Fix several uninitialized member from ctor

Eigen/src/Eigenvalues/EigenSolver.h

@@ -110,7 +110,7 @@ template<typename _MatrixType> class EigenSolver
       *
       * \sa compute() for an example.
       */
-    EigenSolver() : m_eivec(), m_eivalues(), m_isInitialized(false), m_realSchur(), m_matT(), m_tmp() {}
+    EigenSolver() : m_eivec(), m_eivalues(), m_isInitialized(false), m_eigenvectorsOk(false), m_realSchur(), m_matT(), m_tmp() {}
 
     /** \brief Default constructor with memory preallocation
       *

Eigen/src/Eigenvalues/RealQZ.h

@@ -90,8 +90,9 @@ namespace Eigen {
         m_Z(size, size),
         m_workspace(size*2),
         m_maxIters(400),
-        m_isInitialized(false)
+        m_isInitialized(false),
-        { }
+        m_computeQZ(true)
+      {}
 
       /** \brief Constructor; computes real QZ decomposition of given matrices
        * 
@@ -108,9 +109,11 @@ namespace Eigen {
         m_Z(A.rows(),A.cols()),
         m_workspace(A.rows()*2),
         m_maxIters(400),
-        m_isInitialized(false) {
+        m_isInitialized(false),
-          compute(A, B, computeQZ);
+        m_computeQZ(true)
-        }
+      {
+        compute(A, B, computeQZ);
+      }
 
       /** \brief Returns matrix Q in the QZ decomposition. 
        *

Eigen/src/Eigenvalues/SelfAdjointEigenSolver.h

@@ -122,7 +122,8 @@ template<typename _MatrixType> class SelfAdjointEigenSolver
           m_eivalues(),
           m_subdiag(),
           m_info(InvalidInput),
-          m_isInitialized(false)
+          m_isInitialized(false),
+          m_eigenvectorsOk(false)
     { }
 
     /** \brief Constructor, pre-allocates memory for dynamic-size matrices.
@@ -142,7 +143,8 @@ template<typename _MatrixType> class SelfAdjointEigenSolver
         : m_eivec(size, size),
           m_eivalues(size),
           m_subdiag(size > 1 ? size - 1 : 1),
-          m_isInitialized(false)
+          m_isInitialized(false),
+          m_eigenvectorsOk(false)
     {}
 
     /** \brief Constructor; computes eigendecomposition of given matrix.
@@ -166,7 +168,8 @@ template<typename _MatrixType> class SelfAdjointEigenSolver
       : m_eivec(matrix.rows(), matrix.cols()),
         m_eivalues(matrix.cols()),
         m_subdiag(matrix.rows() > 1 ? matrix.rows() - 1 : 1),
-        m_isInitialized(false)
+        m_isInitialized(false),
+        m_eigenvectorsOk(false)
     {
       compute(matrix.derived(), options);
     }

Eigen/src/IterativeLinearSolvers/IncompleteCholesky.h

@@ -76,12 +76,12 @@ class IncompleteCholesky : public SparseSolverBase<IncompleteCholesky<Scalar,_Up
       *
       * \sa IncompleteCholesky(const MatrixType&)
       */
-    IncompleteCholesky() : m_initialShift(1e-3),m_factorizationIsOk(false) {}
+    IncompleteCholesky() : m_initialShift(1e-3),m_analysisIsOk(false),m_factorizationIsOk(false) {}
     
     /** Constructor computing the incomplete factorization for the given matrix \a matrix.
       */
     template<typename MatrixType>
-    IncompleteCholesky(const MatrixType& matrix) : m_initialShift(1e-3),m_factorizationIsOk(false)
+    IncompleteCholesky(const MatrixType& matrix) : m_initialShift(1e-3),m_analysisIsOk(false),m_factorizationIsOk(false)
     {
       compute(matrix);
     }

Eigen/src/PardisoSupport/PardisoSupport.h

@@ -123,6 +123,7 @@ class PardisoImpl : public SparseSolverBase<Derived>
     };
 
