Import LAPACK: LAPACKE directory

lapack-netlib/LAPACKE/CMakeLists.txt

@@ -81,7 +81,7 @@ configure_file(${CMAKE_CURRENT_SOURCE_DIR}/cmake/lapacke-config-install.cmake.in
 install(FILES
   ${CMAKE_CURRENT_BINARY_DIR}/CMakeFiles/lapacke-config.cmake
   ${LAPACK_BINARY_DIR}/lapacke-config-version.cmake
-  DESTINATION lib/cmake/lapacke-${LAPACK_VERSION}
+  DESTINATION ${LIBRARY_DIR}/cmake/lapacke-${LAPACK_VERSION}
   )
 
 install(EXPORT lapacke-targets

lapack-netlib/LAPACKE/include/lapacke.h

@@ -180,14 +180,14 @@ lapack_int LAPACKE_zbdsqr( int matrix_layout, char uplo, lapack_int n,
                            lapack_int ldc );
 lapack_int LAPACKE_sbdsvdx( int matrix_layout, char uplo, char jobz, char range,
                            lapack_int n, float* d, float* e,
-                           lapack_int vl, lapack_int vu,
-                           lapack_int il, lapack_int iu, lapack_int ns,
+                           float vl, float vu,
+                           lapack_int il, lapack_int iu, lapack_int* ns,
                            float* s, float* z, lapack_int ldz,
                            lapack_int* superb );
 lapack_int LAPACKE_dbdsvdx( int matrix_layout, char uplo, char jobz, char range,
                            lapack_int n, double* d, double* e,
-                           lapack_int vl, lapack_int vu,
-                           lapack_int il, lapack_int iu, lapack_int ns,
+                           double vl, double vu,
+                           lapack_int il, lapack_int iu, lapack_int* ns,
                            double* s, double* z, lapack_int ldz,
                            lapack_int* superb );
 lapack_int LAPACKE_sdisna( char job, lapack_int m, lapack_int n, const float* d,
@@ -999,29 +999,29 @@ lapack_int LAPACKE_zgesvd( int matrix_layout, char jobu, char jobvt,
 
 lapack_int LAPACKE_sgesvdx( int matrix_layout, char jobu, char jobvt, char range,
                            lapack_int m, lapack_int n, float* a,
-                           lapack_int lda, lapack_int vl, lapack_int vu,
-                           lapack_int il, lapack_int iu, lapack_int ns,
+                           lapack_int lda, float vl, float vu,
+                           lapack_int il, lapack_int iu, lapack_int* ns,
                            float* s, float* u, lapack_int ldu,
                            float* vt, lapack_int ldvt,
                            lapack_int* superb );
 lapack_int LAPACKE_dgesvdx( int matrix_layout, char jobu, char jobvt, char range,
                            lapack_int m, lapack_int n, double* a,
-                           lapack_int lda, lapack_int vl, lapack_int vu,
-                           lapack_int il, lapack_int iu, lapack_int ns,
+                           lapack_int lda, double vl, double vu,
+                           lapack_int il, lapack_int iu, lapack_int* ns,
                            double* s, double* u, lapack_int ldu,
                            double* vt, lapack_int ldvt,
                            lapack_int* superb );
 lapack_int LAPACKE_cgesvdx( int matrix_layout, char jobu, char jobvt, char range,
                            lapack_int m, lapack_int n, lapack_complex_float* a,
-                           lapack_int lda, lapack_int vl, lapack_int vu,
-                           lapack_int il, lapack_int iu, lapack_int ns,
+                           lapack_int lda, float vl, float vu,
+                           lapack_int il, lapack_int iu, lapack_int* ns,
                            float* s, lapack_complex_float* u, lapack_int ldu,
                            lapack_complex_float* vt, lapack_int ldvt,
                            lapack_int* superb );
 lapack_int LAPACKE_zgesvdx( int matrix_layout, char jobu, char jobvt, char range,
                            lapack_int m, lapack_int n, lapack_complex_double* a,
-                           lapack_int lda, lapack_int vl, lapack_int vu,
-                           lapack_int il, lapack_int iu, lapack_int ns,
+                           lapack_int lda, double vl, double vu,
+                           lapack_int il, lapack_int iu, lapack_int* ns,
                            double* s, lapack_complex_double* u, lapack_int ldu,
                            lapack_complex_double* vt, lapack_int ldvt,
                            lapack_int* superb );
@@ -4738,17 +4738,17 @@ lapack_int LAPACKE_dbdsdc_work( int matrix_layout, char uplo, char compq,
                                 lapack_int* iwork );
 
 lapack_int LAPACKE_sbdsvdx_work( int matrix_layout, char uplo, char jobz, char range,
-                           		lapack_int n, float* d, float* e,
-                           		lapack_int vl, lapack_int vu,
-                           		lapack_int il, lapack_int iu, lapack_int ns,
-                           		float* s, float* z, lapack_int ldz,	
-                                float* work, lapack_int* iwork );
+                                 lapack_int n, float* d, float* e,
+                                 float vl, float vu,
+                                 lapack_int il, lapack_int iu, lapack_int* ns,
+                                 float* s, float* z, lapack_int ldz,
+                                 float* work, lapack_int* iwork );
 lapack_int LAPACKE_dbdsvdx_work( int matrix_layout, char uplo, char jobz, char range,
-                           		lapack_int n, double* d, double* e,
-                           		lapack_int vl, lapack_int vu,
-                           		lapack_int il, lapack_int iu, lapack_int ns,
-                           		double* s, double* z, lapack_int ldz,	
-                                double* work, lapack_int* iwork );
+                                 lapack_int n, double* d, double* e,
+                                 double vl, double vu,
+                                 lapack_int il, lapack_int iu, lapack_int* ns,
+                                 double* s, double* z, lapack_int ldz,
+                                 double* work, lapack_int* iwork );
 
 lapack_int LAPACKE_sbdsqr_work( int matrix_layout, char uplo, lapack_int n,
                                 lapack_int ncvt, lapack_int nru, lapack_int ncc,
@@ -5760,30 +5760,30 @@ lapack_int LAPACKE_zgesvd_work( int matrix_layout, char jobu, char jobvt,
 
