ONLY_FOR_ARRAYS_NOT_MATRICES,
+        YOU_PASSED_A_ROW_VECTOR_BUT_A_COLUMN_VECTOR_WAS_EXPECTED,
+        YOU_PASSED_A_COLUMN_VECTOR_BUT_A_ROW_VECTOR_WAS_EXPECTED,
+        THE_INDEX_TYPE_MUST_BE_A_SIGNED_TYPE
       };
     };
 
@@ -195,4 +201,15 @@
       EIGEN_STATIC_ASSERT(internal::is_lvalue<Derived>::value, \
                           THIS_EXPRESSION_IS_NOT_A_LVALUE__IT_IS_READ_ONLY)
 
+#define EIGEN_STATIC_ASSERT_ARRAYXPR(Derived) \
+      EIGEN_STATIC_ASSERT((internal::is_same<typename internal::traits<Derived>::XprKind, ArrayXpr>::value), \
+                          THIS_METHOD_IS_ONLY_FOR_ARRAYS_NOT_MATRICES)
+
+#define EIGEN_STATIC_ASSERT_SAME_XPR_KIND(Derived1, Derived2) \
+      EIGEN_STATIC_ASSERT((internal::is_same<typename internal::traits<Derived1>::XprKind, \
+                                             typename internal::traits<Derived2>::XprKind \
+                                            >::value), \
+                          YOU_CANNOT_MIX_ARRAYS_AND_MATRICES)
+
+
 #endif // EIGEN_STATIC_ASSERT_H

Eigen/src/Core/util/XprHelper.h

@@ -125,10 +125,9 @@ class compute_matrix_flags
       aligned_bit =
       (
             ((Options&DontAlign)==0)
-        &&  packet_traits<Scalar>::Vectorizable
         && (
 #if EIGEN_ALIGN_STATICALLY
-             ((!is_dynamic_size_storage) && (((MaxCols*MaxRows) % packet_traits<Scalar>::size) == 0))
+             ((!is_dynamic_size_storage) && (((MaxCols*MaxRows*sizeof(Scalar)) % 16) == 0))
 #else
              0
 #endif

Eigen/src/Eigen2Support/CMakeLists.txt

@@ -4,3 +4,5 @@ INSTALL(FILES
   ${Eigen_Eigen2Support_SRCS}
   DESTINATION ${INCLUDE_INSTALL_DIR}/Eigen/src/Eigen2Support COMPONENT Devel
   )
+
+ADD_SUBDIRECTORY(Geometry)

Eigen/src/Eigen2Support/Cwise.h

@@ -55,6 +55,9 @@
   * Example: \include MatrixBase_cwise_const.cpp
   * Output: \verbinclude MatrixBase_cwise_const.out
   *
+  * This class can be extended with the help of the plugin mechanism described on the page
+  * \ref TopicCustomizingEigen by defining the preprocessor symbol \c EIGEN_CWISE_PLUGIN.
+  *
   * \sa MatrixBase::cwise() const, MatrixBase::cwise()
   */
 template<typename ExpressionType> class Cwise
@@ -79,13 +82,17 @@ template<typename ExpressionType> class Cwise
     const EIGEN_CWISE_BINOP_RETURN_TYPE(internal::scalar_quotient_op)
     operator/(const MatrixBase<OtherDerived> &other) const;
 
+    /** \deprecated ArrayBase::min() */
     template<typename OtherDerived>
     const EIGEN_CWISE_BINOP_RETURN_TYPE(internal::scalar_min_op)
-    min(const MatrixBase<OtherDerived> &other) const;
+    (min)(const MatrixBase<OtherDerived> &other) const
+    { return EIGEN_CWISE_BINOP_RETURN_TYPE(internal::scalar_min_op)(_expression(), other.derived()); }
 
+    /** \deprecated ArrayBase::max() */
     template<typename OtherDerived>
     const EIGEN_CWISE_BINOP_RETURN_TYPE(internal::scalar_max_op)
-    max(const MatrixBase<OtherDerived> &other) const;
+    (max)(const MatrixBase<OtherDerived> &other) const
+    { return EIGEN_CWISE_BINOP_RETURN_TYPE(internal::scalar_max_op)(_expression(), other.derived()); }
 
     const EIGEN_CWISE_UNOP_RETURN_TYPE(internal::scalar_abs_op)      abs() const;
     const EIGEN_CWISE_UNOP_RETURN_TYPE(internal::scalar_abs2_op)     abs2() const;