     PardisoImpl()
+      : m_analysisIsOk(false), m_factorizationIsOk(false), m_pt(0)
     {
       eigen_assert((sizeof(StorageIndex) >= sizeof(_INTEGER_t) && sizeof(StorageIndex) <= 8) && "Non-supported index type");
       m_iparm.setZero();

Eigen/src/SPQRSupport/SuiteSparseQRSupport.h

@@ -74,13 +74,35 @@ class SPQR : public SparseSolverBase<SPQR<_MatrixType> >
     };
   public:
     SPQR() 
-      : m_ordering(SPQR_ORDERING_DEFAULT), m_allow_tol(SPQR_DEFAULT_TOL), m_tolerance (NumTraits<Scalar>::epsilon()), m_useDefaultThreshold(true)
+      : m_analysisIsOk(false),
+        m_factorizationIsOk(false),
+        m_isRUpToDate(false),
+        m_ordering(SPQR_ORDERING_DEFAULT),
+        m_allow_tol(SPQR_DEFAULT_TOL),
+        m_tolerance (NumTraits<Scalar>::epsilon()),
+        m_cR(0),
+        m_E(0),
+        m_H(0),
+        m_HPinv(0),
+        m_HTau(0),
+        m_useDefaultThreshold(true)
     { 
       cholmod_l_start(&m_cc);
     }
     
     explicit SPQR(const _MatrixType& matrix)
-    : m_ordering(SPQR_ORDERING_DEFAULT), m_allow_tol(SPQR_DEFAULT_TOL), m_tolerance (NumTraits<Scalar>::epsilon()), m_useDefaultThreshold(true)
+      : m_analysisIsOk(false),
+        m_factorizationIsOk(false),
+        m_isRUpToDate(false),
+        m_ordering(SPQR_ORDERING_DEFAULT),
+        m_allow_tol(SPQR_DEFAULT_TOL),
+        m_tolerance (NumTraits<Scalar>::epsilon()),
+        m_cR(0),
+        m_E(0),
+        m_H(0),
+        m_HPinv(0),
+        m_HTau(0),
+        m_useDefaultThreshold(true)
     {
       cholmod_l_start(&m_cc);
       compute(matrix);

Eigen/src/SVD/BDCSVD.h

@@ -110,7 +110,7 @@ public:
    * The default constructor is useful in cases in which the user intends to
    * perform decompositions via BDCSVD::compute(const MatrixType&).
    */
-  BDCSVD() : m_algoswap(16), m_numIters(0)
+  BDCSVD() : m_algoswap(16), m_isTranspose(false), m_compU(false), m_compV(false), m_numIters(0)
   {}
 
 

Eigen/src/SVD/SVDBase.h

@@ -245,6 +245,10 @@ protected:
     : m_isInitialized(false),
       m_isAllocated(false),
       m_usePrescribedThreshold(false),
+      m_computeFullU(false),
+      m_computeThinU(false),
+      m_computeFullV(false),
+      m_computeThinV(false),
       m_computationOptions(0),
       m_rows(-1), m_cols(-1), m_diagSize(0)
   {

Eigen/src/SparseCholesky/SimplicialCholesky.h

@@ -80,11 +80,19 @@ class SimplicialCholeskyBase : public SparseSolverBase<Derived>
 
     /** Default constructor */
     SimplicialCholeskyBase()
-      : m_info(Success), m_shiftOffset(0), m_shiftScale(1)
+      : m_info(Success),
+        m_factorizationIsOk(false),
+        m_analysisIsOk(false),
+        m_shiftOffset(0),
+        m_shiftScale(1)
     {}
 
     explicit SimplicialCholeskyBase(const MatrixType& matrix)
-      : m_info(Success), m_shiftOffset(0), m_shiftScale(1)
+      : m_info(Success),
+        m_factorizationIsOk(false),
+        m_analysisIsOk(false),
+        m_shiftOffset(0),
+        m_shiftScale(1)
     {
       derived().compute(matrix);
     }