 lapack_int LAPACKE_sgesvdx_work( int matrix_layout, char jobu, char jobvt, char range,
                            		lapack_int m, lapack_int n, float* a,
-                          		lapack_int lda, lapack_int vl, lapack_int vu,
-                           		lapack_int il, lapack_int iu, lapack_int ns,
+                          		lapack_int lda, float vl, float vu,
+                           		lapack_int il, lapack_int iu, lapack_int* ns,
                            		float* s, float* u, lapack_int ldu,
                            		float* vt, lapack_int ldvt,	
                                 float* work, lapack_int lwork, lapack_int* iwork );
 lapack_int LAPACKE_dgesvdx_work( int matrix_layout, char jobu, char jobvt, char range,
                            		lapack_int m, lapack_int n, double* a,
-                          		lapack_int lda, lapack_int vl, lapack_int vu,
-                           		lapack_int il, lapack_int iu, lapack_int ns,
+                          		lapack_int lda, double vl, double vu,
+                           		lapack_int il, lapack_int iu, lapack_int* ns,
                            		double* s, double* u, lapack_int ldu,
                            		double* vt, lapack_int ldvt,	
                                 double* work, lapack_int lwork, lapack_int* iwork );
 lapack_int LAPACKE_cgesvdx_work( int matrix_layout, char jobu, char jobvt, char range,
                            		lapack_int m, lapack_int n, lapack_complex_float* a,
-                          		lapack_int lda, lapack_int vl, lapack_int vu,
-                           		lapack_int il, lapack_int iu, lapack_int ns,
+                          		lapack_int lda, float vl, float vu,
+                           		lapack_int il, lapack_int iu, lapack_int* ns,
                            		float* s, lapack_complex_float* u, lapack_int ldu,
                            		lapack_complex_float* vt, lapack_int ldvt,	
                                 lapack_complex_float* work, lapack_int lwork,
                                 float* rwork, lapack_int* iwork );
 lapack_int LAPACKE_zgesvdx_work( int matrix_layout, char jobu, char jobvt, char range,
                            		lapack_int m, lapack_int n, lapack_complex_double* a,
-                          		lapack_int lda, lapack_int vl, lapack_int vu,
-                           		lapack_int il, lapack_int iu, lapack_int ns,
+                          		lapack_int lda, double vl, double vu,
+                           		lapack_int il, lapack_int iu, lapack_int* ns,
                            		double* s, lapack_complex_double* u, lapack_int ldu,
                            		lapack_complex_double* vt, lapack_int ldvt,	
                                 lapack_complex_double* work, lapack_int lwork, 
@@ -10472,11 +10472,11 @@ lapack_int LAPACKE_cbbcsd_work( int matrix_layout, char jobu1, char jobu2,
                                 float* b22d, float* b22e, float* rwork,
                                 lapack_int lrwork );
 lapack_int LAPACKE_cheswapr( int matrix_layout, char uplo, lapack_int n,
-                             lapack_complex_float* a, lapack_int i1,
-                             lapack_int i2 );
+                             lapack_complex_float* a, lapack_int lda,
+                             lapack_int i1, lapack_int i2 );
 lapack_int LAPACKE_cheswapr_work( int matrix_layout, char uplo, lapack_int n,
-                                  lapack_complex_float* a, lapack_int i1,
-                                  lapack_int i2 );
+                                  lapack_complex_float* a, lapack_int lda,
+                                  lapack_int i1, lapack_int i2 );
 lapack_int LAPACKE_chetri2( int matrix_layout, char uplo, lapack_int n,
                             lapack_complex_float* a, lapack_int lda,
                             const lapack_int* ipiv );
@@ -10502,17 +10502,17 @@ lapack_int LAPACKE_chetrs2_work( int matrix_layout, char uplo, lapack_int n,
                                  lapack_complex_float* work );
 lapack_int LAPACKE_csyconv( int matrix_layout, char uplo, char way, lapack_int n,
                             lapack_complex_float* a, lapack_int lda,
-                            const lapack_int* ipiv, lapack_complex_float* work  );
+                            const lapack_int* ipiv, lapack_complex_float* e  );
 lapack_int LAPACKE_csyconv_work( int matrix_layout, char uplo, char way,
                                  lapack_int n, lapack_complex_float* a,
                                  lapack_int lda, const lapack_int* ipiv,
-                                 lapack_complex_float* work );
+                                 lapack_complex_float* e );
 lapack_int LAPACKE_csyswapr( int matrix_layout, char uplo, lapack_int n,
-                             lapack_complex_float* a, lapack_int i1,
-                             lapack_int i2 );
+                             lapack_complex_float* a, lapack_int lda,
+                             lapack_int i1, lapack_int i2 );
 lapack_int LAPACKE_csyswapr_work( int matrix_layout, char uplo, lapack_int n,
-                                  lapack_complex_float* a, lapack_int i1,
-                                  lapack_int i2 );
+                                  lapack_complex_float* a, lapack_int lda,
+                                  lapack_int i1, lapack_int i2 );
 lapack_int LAPACKE_csytri2( int matrix_layout, char uplo, lapack_int n,
                             lapack_complex_float* a, lapack_int lda,
                             const lapack_int* ipiv );
@@ -10590,14 +10590,14 @@ lapack_int LAPACKE_cuncsd2by1( int matrix_layout, char jobu1, char jobu2,
                            char jobv1t, lapack_int m, lapack_int p, lapack_int q,
                            lapack_complex_float* x11, lapack_int ldx11,
                            lapack_complex_float* x21, lapack_int ldx21,
-                           lapack_complex_float* theta, lapack_complex_float* u1,
+                           float* theta, lapack_complex_float* u1,
                            lapack_int ldu1, lapack_complex_float* u2,
                            lapack_int ldu2, lapack_complex_float* v1t, lapack_int ldv1t );
 lapack_int LAPACKE_cuncsd2by1_work( int matrix_layout, char jobu1, char jobu2,
                                 char jobv1t, lapack_int m, lapack_int p,
                                 lapack_int q, lapack_complex_float* x11, lapack_int ldx11,
                                 lapack_complex_float* x21, lapack_int ldx21,
-                                lapack_complex_float* theta, lapack_complex_float* u1,
+                                float* theta, lapack_complex_float* u1,
                                 lapack_int ldu1, lapack_complex_float* u2, 
                                 lapack_int ldu2, lapack_complex_float* v1t,
                                 lapack_int ldv1t, lapack_complex_float* work, 
@@ -10669,20 +10669,22 @@ lapack_int LAPACKE_dorcsd2by1_work( int matrix_layout, char jobu1, char jobu2,
                            lapack_int ldu2, double* v1t, lapack_int ldv1t,
                            double* work, lapack_int lwork, lapack_int* iwork );
 lapack_int LAPACKE_dsyconv( int matrix_layout, char uplo, char way, lapack_int n,
-                            double* a, lapack_int lda, const lapack_int* ipiv, double* work);
+                            double* a, lapack_int lda, const lapack_int* ipiv, double* e);
 lapack_int LAPACKE_dsyconv_work( int matrix_layout, char uplo, char way,
                                  lapack_int n, double* a, lapack_int lda,
-                                 const lapack_int* ipiv, double* work );
+                                 const lapack_int* ipiv, double* e );
 lapack_int LAPACKE_dsyswapr( int matrix_layout, char uplo, lapack_int n,
-                             double* a, lapack_int i1, lapack_int i2 );
+                             double* a, lapack_int lda, lapack_int i1,
+                             lapack_int i2 );
 lapack_int LAPACKE_dsyswapr_work( int matrix_layout, char uplo, lapack_int n,
-                                  double* a, lapack_int i1, lapack_int i2 );
+                                  double* a, lapack_int lda, lapack_int i1,
+                                  lapack_int i2 );
 lapack_int LAPACKE_dsytri2( int matrix_layout, char uplo, lapack_int n,
                             double* a, lapack_int lda, const lapack_int* ipiv );
 lapack_int LAPACKE_dsytri2_work( int matrix_layout, char uplo, lapack_int n,
                                  double* a, lapack_int lda,
                                  const lapack_int* ipiv,
-                                 lapack_complex_double* work, lapack_int lwork );
+                                 double* work, lapack_int lwork );
 lapack_int LAPACKE_dsytri2x( int matrix_layout, char uplo, lapack_int n,
                              double* a, lapack_int lda, const lapack_int* ipiv,
                              lapack_int nb );
@@ -10762,20 +10764,22 @@ lapack_int LAPACKE_sorcsd2by1_work( int matrix_layout, char jobu1, char jobu2,
                            lapack_int ldu2, float* v1t, lapack_int ldv1t,
                            float* work, lapack_int lwork, lapack_int* iwork );
 lapack_int LAPACKE_ssyconv( int matrix_layout, char uplo, char way, lapack_int n,
-                            float* a, lapack_int lda, const lapack_int* ipiv, float* work );
+                            float* a, lapack_int lda, const lapack_int* ipiv, float* e );
 lapack_int LAPACKE_ssyconv_work( int matrix_layout, char uplo, char way,
                                  lapack_int n, float* a, lapack_int lda,
-                                 const lapack_int* ipiv, float* work );
+                                 const lapack_int* ipiv, float* e );
 lapack_int LAPACKE_ssyswapr( int matrix_layout, char uplo, lapack_int n,
-                             float* a, lapack_int i1, lapack_int i2 );
+                             float* a, lapack_int lda, lapack_int i1,
+                             lapack_int i2 );
 lapack_int LAPACKE_ssyswapr_work( int matrix_layout, char uplo, lapack_int n,
-                                  float* a, lapack_int i1, lapack_int i2 );
+                                  float* a, lapack_int lda, lapack_int i1,
+                                  lapack_int i2 );
 lapack_int LAPACKE_ssytri2( int matrix_layout, char uplo, lapack_int n, float* a,
                             lapack_int lda, const lapack_int* ipiv );
 lapack_int LAPACKE_ssytri2_work( int matrix_layout, char uplo, lapack_int n,
                                  float* a, lapack_int lda,
                                  const lapack_int* ipiv,
-                                 lapack_complex_float* work, lapack_int lwork );
+                                 float* work, lapack_int lwork );
 lapack_int LAPACKE_ssytri2x( int matrix_layout, char uplo, lapack_int n,
                              float* a, lapack_int lda, const lapack_int* ipiv,
                              lapack_int nb );
@@ -10813,11 +10817,11 @@ lapack_int LAPACKE_zbbcsd_work( int matrix_layout, char jobu1, char jobu2,
                                 double* b22d, double* b22e, double* rwork,
                                 lapack_int lrwork );
 lapack_int LAPACKE_zheswapr( int matrix_layout, char uplo, lapack_int n,
-                             lapack_complex_double* a, lapack_int i1,
-                             lapack_int i2 );
+                             lapack_complex_double* a, lapack_int lda,
+                             lapack_int i1, lapack_int i2 );
 lapack_int LAPACKE_zheswapr_work( int matrix_layout, char uplo, lapack_int n,
-                                  lapack_complex_double* a, lapack_int i1,
-                                  lapack_int i2 );
+                                  lapack_complex_double* a, lapack_int lda,
+                                  lapack_int i1, lapack_int i2 );
 lapack_int LAPACKE_zhetri2( int matrix_layout, char uplo, lapack_int n,
                             lapack_complex_double* a, lapack_int lda,
                             const lapack_int* ipiv );
@@ -10843,17 +10847,17 @@ lapack_int LAPACKE_zhetrs2_work( int matrix_layout, char uplo, lapack_int n,
                                  lapack_complex_double* work );
 lapack_int LAPACKE_zsyconv( int matrix_layout, char uplo, char way, lapack_int n,
                             lapack_complex_double* a, lapack_int lda,
-                            const lapack_int* ipiv, lapack_complex_double* work );
+                            const lapack_int* ipiv, lapack_complex_double* e );
 lapack_int LAPACKE_zsyconv_work( int matrix_layout, char uplo, char way,
                                  lapack_int n, lapack_complex_double* a,
                                  lapack_int lda, const lapack_int* ipiv,
-                                 lapack_complex_double* work );
+                                 lapack_complex_double* e );
 lapack_int LAPACKE_zsyswapr( int matrix_layout, char uplo, lapack_int n,
-                             lapack_complex_double* a, lapack_int i1,
-                             lapack_int i2 );
+                             lapack_complex_double* a, lapack_int lda,
+                             lapack_int i1, lapack_int i2 );
 lapack_int LAPACKE_zsyswapr_work( int matrix_layout, char uplo, lapack_int n,
-                                  lapack_complex_double* a, lapack_int i1,
-                                  lapack_int i2 );
+                                  lapack_complex_double* a, lapack_int lda,
+                                  lapack_int i1, lapack_int i2 );
 lapack_int LAPACKE_zsytri2( int matrix_layout, char uplo, lapack_int n,
                             lapack_complex_double* a, lapack_int lda,
                             const lapack_int* ipiv );
@@ -10931,14 +10935,14 @@ lapack_int LAPACKE_zuncsd2by1( int matrix_layout, char jobu1, char jobu2,
                            char jobv1t, lapack_int m, lapack_int p, lapack_int q,
                            lapack_complex_double* x11, lapack_int ldx11,
                            lapack_complex_double* x21, lapack_int ldx21,
-                           lapack_complex_double* theta, lapack_complex_double* u1,
+                           double* theta, lapack_complex_double* u1,
                            lapack_int ldu1, lapack_complex_double* u2,
                            lapack_int ldu2, lapack_complex_double* v1t, lapack_int ldv1t );
 lapack_int LAPACKE_zuncsd2by1_work( int matrix_layout, char jobu1, char jobu2,
                                 char jobv1t, lapack_int m, lapack_int p,
                                 lapack_int q, lapack_complex_double* x11, lapack_int ldx11,
                                 lapack_complex_double* x21, lapack_int ldx21,
-                                lapack_complex_double* theta, lapack_complex_double* u1,
+                                double* theta, lapack_complex_double* u1,
                                 lapack_int ldu1, lapack_complex_double* u2, 
                                 lapack_int ldu2, lapack_complex_double* v1t,
                                 lapack_int ldv1t, lapack_complex_double* work, 
@@ -11243,7 +11247,7 @@ lapack_int LAPACKE_stprfb_work( int matrix_layout, char side, char trans,
                                 lapack_int n, lapack_int k, lapack_int l,
                                 const float* v, lapack_int ldv, const float* t,
                                 lapack_int ldt, float* a, lapack_int lda,
-                                float* b, lapack_int ldb, const float* work,
+                                float* b, lapack_int ldb, float* work,
                                 lapack_int ldwork );
 lapack_int LAPACKE_dtprfb_work( int matrix_layout, char side, char trans,
                                 char direct, char storev, lapack_int m,
@@ -11251,7 +11255,7 @@ lapack_int LAPACKE_dtprfb_work( int matrix_layout, char side, char trans,
                                 const double* v, lapack_int ldv,
                                 const double* t, lapack_int ldt, double* a,
                                 lapack_int lda, double* b, lapack_int ldb,
-                                const double* work, lapack_int ldwork );
+                                double* work, lapack_int ldwork );
 lapack_int LAPACKE_ctprfb_work( int matrix_layout, char side, char trans,
                                 char direct, char storev, lapack_int m,
                                 lapack_int n, lapack_int k, lapack_int l,
@@ -11417,9 +11421,9 @@ lapack_int LAPACKE_zsyr_work( int matrix_layout, char uplo, lapack_int n,
                                   const lapack_complex_double* x,
                                   lapack_int incx, lapack_complex_double* a,
                                   lapack_int lda );
-void LAPACKE_ilaver( const lapack_int* vers_major,
-                     const lapack_int* vers_minor,
-                     const lapack_int* vers_patch );
+void LAPACKE_ilaver( lapack_int* vers_major,
+                     lapack_int* vers_minor,
+                     lapack_int* vers_patch );
 
 
 #define LAPACK_sgetrf LAPACK_GLOBAL(sgetrf,SGETRF)
@@ -14784,13 +14788,13 @@ void LAPACK_dbdsdc( char* uplo, char* compq, lapack_int* n, double* d,
                     lapack_int* iwork, lapack_int *info );
 void LAPACK_sbdsvdx( char* uplo, char* jobz, char* range,
                      lapack_int* n, float* d, float* e,
-                     lapack_int* vl, lapack_int* vu,
+                     float* vl, float* vu,
                      lapack_int* il, lapack_int* iu, lapack_int* ns,
                      float* s, float* z, lapack_int* ldz,
                      float* work, lapack_int *iwork, lapack_int *info );
 void LAPACK_dbdsvdx( char* uplo, char* jobz, char* range,
                      lapack_int* n, double* d, double* e,
-                     lapack_int* vl, lapack_int* vu,
+                     double* vl, double* vu,
                      lapack_int* il, lapack_int* iu, lapack_int* ns,
                      double* s, double* z, lapack_int* ldz,
                      double* work, lapack_int *iwork, lapack_int *info );
@@ -16115,24 +16119,24 @@ void LAPACK_zgesvd( char* jobu, char* jobvt, lapack_int* m, lapack_int* n,
                     lapack_complex_double* work, lapack_int* lwork,
                     double* rwork, lapack_int *info );
 void LAPACK_sgesvdx( char* jobu, char* jobvt, char* range, lapack_int* m, lapack_int* n,
-                    float* a, lapack_int* lda, lapack_int* vl, lapack_int* vu,
+                    float* a, lapack_int* lda, float* vl, float* vu,
                     lapack_int* il, lapack_int* iu, lapack_int* ns, float* s, float* u,
                     lapack_int* ldu, float* vt, lapack_int* ldvt, float* work,
                     lapack_int* lwork, lapack_int *iwork, lapack_int *info );
 void LAPACK_dgesvdx( char* jobu, char* jobvt, char* range, lapack_int* m, lapack_int* n,
-                    double* a, lapack_int* lda, lapack_int* vl, lapack_int* vu,
+                    double* a, lapack_int* lda, double* vl, double* vu,
                     lapack_int* il, lapack_int* iu, lapack_int* ns, double* s, double* u,
                     lapack_int* ldu, double* vt, lapack_int* ldvt, double* work,
                     lapack_int* lwork, lapack_int *iwork, lapack_int *info );
 void LAPACK_cgesvdx( char* jobu, char* jobvt, char* range, lapack_int* m, lapack_int* n,
-                    lapack_complex_float* a, lapack_int* lda, lapack_int* vl, lapack_int* vu,
+                    lapack_complex_float* a, lapack_int* lda, float* vl, float* vu,
                     lapack_int* il, lapack_int* iu, lapack_int* ns, float* s,
                     lapack_complex_float* u, lapack_int* ldu,
                     lapack_complex_float* vt, lapack_int* ldvt,
                     lapack_complex_float* work, lapack_int* lwork, float* rwork,
                     lapack_int *iwork, lapack_int *info );
 void LAPACK_zgesvdx( char* jobu, char* jobvt, char* range, lapack_int* m, lapack_int* n,
-                    lapack_complex_double* a, lapack_int* lda, lapack_int* vl, lapack_int* vu,
+                    lapack_complex_double* a, lapack_int* lda, double* vl, double* vu,
                     lapack_int* il, lapack_int* iu, lapack_int* ns, double* s,
                     lapack_complex_double* u, lapack_int* ldu,
                     lapack_complex_double* vt, lapack_int* ldvt,
@@ -17060,9 +17064,8 @@ void LAPACK_cbbcsd( char* jobu1, char* jobu2,
                     float* b12e, float* b21d, float* b21e,
                     float* b22d, float* b22e, float* rwork,
                     lapack_int* lrwork , lapack_int *info );
-void LAPACK_cheswapr( char* uplo, lapack_int* n,
-                      lapack_complex_float* a, lapack_int* i1,
-                      lapack_int* i2 );
+void LAPACK_cheswapr( char* uplo, lapack_int* n, lapack_complex_float* a,
+                      lapack_int* lda, lapack_int* i1, lapack_int* i2 );
 void LAPACK_chetri2( char* uplo, lapack_int* n,
                      lapack_complex_float* a, lapack_int* lda,
                      const lapack_int* ipiv,
@@ -17079,10 +17082,10 @@ void LAPACK_chetrs2( char* uplo, lapack_int* n,
 void LAPACK_csyconv( char* uplo, char* way,
                      lapack_int* n, lapack_complex_float* a,
                      lapack_int* lda, const lapack_int* ipiv,
-                     lapack_complex_float* work , lapack_int *info );
+                     lapack_complex_float* e , lapack_int *info );
 void LAPACK_csyswapr( char* uplo, lapack_int* n,
-                      lapack_complex_float* a, lapack_int* i1,
-                      lapack_int* i2 );
+                      lapack_complex_float* a, lapack_int* lda,
+                      lapack_int* i1, lapack_int* i2 );
 void LAPACK_csytri2( char* uplo, lapack_int* n,
                      lapack_complex_float* a, lapack_int* lda,
                      const lapack_int* ipiv,
@@ -17127,7 +17130,7 @@ void LAPACK_cuncsd2by1( char* jobu1, char* jobu2,
                     char* jobv1t, lapack_int* m, lapack_int* p,
                     lapack_int* q, lapack_complex_float* x11,
                     lapack_int* ldx11, lapack_complex_float* x21,
-                    lapack_int* ldx21, lapack_complex_float* theta,
+                    lapack_int* ldx21, float* theta,
                     lapack_complex_float* u1, lapack_int* ldu1,
                     lapack_complex_float* u2, lapack_int* ldu2,
                     lapack_complex_float* v1t, lapack_int* ldv1t,
@@ -17173,13 +17176,13 @@ void LAPACK_dorcsd2by1( char* jobu1, char* jobu2,
                     lapack_int* iwork , lapack_int *info );
 void LAPACK_dsyconv( char* uplo, char* way,
                      lapack_int* n, double* a, lapack_int* lda,
-                     const lapack_int* ipiv, double* work , lapack_int *info );
-void LAPACK_dsyswapr( char* uplo, lapack_int* n,
-                      double* a, lapack_int* i1, lapack_int* i2 );
+                     const lapack_int* ipiv, double* e , lapack_int *info );
+void LAPACK_dsyswapr( char* uplo, lapack_int* n, double* a,  
+                      lapack_int* lda, lapack_int* i1, lapack_int* i2 );
 void LAPACK_dsytri2( char* uplo, lapack_int* n,
                      double* a, lapack_int* lda,
                      const lapack_int* ipiv,
-                     lapack_complex_double* work, lapack_int* lwork , lapack_int *info );
+                     double* work, lapack_int* lwork , lapack_int *info );
 void LAPACK_dsytri2x( char* uplo, lapack_int* n,
                       double* a, lapack_int* lda,
                       const lapack_int* ipiv, double* work,
@@ -17227,13 +17230,13 @@ void LAPACK_sorcsd2by1( char* jobu1, char* jobu2,
                     lapack_int* iwork , lapack_int *info );
 void LAPACK_ssyconv( char* uplo, char* way,
                      lapack_int* n, float* a, lapack_int* lda,
-                     const lapack_int* ipiv, float* work , lapack_int *info );
-void LAPACK_ssyswapr( char* uplo, lapack_int* n,
-                      float* a, lapack_int* i1, lapack_int* i2 );
+                     const lapack_int* ipiv, float* e , lapack_int *info );
+void LAPACK_ssyswapr( char* uplo, lapack_int* n, float* a, 
+                      lapack_int* lda, lapack_int* i1, lapack_int* i2 );
 void LAPACK_ssytri2( char* uplo, lapack_int* n,
                      float* a, lapack_int* lda,
                      const lapack_int* ipiv,
-                     lapack_complex_float* work, lapack_int* lwork , lapack_int *info );
+                     float* work, lapack_int* lwork , lapack_int *info );
 void LAPACK_ssytri2x( char* uplo, lapack_int* n,
                       float* a, lapack_int* lda,
                       const lapack_int* ipiv, float* work,
@@ -17254,9 +17257,8 @@ void LAPACK_zbbcsd( char* jobu1, char* jobu2,
                     double* b12e, double* b21d, double* b21e,
                     double* b22d, double* b22e, double* rwork,
                     lapack_int* lrwork , lapack_int *info );
-void LAPACK_zheswapr( char* uplo, lapack_int* n,
-                      lapack_complex_double* a, lapack_int* i1,
-                      lapack_int* i2 );
+void LAPACK_zheswapr( char* uplo, lapack_int* n, lapack_complex_double* a,
+                      lapack_int* lda, lapack_int* i1, lapack_int* i2 );
 void LAPACK_zhetri2( char* uplo, lapack_int* n,
                      lapack_complex_double* a, lapack_int* lda,
                      const lapack_int* ipiv,
@@ -17274,9 +17276,9 @@ void LAPACK_zhetrs2( char* uplo, lapack_int* n,
 void LAPACK_zsyconv( char* uplo, char* way,
                      lapack_int* n, lapack_complex_double* a,
                      lapack_int* lda, const lapack_int* ipiv,
-                     lapack_complex_double* work , lapack_int *info );
-void LAPACK_zsyswapr( char* uplo, lapack_int* n,
-                      lapack_complex_double* a, lapack_int* i1,
+                     lapack_complex_double* e , lapack_int *info );
+void LAPACK_zsyswapr( char* uplo, lapack_int* n, lapack_complex_double* a, 
+                      lapack_int* lda, lapack_int* i1,
                       lapack_int* i2 );
 void LAPACK_zsytri2( char* uplo, lapack_int* n,
                      lapack_complex_double* a, lapack_int* lda,
@@ -17323,7 +17325,7 @@ void LAPACK_zuncsd2by1( char* jobu1, char* jobu2,
                     char* jobv1t, lapack_int* m, lapack_int* p,
                     lapack_int* q, lapack_complex_double* x11,
                     lapack_int* ldx11, lapack_complex_double* x21,
-                    lapack_int* ldx21, lapack_complex_double* theta,
+                    lapack_int* ldx21, double* theta,
                     lapack_complex_double* u1, lapack_int* ldu1,
                     lapack_complex_double* u2, lapack_int* ldu2,
                     lapack_complex_double* v1t, lapack_int* ldv1t,

lapack-netlib/LAPACKE/include/lapacke_config.h

@@ -34,13 +34,6 @@
 #ifndef _LAPACKE_CONFIG_H_
 #define _LAPACKE_CONFIG_H_
 
-// For Android prior to API 21 (no <complex> include)
-#if defined(__ANDROID__)
-#if __ANDROID_API__ < 21
-#define LAPACK_COMPLEX_STRUCTURE
-#endif
-#endif
-
 #ifdef __cplusplus
 #if defined(LAPACK_COMPLEX_CPP)
 #include <complex>

lapack-netlib/LAPACKE/include/lapacke_mangling.h

@@ -0,0 +1,17 @@
+#ifndef LAPACK_HEADER_INCLUDED
+#define LAPACK_HEADER_INCLUDED
+
+#ifndef LAPACK_GLOBAL
+#if defined(LAPACK_GLOBAL_PATTERN_LC) || defined(ADD_)
+#define LAPACK_GLOBAL(lcname,UCNAME)  lcname##_
+#elif defined(LAPACK_GLOBAL_PATTERN_UC) || defined(UPPER)
+#define LAPACK_GLOBAL(lcname,UCNAME)  UCNAME
+#elif defined(LAPACK_GLOBAL_PATTERN_MC) || defined(NOCHANGE)
+#define LAPACK_GLOBAL(lcname,UCNAME)  lcname
+#else
+#define LAPACK_GLOBAL(lcname,UCNAME)  lcname##_
+#endif
+#endif
+
+#endif
+

lapack-netlib/LAPACKE/src/Makefile

@@ -34,7 +34,7 @@
 #
 include ../../make.inc
 
-SRC_OBJA = \
+SRC_OBJ = \
 lapacke_cbbcsd.o \
 lapacke_cbbcsd_work.o \
 lapacke_cbdsqr.o \
@@ -996,9 +996,7 @@ lapacke_dsytri2.o \
 lapacke_dsytri2_work.o \
 lapacke_dsytri2x.o \
 lapacke_dsytri2x_work.o \
-lapacke_dsytri_work.o
-
-SRC_OBJB = \
+lapacke_dsytri_work.o \
 lapacke_dsytrs.o \
 lapacke_dsytrs_rook.o \
 lapacke_dsytrs2.o \
@@ -2191,8 +2189,7 @@ lapacke_slagsy_work.o \
 lapacke_zlagsy.o \
 lapacke_zlagsy_work.o
 
-ALLOBJA = $(SRC_OBJA)
-ALLOBJB = $(SRC_OBJB) $(MATGEN_OBJ)
+ALLOBJ = $(SRC_OBJ) $(MATGEN_OBJ) 
 
 ifdef USEXBLAS
 ALLXOBJ = $(SXLASRC) $(DXLASRC) $(CXLASRC) $(ZXLASRC)
@@ -2206,9 +2203,8 @@ OBJ_FILES := $(C_FILES:.o=.o)
 
 all: ../../$(LAPACKELIB)
 
-../../$(LAPACKELIB): $(ALLOBJA) $(ALLOBJB) $(ALLXOBJ) $(DEPRECATED)
-	$(ARCH) $(ARCHFLAGS) ../../$(LAPACKELIB) $(ALLOBJA)
-	$(ARCH) $(ARCHFLAGS) ../../$(LAPACKELIB) $(ALLOBJB) $(ALLXOBJ) $(DEPRECATED)
+../../$(LAPACKELIB): $(ALLOBJ) $(ALLXOBJ) $(DEPRECATED)
+	$(ARCH) $(ARCHFLAGS) ../../$(LAPACKELIB) $(ALLOBJ) $(ALLXOBJ) $(DEPRECATED)
 	$(RANLIB) ../../$(LAPACKELIB)
 
 .c.o:

lapack-netlib/LAPACKE/src/lapacke_cgejsv.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function cgejsv
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -119,7 +119,7 @@ lapack_int LAPACKE_cgejsv( int matrix_layout, char joba, char jobu, char jobv,
      (  (       ( LAPACKE_lsame( jobu, 'u' ) ||  LAPACKE_lsame( jobu, 'f' ) ) &&
                ( LAPACKE_lsame( jobv, 'v' ) ||  LAPACKE_lsame( jobv, 'j' ) ) &&
                ( LAPACKE_lsame( jobt, 't' ) ||  LAPACKE_lsame( joba, 'f' ) || LAPACKE_lsame( joba, 'g' ) ))? MAX(7,2*n) :
-                       1 ))))))));  
+                       7 ))))))));
     lapack_int* iwork = NULL;
     float* rwork = NULL;
     lapack_complex_float* cwork = NULL;
@@ -136,30 +136,29 @@ lapack_int LAPACKE_cgejsv( int matrix_layout, char joba, char jobu, char jobv,
     if( LAPACKE_cge_nancheck( matrix_layout, m, n, a, lda ) ) {
         return -10;
     }
-    if( LAPACKE_lsame( jobu, 'f' ) || LAPACKE_lsame( jobu, 'u' ) ||
-        LAPACKE_lsame( jobu, 'w' ) ) {
-        if( LAPACKE_cge_nancheck( matrix_layout, nu, n, u, ldu ) ) {
-            return -13;
-        }
-    }
-    if( LAPACKE_lsame( jobv, 'j' ) || LAPACKE_lsame( jobv, 'v' ) ||
-        LAPACKE_lsame( jobv, 'w' ) ) {
-        if( LAPACKE_cge_nancheck( matrix_layout, nv, n, v, ldv ) ) {
-            return -15;
-        }
-    }
 #endif
     /* Allocate memory for working array(s) */
-    iwork = (lapack_int*)LAPACKE_malloc( sizeof(lapack_int) * MAX(1,m+3*n) );
+    iwork = (lapack_int*)LAPACKE_malloc( sizeof(lapack_int) * MAX(3,m+2*n) );
     if( iwork == NULL ) {
         info = LAPACK_WORK_MEMORY_ERROR;
         goto exit_level_0;
     }
+    lwork = MAX( lwork, 1 );
+    { /* FIXUP LWORK */
+        int want_u = LAPACKE_lsame( jobu, 'u' ) || LAPACKE_lsame( jobu, 'f' );
+        int want_v = LAPACKE_lsame( jobv, 'v' ) || LAPACKE_lsame( jobv, 'j' );
+        int want_sce = LAPACKE_lsame( joba, 'e' ) || LAPACKE_lsame( joba, 'g' );
+        if( !want_u && !want_v && !want_sce )  lwork = MAX( lwork, 2*n+1 ); // 1.1
+        if( !want_u && !want_v && want_sce )   lwork = MAX( lwork, n*n+3*n ); // 1.2
+        if( want_u && LAPACKE_lsame( jobv, 'v' ) ) lwork = MAX( lwork, 5*n+2*n*n ); // 4.1
+        if( want_u && LAPACKE_lsame( jobv, 'j' ) ) lwork = MAX( lwork, 4*n+n*n ); // 4.2
+    }
     cwork = (lapack_complex_float*)LAPACKE_malloc( sizeof(lapack_complex_float) * lwork );
     if( cwork == NULL ) {
         info = LAPACK_WORK_MEMORY_ERROR;
         goto exit_level_1;
     }
+    lrwork = MAX3( lrwork, 7, n+2*m );
     rwork = (float*)LAPACKE_malloc( sizeof(float) * lrwork );
     if( rwork == NULL ) {
         info = LAPACK_WORK_MEMORY_ERROR;

lapack-netlib/LAPACKE/src/lapacke_cgejsv_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function cgejsv
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -46,7 +46,7 @@ lapack_int LAPACKE_cgejsv_work( int matrix_layout, char joba, char jobu,
     if( matrix_layout == LAPACK_COL_MAJOR ) {
         /* Call LAPACK function and adjust info */
         LAPACK_cgejsv( &joba, &jobu, &jobv, &jobr, &jobt, &jobp, &m, &n, a,
-                       &lda, sva, u, &ldu, v, &ldv, cwork, &lwork, rwork, &lwork, 
+                       &lda, sva, u, &ldu, v, &ldv, cwork, &lwork, rwork, &lrwork,
                        iwork, &info );
         if( info < 0 ) {
             info = info - 1;
@@ -54,6 +54,8 @@ lapack_int LAPACKE_cgejsv_work( int matrix_layout, char joba, char jobu,
     } else if( matrix_layout == LAPACK_ROW_MAJOR ) {
         lapack_int nu = LAPACKE_lsame( jobu, 'n' ) ? 1 : m;
         lapack_int nv = LAPACKE_lsame( jobv, 'n' ) ? 1 : n;
+        lapack_int ncols_u = LAPACKE_lsame( jobu, 'n' ) ? 1 :
+            LAPACKE_lsame( jobu, 'f' ) ? m : n;
         lapack_int lda_t = MAX(1,m);
         lapack_int ldu_t = MAX(1,nu);
         lapack_int ldv_t = MAX(1,nv);
@@ -66,7 +68,7 @@ lapack_int LAPACKE_cgejsv_work( int matrix_layout, char joba, char jobu,
             LAPACKE_xerbla( "LAPACKE_cgejsv_work", info );
             return info;
         }
-        if( ldu < n ) {
+        if( ldu < ncols_u ) {
             info = -14;
             LAPACKE_xerbla( "LAPACKE_cgejsv_work", info );
             return info;
@@ -86,7 +88,7 @@ lapack_int LAPACKE_cgejsv_work( int matrix_layout, char joba, char jobu,
         if( LAPACKE_lsame( jobu, 'f' ) || LAPACKE_lsame( jobu, 'u' ) ||
             LAPACKE_lsame( jobu, 'w' ) ) {
             u_t = (lapack_complex_float*)
-               LAPACKE_malloc( sizeof(lapack_complex_float) * ldu_t * MAX(1,n) );
+               LAPACKE_malloc( sizeof(lapack_complex_float) * ldu_t * MAX(1,ncols_u) );
             if( u_t == NULL ) {
                 info = LAPACK_TRANSPOSE_MEMORY_ERROR;
                 goto exit_level_1;
@@ -103,14 +105,6 @@ lapack_int LAPACKE_cgejsv_work( int matrix_layout, char joba, char jobu,
         }
         /* Transpose input matrices */
         LAPACKE_cge_trans( matrix_layout, m, n, a, lda, a_t, lda_t );
-        if( LAPACKE_lsame( jobu, 'f' ) || LAPACKE_lsame( jobu, 'u' ) ||
-            LAPACKE_lsame( jobu, 'w' ) ) {
-            LAPACKE_cge_trans( matrix_layout, nu, n, u, ldu, u_t, ldu_t );
-        }
-        if( LAPACKE_lsame( jobv, 'j' ) || LAPACKE_lsame( jobv, 'v' ) ||
-            LAPACKE_lsame( jobv, 'w' ) ) {
-            LAPACKE_cge_trans( matrix_layout, nv, n, v, ldv, v_t, ldv_t );
-        }
         /* Call LAPACK function and adjust info */
         LAPACK_cgejsv( &joba, &jobu, &jobv, &jobr, &jobt, &jobp, &m, &n, a_t,
                        &lda_t, sva, u_t, &ldu_t, v_t, &ldv_t, cwork, &lwork,
@@ -121,7 +115,7 @@ lapack_int LAPACKE_cgejsv_work( int matrix_layout, char joba, char jobu,
         /* Transpose output matrices */
         if( LAPACKE_lsame( jobu, 'f' ) || LAPACKE_lsame( jobu, 'u' ) ||
             LAPACKE_lsame( jobu, 'w' ) ) {
-            LAPACKE_cge_trans( LAPACK_COL_MAJOR, nu, n, u_t, ldu_t, u, ldu );
+            LAPACKE_cge_trans( LAPACK_COL_MAJOR, nu, ncols_u, u_t, ldu_t, u, ldu );
         }
         if( LAPACKE_lsame( jobv, 'j' ) || LAPACKE_lsame( jobv, 'v' ) ||
             LAPACKE_lsame( jobv, 'w' ) ) {

lapack-netlib/LAPACKE/src/lapacke_cgemqrt.c

@@ -28,7 +28,7 @@
 ******************************************************************************
 * Contents: Native high-level C interface to LAPACK function cgemqrt
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -40,6 +40,7 @@ lapack_int LAPACKE_cgemqrt( int matrix_layout, char side, char trans,
                             lapack_int ldt, lapack_complex_float* c,
                             lapack_int ldc )
 {
+    lapack_int nrows_v;
     lapack_int info = 0;
     lapack_complex_float* work = NULL;
     if( matrix_layout != LAPACK_COL_MAJOR && matrix_layout != LAPACK_ROW_MAJOR ) {
@@ -48,13 +49,15 @@ lapack_int LAPACKE_cgemqrt( int matrix_layout, char side, char trans,
     }
 #ifndef LAPACK_DISABLE_NAN_CHECK
     /* Optionally check input matrices for NaNs */
+    nrows_v = LAPACKE_lsame( side, 'L' ) ? m :
+                         ( LAPACKE_lsame( side, 'R' ) ? n : 0 );
     if( LAPACKE_cge_nancheck( matrix_layout, m, n, c, ldc ) ) {
         return -12;
     }
-    if( LAPACKE_cge_nancheck( matrix_layout, ldt, nb, t, ldt ) ) {
+    if( LAPACKE_cge_nancheck( matrix_layout, nb, k, t, ldt ) ) {
         return -10;
     }
-    if( LAPACKE_cge_nancheck( matrix_layout, ldv, k, v, ldv ) ) {
+    if( LAPACKE_cge_nancheck( matrix_layout, nrows_v, k, v, ldv ) ) {
         return -8;
     }
 #endif

lapack-netlib/LAPACKE/src/lapacke_cgesvdx.c

@@ -28,15 +28,15 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function cgesvdx
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_cgesvdx( int matrix_layout, char jobu, char jobvt, char range,
                            lapack_int m, lapack_int n, lapack_complex_float* a,
-                           lapack_int lda, lapack_int vl, lapack_int vu,
-                           lapack_int il, lapack_int iu, lapack_int ns,
+                           lapack_int lda, float vl, float vu,
+                           lapack_int il, lapack_int iu, lapack_int* ns,
                            float* s, lapack_complex_float* u, lapack_int ldu,
                            lapack_complex_float* vt, lapack_int ldvt,
                            lapack_int* superb )
@@ -44,9 +44,9 @@ lapack_int LAPACKE_cgesvdx( int matrix_layout, char jobu, char jobvt, char range
     lapack_int info = 0;
     lapack_int lwork = -1;
     lapack_complex_float* work = NULL;
-    lapack_complex_float work_query;    
+    lapack_complex_float work_query;
     float* rwork = NULL;
-    lapack_int lrwork = MIN(m,n)*(MIN(m,n)*2+15*MIN(m,n));
+    lapack_int lrwork = MAX(1, MIN(m,n)*(MIN(m,n)*2+15*MIN(m,n)));
     lapack_int* iwork = NULL;
     lapack_int i;
     if( matrix_layout != LAPACK_COL_MAJOR && matrix_layout != LAPACK_ROW_MAJOR ) {
@@ -68,18 +68,18 @@ lapack_int LAPACKE_cgesvdx( int matrix_layout, char jobu, char jobvt, char range
     }
     lwork = LAPACK_C2INT (work_query);
     /* Allocate memory for work arrays */
-    rwork = (float*)LAPACKE_malloc( sizeof(float) * lwork );
-    if( work == NULL ) {
-        info = LAPACK_WORK_MEMORY_ERROR;
-        goto exit_level_0;
-    }
     work = (lapack_complex_float*)
-    	LAPACKE_malloc( sizeof(lapack_complex_float) * lrwork );
+    	LAPACKE_malloc( sizeof(lapack_complex_float) * lwork );
     if( work == NULL ) {
         info = LAPACK_WORK_MEMORY_ERROR;
         goto exit_level_1;
     }
-    iwork = (lapack_int*)LAPACKE_malloc( sizeof(lapack_int) * (12*MIN(m,n)) );
+    rwork = (float*)LAPACKE_malloc( sizeof(float) * lrwork );
+    if( rwork == NULL ) {
+        info = LAPACK_WORK_MEMORY_ERROR;
+        goto exit_level_0;
+    }
+    iwork = (lapack_int*)LAPACKE_malloc( sizeof(lapack_int) * MAX(1,(12*MIN(m,n))) );
     if( iwork == NULL ) {
         info = LAPACK_WORK_MEMORY_ERROR;
         goto exit_level_2;

lapack-netlib/LAPACKE/src/lapacke_cgesvdx_work.c

@@ -28,15 +28,15 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function cgesvdx
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_cgesvdx_work( int matrix_layout, char jobu, char jobvt, char range,
                            		lapack_int m, lapack_int n, lapack_complex_float* a,
-                          		lapack_int lda, lapack_int vl, lapack_int vu,
-                           		lapack_int il, lapack_int iu, lapack_int ns,
+                          		lapack_int lda, float vl, float vu,
+                           		lapack_int il, lapack_int iu, lapack_int* ns,
                            		float* s, lapack_complex_float* u, lapack_int ldu,
                            		lapack_complex_float* vt, lapack_int ldvt,	
                                 lapack_complex_float* work, lapack_int lwork,
@@ -46,21 +46,23 @@ lapack_int LAPACKE_cgesvdx_work( int matrix_layout, char jobu, char jobvt, char
     if( matrix_layout == LAPACK_COL_MAJOR ) {
         /* Call LAPACK function and adjust info */
         LAPACK_cgesvdx( &jobu, &jobvt, &range, &m, &n, a, &lda, &vl, &vu,
-            			&il, &iu, &ns, s, u, &ldu, vt, &ldvt,
+            			&il, &iu, ns, s, u, &ldu, vt, &ldvt,
                         work, &lwork, rwork, iwork, &info );
         if( info < 0 ) {
             info = info - 1;
         }
     } else if( matrix_layout == LAPACK_ROW_MAJOR ) {
-        lapack_int nrows_u = ( LAPACKE_lsame( jobu, 'a' ) ||
-                             LAPACKE_lsame( jobu, 's' ) ) ? m : 1;
-        lapack_int ncols_u = LAPACKE_lsame( jobu, 'a' ) ? m :
-                             ( LAPACKE_lsame( jobu, 's' ) ? MIN(m,n) : 1);
-        lapack_int nrows_vt = LAPACKE_lsame( jobvt, 'a' ) ? n :
-                              ( LAPACKE_lsame( jobvt, 's' ) ? MIN(m,n) : 1);
+        lapack_int nrows_u = LAPACKE_lsame( jobu, 'v' ) ? m : 0;
+        lapack_int ncols_u = LAPACKE_lsame( jobu, 'v' ) ?
+                             ( LAPACKE_lsame( range, 'i' ) ? MAX(iu - il + 1, 0) : MIN(m,n)) : 0;
+        lapack_int nrows_vt = LAPACKE_lsame( jobvt, 'v' ) ?
+                             ( LAPACKE_lsame( range, 'i' ) ? MAX(iu - il + 1, 0) : MIN(m,n)) : 0;
+        lapack_int ncols_vt = LAPACKE_lsame( jobvt, 'v' ) ? n : 0;
+
         lapack_int lda_t = MAX(1,m);
         lapack_int ldu_t = MAX(1,nrows_u);
         lapack_int ldvt_t = MAX(1,nrows_vt);
+
         lapack_complex_float* a_t = NULL;
         lapack_complex_float* u_t = NULL;
         lapack_complex_float* vt_t = NULL;
@@ -75,7 +77,7 @@ lapack_int LAPACKE_cgesvdx_work( int matrix_layout, char jobu, char jobvt, char
             LAPACKE_xerbla( "LAPACKE_cgesvdx_work", info );
             return info;
         }
-        if( ldvt < n ) {
+        if( ldvt < ncols_vt ) {
             info = -18;
             LAPACKE_xerbla( "LAPACKE_cgesvdx_work", info );
             return info;
@@ -83,7 +85,7 @@ lapack_int LAPACKE_cgesvdx_work( int matrix_layout, char jobu, char jobvt, char
         /* Query optimal working array(s) size if requested */
         if( lwork == -1 ) {
             LAPACK_cgesvdx( &jobu, &jobvt, &range, &m, &n, a, &lda_t, &vl, &vu,
-            				&il, &iu, &ns, s, u, &ldu_t, vt,
+            				&il, &iu, ns, s, u, &ldu_t, vt,
                             &ldvt_t, work, &lwork, rwork, iwork, &info );
             return (info < 0) ? (info - 1) : info;
         }
@@ -94,7 +96,7 @@ lapack_int LAPACKE_cgesvdx_work( int matrix_layout, char jobu, char jobvt, char
             info = LAPACK_TRANSPOSE_MEMORY_ERROR;
             goto exit_level_0;
         }
-        if( LAPACKE_lsame( jobu, 'a' ) || LAPACKE_lsame( jobu, 's' ) ) {
+        if( LAPACKE_lsame( jobu, 'v' ) ) {
             u_t = (lapack_complex_float*)
                 LAPACKE_malloc( sizeof(lapack_complex_float) * ldu_t * MAX(1,ncols_u) );
             if( u_t == NULL ) {
@@ -102,7 +104,7 @@ lapack_int LAPACKE_cgesvdx_work( int matrix_layout, char jobu, char jobvt, char
                 goto exit_level_1;
             }
         }
-        if( LAPACKE_lsame( jobvt, 'a' ) || LAPACKE_lsame( jobvt, 's' ) ) {
+        if( LAPACKE_lsame( jobvt, 'v' ) ) {
             vt_t = (lapack_complex_float*)
                 LAPACKE_malloc( sizeof(lapack_complex_float) * ldvt_t * MAX(1,n) );
             if( vt_t == NULL ) {
@@ -113,28 +115,28 @@ lapack_int LAPACKE_cgesvdx_work( int matrix_layout, char jobu, char jobvt, char
         /* Transpose input matrices */
         LAPACKE_cge_trans( matrix_layout, m, n, a, lda, a_t, lda_t );
         /* Call LAPACK function and adjust info */
-        LAPACK_cgesvdx( &jobu, &jobvt, &range, &m, &n, a, &lda_t, &vl, &vu,
-            				&il, &iu, &ns, s, u, &ldu_t, vt,
+        LAPACK_cgesvdx( &jobu, &jobvt, &range, &m, &n, a_t, &lda_t, &vl, &vu,
+            				&il, &iu, ns, s, u_t, &ldu_t, vt_t,
                             &ldvt_t, work, &lwork, rwork, iwork, &info );
         if( info < 0 ) {
             info = info - 1;
         }
         /* Transpose output matrices */
         LAPACKE_cge_trans( LAPACK_COL_MAJOR, m, n, a_t, lda_t, a, lda );
-        if( LAPACKE_lsame( jobu, 'a' ) || LAPACKE_lsame( jobu, 's' ) ) {
+        if( LAPACKE_lsame( jobu, 'v' ) ) {
             LAPACKE_cge_trans( LAPACK_COL_MAJOR, nrows_u, ncols_u, u_t, ldu_t,
                                u, ldu );
         }
-        if( LAPACKE_lsame( jobvt, 'a' ) || LAPACKE_lsame( jobvt, 's' ) ) {
+        if( LAPACKE_lsame( jobvt, 'v' ) ) {
             LAPACKE_cge_trans( LAPACK_COL_MAJOR, nrows_vt, n, vt_t, ldvt_t, vt,
                                ldvt );
         }
         /* Release memory and exit */
-        if( LAPACKE_lsame( jobvt, 'a' ) || LAPACKE_lsame( jobvt, 's' ) ) {
+        if( LAPACKE_lsame( jobvt, 'v' ) ) {
             LAPACKE_free( vt_t );
         }
 exit_level_2:
-        if( LAPACKE_lsame( jobu, 'a' ) || LAPACKE_lsame( jobu, 's' ) ) {
+        if( LAPACKE_lsame( jobu, 'v' ) ) {
             LAPACKE_free( u_t );
         }
 exit_level_1:

lapack-netlib/LAPACKE/src/lapacke_cgesvj.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function cgesvj
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -52,8 +52,8 @@ lapack_int LAPACKE_cgesvj( int matrix_layout, char joba, char jobu, char jobv,
     }
 #ifndef LAPACK_DISABLE_NAN_CHECK
     /* Optionally check input matrices for NaNs */
-    nrows_v = LAPACKE_lsame( jobv, 'v' ) ? n :
-                         ( LAPACKE_lsame( jobv, 'a' ) ? mv : 1);
+    nrows_v = LAPACKE_lsame( jobv, 'v' ) ? MAX(0,n) :
+            ( LAPACKE_lsame( jobv, 'a' ) ? MAX(0,mv) : 0);
     if( LAPACKE_cge_nancheck( matrix_layout, m, n, a, lda ) ) {
         return -7;
     }

lapack-netlib/LAPACKE/src/lapacke_cgesvj_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function cgesvj
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -50,8 +50,8 @@ lapack_int LAPACKE_cgesvj_work( int matrix_layout, char joba, char jobu,
             info = info - 1;
         }
     } else if( matrix_layout == LAPACK_ROW_MAJOR ) {
-        lapack_int nrows_v = LAPACKE_lsame( jobv, 'v' ) ? n :
-                             ( LAPACKE_lsame( jobv, 'a' ) ? mv : 1);
+        lapack_int nrows_v = LAPACKE_lsame( jobv, 'v' ) ? MAX(0,n) :
+                           ( LAPACKE_lsame( jobv, 'a' ) ? MAX(0,mv) : 0);
         lapack_int lda_t = MAX(1,m);
         lapack_int ldv_t = MAX(1,nrows_v);
         lapack_complex_float* a_t = NULL;

lapack-netlib/LAPACKE/src/lapacke_cggsvd3_work.c

@@ -93,9 +93,9 @@ lapack_int LAPACKE_cggsvd3_work( int matrix_layout, char jobu, char jobv,
         }
         /* Query optimal working array(s) size if requested */
         if( lwork == -1 ) {
-            LAPACK_cggsvd3( &jobu, &jobv, &jobq, &m, &n, &p, k, l, a_t,
-                            &lda_t, b_t, &ldb_t, alpha, beta, u_t, &ldu_t,
-                            v_t, &ldv_t, q_t, &ldq_t, work, &lwork, rwork,
+            LAPACK_cggsvd3( &jobu, &jobv, &jobq, &m, &n, &p, k, l, a,
+                            &lda_t, b, &ldb_t, alpha, beta, u, &ldu_t,
+                            v, &ldv_t, q, &ldq_t, work, &lwork, rwork,
                             iwork, &info );
             return (info < 0) ? (info - 1) : info;
         }

lapack-netlib/LAPACKE/src/lapacke_cggsvp3_work.c

@@ -87,16 +87,16 @@ lapack_int LAPACKE_cggsvp3_work( int matrix_layout, char jobu, char jobv,
             LAPACKE_xerbla( "LAPACKE_cggsvp3_work", info );
             return info;
         }
-        if( ldv < m ) {
+        if( ldv < p ) {
             info = -19;
             LAPACKE_xerbla( "LAPACKE_cggsvp3_work", info );
             return info;
         }
         /* Query optimal working array(s) size if requested */
         if( lwork == -1 ) {
-          LAPACK_cggsvp3( &jobu, &jobv, &jobq, &m, &p, &n, a_t, &lda_t, b_t,
-                          &ldb_t, &tola, &tolb, k, l, u_t, &ldu_t, v_t, &ldv_t,
-                          q_t, &ldq_t, iwork, rwork, tau, work, &lwork, &info );
+          LAPACK_cggsvp3( &jobu, &jobv, &jobq, &m, &p, &n, a, &lda_t, b,
+                          &ldb_t, &tola, &tolb, k, l, u, &ldu_t, v, &ldv_t,
+                          q, &ldq_t, iwork, rwork, tau, work, &lwork, &info );
           return (info < 0) ? (info - 1) : info;
         }
         /* Allocate memory for temporary array(s) */
@@ -124,7 +124,7 @@ lapack_int LAPACKE_cggsvp3_work( int matrix_layout, char jobu, char jobv,
         if( LAPACKE_lsame( jobv, 'v' ) ) {
             v_t = (lapack_complex_float*)
                 LAPACKE_malloc( sizeof(lapack_complex_float) *
-                                ldv_t * MAX(1,m) );
+                                ldv_t * MAX(1,p) );
             if( v_t == NULL ) {
                 info = LAPACK_TRANSPOSE_MEMORY_ERROR;
                 goto exit_level_3;
@@ -156,7 +156,7 @@ lapack_int LAPACKE_cggsvp3_work( int matrix_layout, char jobu, char jobv,
             LAPACKE_cge_trans( LAPACK_COL_MAJOR, m, m, u_t, ldu_t, u, ldu );
         }
         if( LAPACKE_lsame( jobv, 'v' ) ) {
-            LAPACKE_cge_trans( LAPACK_COL_MAJOR, p, m, v_t, ldv_t, v, ldv );
+            LAPACKE_cge_trans( LAPACK_COL_MAJOR, p, p, v_t, ldv_t, v, ldv );
         }
         if( LAPACKE_lsame( jobq, 'q' ) ) {
             LAPACKE_cge_trans( LAPACK_COL_MAJOR, n, n, q_t, ldq_t, q, ldq );

lapack-netlib/LAPACKE/src/lapacke_chbtrd.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function chbtrd
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -49,7 +49,7 @@ lapack_int LAPACKE_chbtrd( int matrix_layout, char vect, char uplo, lapack_int n
     if( LAPACKE_chb_nancheck( matrix_layout, uplo, n, kd, ab, ldab ) ) {
         return -6;
     }
-    if( LAPACKE_lsame( vect, 'u' ) || LAPACKE_lsame( vect, 'v' ) ) {
+    if( LAPACKE_lsame( vect, 'u' ) ) {
         if( LAPACKE_cge_nancheck( matrix_layout, n, n, q, ldq ) ) {
             return -10;
         }

lapack-netlib/LAPACKE/src/lapacke_cheswapr.c

@@ -28,14 +28,14 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function cheswapr
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_cheswapr( int matrix_layout, char uplo, lapack_int n,
-                             lapack_complex_float* a, lapack_int i1,
-                             lapack_int i2 )
+                             lapack_complex_float* a, lapack_int lda,
+                             lapack_int i1, lapack_int i2 )
 {
     if( matrix_layout != LAPACK_COL_MAJOR && matrix_layout != LAPACK_ROW_MAJOR ) {
         LAPACKE_xerbla( "LAPACKE_cheswapr", -1 );
@@ -43,9 +43,9 @@ lapack_int LAPACKE_cheswapr( int matrix_layout, char uplo, lapack_int n,
     }
 #ifndef LAPACK_DISABLE_NAN_CHECK
     /* Optionally check input matrices for NaNs */
-    if( LAPACKE_che_nancheck( matrix_layout, uplo, n, a, n ) ) {
+    if( LAPACKE_che_nancheck( matrix_layout, uplo, n, a, lda ) ) {
         return -4;
     }
 #endif
-    return LAPACKE_cheswapr_work( matrix_layout, uplo, n, a, i1, i2 );
+    return LAPACKE_cheswapr_work( matrix_layout, uplo, n, a, lda, i1, i2 );
 }

lapack-netlib/LAPACKE/src/lapacke_cheswapr_work.c

@@ -28,38 +28,39 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function cheswapr
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_cheswapr_work( int matrix_layout, char uplo, lapack_int n,
-                                  lapack_complex_float* a, lapack_int i1,
-                                  lapack_int i2 )
+                                  lapack_complex_float* a, lapack_int lda,
+                                  lapack_int i1, lapack_int i2 )
 {
     lapack_int info = 0;
     if( matrix_layout == LAPACK_COL_MAJOR ) {
         /* Call LAPACK function and adjust info */
-        LAPACK_cheswapr( &uplo, &n, a, &i1, &i2 );
+        LAPACK_cheswapr( &uplo, &n, a, &lda, &i1, &i2 );
         if( info < 0 ) {
             info = info - 1;
         }
     } else if( matrix_layout == LAPACK_ROW_MAJOR ) {
+        lapack_int lda_t = MAX(1,n);
         lapack_complex_float* a_t = NULL;
         /* Allocate memory for temporary array(s) */
         a_t = (lapack_complex_float*)
-            LAPACKE_malloc( sizeof(lapack_complex_float) * n * MAX(1,n) );
+            LAPACKE_malloc( sizeof(lapack_complex_float) * lda_t * MAX(1,n) );
         if( a_t == NULL ) {
             info = LAPACK_TRANSPOSE_MEMORY_ERROR;
             goto exit_level_0;
         }
         /* Transpose input matrices */
-        LAPACKE_che_trans( matrix_layout, uplo, n, a, n, a_t, n );
+        LAPACKE_che_trans( matrix_layout, uplo, n, a, lda, a_t, lda_t );
         /* Call LAPACK function and adjust info */
-        LAPACK_cheswapr( &uplo, &n, a_t, &i1, &i2 );
+        LAPACK_cheswapr( &uplo, &n, a_t, &lda_t, &i1, &i2 );
         info = 0;  /* LAPACK call is ok! */
         /* Transpose output matrices */
-        LAPACKE_che_trans( LAPACK_COL_MAJOR, uplo, n, a_t, n, a, n );
+        LAPACKE_che_trans( LAPACK_COL_MAJOR, uplo, n, a_t, lda_t, a, lda );
         /* Release memory and exit */
         LAPACKE_free( a_t );
 exit_level_0:

lapack-netlib/LAPACKE/src/lapacke_chetri2x.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function chetri2x
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -45,7 +45,7 @@ lapack_int LAPACKE_chetri2x( int matrix_layout, char uplo, lapack_int n,
     }
 #ifndef LAPACK_DISABLE_NAN_CHECK
     /* Optionally check input matrices for NaNs */
-    if( LAPACKE_cge_nancheck( matrix_layout, lda, n, a, lda ) ) {
+    if( LAPACKE_cge_nancheck( matrix_layout, n, n, a, lda ) ) {
         return -4;
     }
 #endif

lapack-netlib/LAPACKE/src/lapacke_clantr.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function clantr
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"

lapack-netlib/LAPACKE/src/lapacke_clarfb_work.c

@@ -123,7 +123,7 @@ lapack_int LAPACKE_clarfb_work( int matrix_layout, char side, char trans,
             LAPACKE_cge_trans( matrix_layout, nrows_v, ncols_v-k, &v[k], ldv,
                                &v_t[k*ldv_t], ldv_t );
         } else if( LAPACKE_lsame( storev, 'r' ) &&
-                   LAPACKE_lsame( direct, 'b' ) ) {
+                   LAPACKE_lsame( direct, 'f' ) ) {
             if( k > ncols_v ) {
                 LAPACKE_xerbla( "LAPACKE_clarfb_work", -8 );
                 return -8;

lapack-netlib/LAPACKE/src/lapacke_clascl.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function slaswp
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -46,50 +46,64 @@ lapack_int LAPACKE_clascl( int matrix_layout, char type, lapack_int kl,
     /* Optionally check input matrices for NaNs */
     switch (type) {
     case 'G':
-       if( LAPACKE_cge_nancheck( matrix_layout, lda, n, a, lda ) ) {
-           return -9;
-           }
+        if( LAPACKE_cge_nancheck( matrix_layout, m, n, a, lda ) ) {
+            return -9;
+        }
         break;
     case 'L':
-       // TYPE = 'L' - lower triangular matrix.
-       if( LAPACKE_ctr_nancheck( matrix_layout, 'L', 'N', n, a, lda ) ) {
-           return -9;
-          }
+        // TYPE = 'L' - lower triangle of general matrix
+        if( matrix_layout == LAPACK_COL_MAJOR &&
+            LAPACKE_cgb_nancheck( matrix_layout, m, n, m-1, 0, a, lda+1 ) ) {
+            return -9;
+        }
+        if( matrix_layout == LAPACK_ROW_MAJOR &&
+            LAPACKE_cgb_nancheck( LAPACK_COL_MAJOR, n, m, 0, m-1, a-m+1, lda+1 ) ) {
+            return -9;
+        }
         break;
     case 'U':
-       // TYPE = 'U' - upper triangular matrix
-       if( LAPACKE_ctr_nancheck( matrix_layout, 'U', 'N', n, a, lda ) ) {
-           return -9;
-           } 
+        // TYPE = 'U' - upper triangle of general matrix
+        if( matrix_layout == LAPACK_COL_MAJOR &&
+            LAPACKE_cgb_nancheck( matrix_layout, m, n, 0, n-1, a-n+1, lda+1 ) ) {
+            return -9;
+        }
+        if( matrix_layout == LAPACK_ROW_MAJOR &&
+            LAPACKE_cgb_nancheck( LAPACK_COL_MAJOR, n, m, n-1, 0, a, lda+1 ) ) {
+            return -9;
+        }
         break;
     case 'H':
-       // TYPE = 'H' - upper Hessenberg matrix   
-       if( LAPACKE_chs_nancheck( matrix_layout, n, a, lda ) ) {
-           return -9;
-           }    
-        break;
+        // TYPE = 'H' - part of upper Hessenberg matrix in general matrix
+        if( matrix_layout == LAPACK_COL_MAJOR &&
+            LAPACKE_cgb_nancheck( matrix_layout, m, n, 1, n-1, a-n+1, lda+1 ) ) {
+            return -9;
+        }
+        if( matrix_layout == LAPACK_ROW_MAJOR &&
+            LAPACKE_cgb_nancheck( LAPACK_COL_MAJOR, n, m, n-1, 1, a-1, lda+1 ) ) {
+            return -9;
+        }
     case 'B':
-       // TYPE = 'B' - A is a symmetric band matrix with lower bandwidth KL
-       //             and upper bandwidth KU and with the only the lower
-       //             half stored.   
-       if( LAPACKE_chb_nancheck( matrix_layout, 'L', n, kl, a, lda ) ) {
-           return -9;
-           }
-         break;
-   case 'Q':
-       // TYPE = 'Q' - A is a symmetric band matrix with lower bandwidth KL
-       //             and upper bandwidth KU and with the only the upper
-       //             half stored.   
-       if( LAPACKE_chb_nancheck( matrix_layout, 'U', n, ku, a, lda ) ) {
-           return -9;
-           }
+        // TYPE = 'B' - lower part of symmetric band matrix (assume m==n)
+        if( LAPACKE_chb_nancheck( matrix_layout, 'L', n, kl, a, lda ) ) {
+            return -9;
+        }
+        break;
+    case 'Q':
+        // TYPE = 'Q' - upper part of symmetric band matrix (assume m==n)
+        if( LAPACKE_chb_nancheck( matrix_layout, 'U', n, ku, a, lda ) ) {
+            return -9;
+        }
         break;
     case 'Z':
-       // TYPE = 'Z' -  A is a band matrix with lower bandwidth KL and upper
-       //             bandwidth KU. See DGBTRF for storage details.        
-       if( LAPACKE_cgb_nancheck( matrix_layout, n, n, kl, kl+ku, a, lda ) ) {
-           return -6;
-           }
+        // TYPE = 'Z' -  band matrix laid out for ?GBTRF
+        if( matrix_layout == LAPACK_COL_MAJOR &&
+            LAPACKE_cgb_nancheck( matrix_layout, m, n, kl, ku, a+kl, lda ) ) {
+            return -9;
+        }
+        if( matrix_layout == LAPACK_ROW_MAJOR &&
+            LAPACKE_cgb_nancheck( matrix_layout, m, n, kl, ku, a+lda*kl, lda ) ) {
+            return -9;
+        }
         break;
     }
 #endif

lapack-netlib/LAPACKE/src/lapacke_clascl_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function slaswp
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -46,7 +46,10 @@ lapack_int LAPACKE_clascl_work( int matrix_layout, char type, lapack_int kl,
             info = info - 1;
         }
     } else if( matrix_layout == LAPACK_ROW_MAJOR ) {
-        lapack_int lda_t = MAX(1,lda);
+        lapack_int nrows_a = LAPACKE_lsame(type, 'b') ? kl + 1 :
+                             LAPACKE_lsame(type, 'q') ? ku + 1 :
+                             LAPACKE_lsame(type, 'z') ? 2 * kl + ku + 1 : m;
+        lapack_int lda_t = MAX(1,nrows_a);
         lapack_complex_float* a_t = NULL;
         /* Check leading dimension(s) */
         if( lda < n ) {
@@ -62,12 +65,14 @@ lapack_int LAPACKE_clascl_work( int matrix_layout, char type, lapack_int kl,
             goto exit_level_0;
         }
         /* Transpose input matrices */
-        LAPACKE_cge_trans( matrix_layout, lda, n, a, lda, a_t, lda_t );
+        LAPACKE_cge_trans( matrix_layout, nrows_a, n, a, lda, a_t, lda_t );
         /* Call LAPACK function and adjust info */
         LAPACK_clascl( &type, &kl, &ku, &cfrom, &cto, &m, &n, a_t, &lda_t, &info);
-        info = 0;  /* LAPACK call is ok! */
+        if( info < 0 ) {
+            info = info - 1;
+        }
         /* Transpose output matrices */
-        LAPACKE_cge_trans( LAPACK_COL_MAJOR, lda, n, a_t, lda_t, a, lda );
+        LAPACKE_cge_trans( LAPACK_COL_MAJOR, nrows_a, n, a_t, lda_t, a, lda );
         /* Release memory and exit */
         LAPACKE_free( a_t );
 exit_level_0:

lapack-netlib/LAPACKE/src/lapacke_cstedc.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function cstedc
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -58,7 +58,7 @@ lapack_int LAPACKE_cstedc( int matrix_layout, char compz, lapack_int n, float* d
     if( LAPACKE_s_nancheck( n-1, e, 1 ) ) {
         return -5;
     }
-    if( LAPACKE_lsame( compz, 'i' ) || LAPACKE_lsame( compz, 'v' ) ) {
+    if( LAPACKE_lsame( compz, 'v' ) ) {
         if( LAPACKE_cge_nancheck( matrix_layout, n, n, z, ldz ) ) {
             return -6;
         }

lapack-netlib/LAPACKE/src/lapacke_csteqr.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function csteqr
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -52,7 +52,7 @@ lapack_int LAPACKE_csteqr( int matrix_layout, char compz, lapack_int n, float* d
     if( LAPACKE_s_nancheck( n-1, e, 1 ) ) {
         return -5;
     }
-    if( LAPACKE_lsame( compz, 'i' ) || LAPACKE_lsame( compz, 'v' ) ) {
+    if( LAPACKE_lsame( compz, 'v' ) ) {
         if( LAPACKE_cge_nancheck( matrix_layout, n, n, z, ldz ) ) {
             return -6;
         }

lapack-netlib/LAPACKE/src/lapacke_csyconv.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function csyconv
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -36,7 +36,7 @@
 lapack_int LAPACKE_csyconv( int matrix_layout, char uplo, char way, lapack_int n,
                             lapack_complex_float* a, lapack_int lda,
                             const lapack_int* ipiv,
-                            lapack_complex_float* work )
+                            lapack_complex_float* e )
 {
     lapack_int info = 0;
     if( matrix_layout != LAPACK_COL_MAJOR && matrix_layout != LAPACK_ROW_MAJOR ) {
@@ -45,13 +45,13 @@ lapack_int LAPACKE_csyconv( int matrix_layout, char uplo, char way, lapack_int n
     }
 #ifndef LAPACK_DISABLE_NAN_CHECK
     /* Optionally check input matrices for NaNs */
-    if( LAPACKE_cge_nancheck( matrix_layout, lda, n, a, lda ) ) {
+    if( LAPACKE_cge_nancheck( matrix_layout, n, n, a, lda ) ) {
         return -5;
     }
 #endif
     /* Call middle-level interface */
     info = LAPACKE_csyconv_work( matrix_layout, uplo, way, n, a, lda, ipiv,
-                                 work );
+                                 e );
 exit_level_0:
     if( info == LAPACK_WORK_MEMORY_ERROR ) {
         LAPACKE_xerbla( "LAPACKE_csyconv", info );

lapack-netlib/LAPACKE/src/lapacke_csyconv_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function csyconv
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -36,12 +36,12 @@
 lapack_int LAPACKE_csyconv_work( int matrix_layout, char uplo, char way,
                                  lapack_int n, lapack_complex_float* a,
                                  lapack_int lda, const lapack_int* ipiv,
-                                 lapack_complex_float* work )
+                                 lapack_complex_float* e )
 {
     lapack_int info = 0;
     if( matrix_layout == LAPACK_COL_MAJOR ) {
         /* Call LAPACK function and adjust info */
-        LAPACK_csyconv( &uplo, &way, &n, a, &lda, ipiv, work, &info );
+        LAPACK_csyconv( &uplo, &way, &n, a, &lda, ipiv, e, &info );
         if( info < 0 ) {
             info = info - 1;
         }
@@ -64,7 +64,7 @@ lapack_int LAPACKE_csyconv_work( int matrix_layout, char uplo, char way,
         /* Transpose input matrices */
         LAPACKE_cge_trans( matrix_layout, lda, n, a, lda, a_t, lda_t );
         /* Call LAPACK function and adjust info */
-        LAPACK_csyconv( &uplo, &way, &n, a_t, &lda_t, ipiv, work, &info );
+        LAPACK_csyconv( &uplo, &way, &n, a_t, &lda_t, ipiv, e, &info );
         if( info < 0 ) {
             info = info - 1;
         }

lapack-netlib/LAPACKE/src/lapacke_csyswapr.c

@@ -28,14 +28,14 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function csyswapr
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_csyswapr( int matrix_layout, char uplo, lapack_int n,
-                             lapack_complex_float* a, lapack_int i1,
-                             lapack_int i2 )
+                             lapack_complex_float* a, lapack_int lda,
+                             lapack_int i1, lapack_int i2 )
 {
     if( matrix_layout != LAPACK_COL_MAJOR && matrix_layout != LAPACK_ROW_MAJOR ) {
         LAPACKE_xerbla( "LAPACKE_csyswapr", -1 );
@@ -43,9 +43,9 @@ lapack_int LAPACKE_csyswapr( int matrix_layout, char uplo, lapack_int n,
     }
 #ifndef LAPACK_DISABLE_NAN_CHECK
     /* Optionally check input matrices for NaNs */
-    if( LAPACKE_csy_nancheck( matrix_layout, uplo, n, a, n ) ) {
+    if( LAPACKE_csy_nancheck( matrix_layout, uplo, n, a, lda ) ) {
         return -4;
     }
 #endif
-    return LAPACKE_csyswapr_work( matrix_layout, uplo, n, a, i1, i2 );
+    return LAPACKE_csyswapr_work( matrix_layout, uplo, n, a, lda, i1, i2 );
 }

lapack-netlib/LAPACKE/src/lapacke_csyswapr_work.c

@@ -28,38 +28,39 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function csyswapr
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_csyswapr_work( int matrix_layout, char uplo, lapack_int n,
-                                  lapack_complex_float* a, lapack_int i1,
-                                  lapack_int i2 )
+                                  lapack_complex_float* a, lapack_int lda,
+                                  lapack_int i1, lapack_int i2 )
 {
     lapack_int info = 0;
     if( matrix_layout == LAPACK_COL_MAJOR ) {
         /* Call LAPACK function and adjust info */
-        LAPACK_csyswapr( &uplo, &n, a, &i1, &i2 );
+        LAPACK_csyswapr( &uplo, &n, a, &lda, &i1, &i2 );
         if( info < 0 ) {
             info = info - 1;
         }
     } else if( matrix_layout == LAPACK_ROW_MAJOR ) {
+        lapack_int lda_t = MAX(1,n);
         lapack_complex_float* a_t = NULL;
         /* Allocate memory for temporary array(s) */
         a_t = (lapack_complex_float*)
-            LAPACKE_malloc( sizeof(lapack_complex_float) * n * MAX(1,n) );
+            LAPACKE_malloc( sizeof(lapack_complex_float) * lda_t * MAX(1,n) );
         if( a_t == NULL ) {
             info = LAPACK_TRANSPOSE_MEMORY_ERROR;
             goto exit_level_0;
         }
         /* Transpose input matrices */
-        LAPACKE_csy_trans( matrix_layout, uplo, n, a, n, a_t, n );
+        LAPACKE_csy_trans( matrix_layout, uplo, n, a, lda, a_t, lda_t );
         /* Call LAPACK function and adjust info */
-        LAPACK_csyswapr( &uplo, &n, a_t, &i1, &i2 );
+        LAPACK_csyswapr( &uplo, &n, a_t, &lda_t, &i1, &i2 );
         info = 0;  /* LAPACK call is ok! */
         /* Transpose output matrices */
-        LAPACKE_csy_trans( LAPACK_COL_MAJOR, uplo, n, a_t, n, a, n );
+        LAPACKE_csy_trans( LAPACK_COL_MAJOR, uplo, n, a_t, lda_t, a, lda );
         /* Release memory and exit */
         LAPACKE_free( a_t );
 exit_level_0:

lapack-netlib/LAPACKE/src/lapacke_ctpmqrt.c

@@ -28,7 +28,7 @@
 ******************************************************************************
 * Contents: Native high-level C interface to LAPACK function ctpmqrt
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -41,6 +41,8 @@ lapack_int LAPACKE_ctpmqrt( int matrix_layout, char side, char trans,
                             lapack_complex_float* a, lapack_int lda,
                             lapack_complex_float* b, lapack_int ldb )
 {
+    lapack_int ncols_a, nrows_a;
+    lapack_int nrows_v;
     lapack_int info = 0;
     lapack_complex_float* work = NULL;
     if( matrix_layout != LAPACK_COL_MAJOR && matrix_layout != LAPACK_ROW_MAJOR ) {
@@ -49,16 +51,22 @@ lapack_int LAPACKE_ctpmqrt( int matrix_layout, char side, char trans,
     }
 #ifndef LAPACK_DISABLE_NAN_CHECK
     /* Optionally check input matrices for NaNs */
-    if( LAPACKE_cge_nancheck( matrix_layout, k, m, a, lda ) ) {
+    ncols_a = LAPACKE_lsame( side, 'L' ) ? n :
+                         ( LAPACKE_lsame( side, 'R' ) ? k : 0 );
+    nrows_a = LAPACKE_lsame( side, 'L' ) ? k :
+                         ( LAPACKE_lsame( side, 'R' ) ? m : 0 );
+    nrows_v = LAPACKE_lsame( side, 'L' ) ? m :
+                         ( LAPACKE_lsame( side, 'R' ) ? n : 0 );
+    if( LAPACKE_cge_nancheck( matrix_layout, nrows_a, ncols_a, a, lda ) ) {
         return -13;
     }
     if( LAPACKE_cge_nancheck( matrix_layout, m, n, b, ldb ) ) {
         return -15;
     }
-    if( LAPACKE_cge_nancheck( matrix_layout, ldt, nb, t, ldt ) ) {
+    if( LAPACKE_cge_nancheck( matrix_layout, nb, k, t, ldt ) ) {
         return -11;
     }
-    if( LAPACKE_cge_nancheck( matrix_layout, ldv, k, v, ldv ) ) {
+    if( LAPACKE_cge_nancheck( matrix_layout, nrows_v, k, v, ldv ) ) {
         return -9;
     }
 #endif

lapack-netlib/LAPACKE/src/lapacke_ctprfb.c

@@ -28,7 +28,7 @@
 ******************************************************************************
 * Contents: Native high-level C interface to LAPACK function ctprfb
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -41,6 +41,7 @@ lapack_int LAPACKE_ctprfb( int matrix_layout, char side, char trans, char direct
                            lapack_complex_float* a, lapack_int lda,
                            lapack_complex_float* b, lapack_int ldb )
 {
+    lapack_int ncols_v, nrows_v;
     lapack_int info = 0;
     lapack_int ldwork;
     lapack_int work_size;
@@ -51,16 +52,28 @@ lapack_int LAPACKE_ctprfb( int matrix_layout, char side, char trans, char direct
     }
 #ifndef LAPACK_DISABLE_NAN_CHECK
     /* Optionally check input matrices for NaNs */
+    if( LAPACKE_lsame( storev, 'C' ) ) {
+        ncols_v = k;
+        nrows_v = LAPACKE_lsame( side, 'L' ) ? m :
+                             ( LAPACKE_lsame( side, 'R' ) ? n : 0 );
+    } else if( LAPACKE_lsame( storev, 'R' ) ) {
+        ncols_v = LAPACKE_lsame( side, 'L' ) ? m :
+                             ( LAPACKE_lsame( side, 'R' ) ? n : 0 );
+        nrows_v = k;
+    } else {
+        ncols_v = 0;
+        nrows_v = 0;
+    }
     if( LAPACKE_cge_nancheck( matrix_layout, k, m, a, lda ) ) {
         return -14;
     }
     if( LAPACKE_cge_nancheck( matrix_layout, m, n, b, ldb ) ) {
         return -16;
     }
-    if( LAPACKE_cge_nancheck( matrix_layout, ldt, k, t, ldt ) ) {
+    if( LAPACKE_cge_nancheck( matrix_layout, k, k, t, ldt ) ) {
         return -12;
     }
-    if( LAPACKE_cge_nancheck( matrix_layout, ldv, k, v, ldv ) ) {
+    if( LAPACKE_cge_nancheck( matrix_layout, nrows_v, ncols_v, v, ldv ) ) {
         return -10;
     }
 #endif

lapack-netlib/LAPACKE/src/lapacke_cuncsd2by1.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function cuncsd2by1
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -37,7 +37,7 @@ lapack_int LAPACKE_cuncsd2by1( int matrix_layout, char jobu1, char jobu2,
                            char jobv1t, lapack_int m, lapack_int p, lapack_int q,
                            lapack_complex_float* x11, lapack_int ldx11,
                            lapack_complex_float* x21, lapack_int ldx21,
-                           lapack_complex_float* theta, lapack_complex_float* u1,
+                           float* theta, lapack_complex_float* u1,
                            lapack_int ldu1, lapack_complex_float* u2,
                            lapack_int ldu2, lapack_complex_float* v1t, lapack_int ldv1t )
 {

lapack-netlib/LAPACKE/src/lapacke_cuncsd2by1_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function cuncsd2by1
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -37,7 +37,7 @@ lapack_int LAPACKE_cuncsd2by1_work( int matrix_layout, char jobu1, char jobu2,
                                 char jobv1t, lapack_int m, lapack_int p,
                                 lapack_int q, lapack_complex_float* x11, lapack_int ldx11,
                                 lapack_complex_float* x21, lapack_int ldx21,
-                                lapack_complex_float* theta, lapack_complex_float* u1,
+                                float* theta, lapack_complex_float* u1,
                                 lapack_int ldu1, lapack_complex_float* u2, 
                                 lapack_int ldu2, lapack_complex_float* v1t,
                                 lapack_int ldv1t, lapack_complex_float* work, 
@@ -99,8 +99,8 @@ lapack_int LAPACKE_cuncsd2by1_work( int matrix_layout, char jobu1, char jobu2,
         /* Query optimal working array(s) size if requested */
         if( lrwork == -1 || lwork == -1 ) {
             LAPACK_cuncsd2by1( &jobu1, &jobu2, &jobv1t, &m, &p,
-                       &q, x11, &ldx11, x21, &ldx21, 
-                       theta, u1, &ldu1, u2, &ldu2, v1t, &ldv1t, 
+                       &q, x11, &ldx11_t, x21, &ldx21_t,
+                       theta, u1, &ldu1_t, u2, &ldu2_t, v1t, &ldv1t_t,
                        work, &lwork, rwork, &lrwork, iwork, &info );
             return (info < 0) ? (info - 1) : info;
         }
@@ -146,8 +146,8 @@ lapack_int LAPACKE_cuncsd2by1_work( int matrix_layout, char jobu1, char jobu2,
                            ldx21_t );
         /* Call LAPACK function and adjust info */
         LAPACK_cuncsd2by1( &jobu1, &jobu2, &jobv1t, &m, &p,
-                       &q, x11, &ldx11, x21, &ldx21, 
-                       theta, u1, &ldu1, u2, &ldu2, v1t, &ldv1t, 
+                       &q, x11_t, &ldx11_t, x21_t, &ldx21_t,
+                       theta, u1_t, &ldu1_t, u2_t, &ldu2_t, v1t_t, &ldv1t_t,
                        work, &lwork, rwork, &lrwork, iwork, &info );
         if( info < 0 ) {
             info = info - 1;

lapack-netlib/LAPACKE/src/lapacke_cunmbr_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function cunmbr
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -73,9 +73,11 @@ lapack_int LAPACKE_cunmbr_work( int matrix_layout, char vect, char side,
             return (info < 0) ? (info - 1) : info;
         }
         /* Allocate memory for temporary array(s) */
-        a_t = (lapack_complex_float*)
-            LAPACKE_malloc( sizeof(lapack_complex_float) *
-                            lda_t * MAX(1,MIN(nq,k)) );
+        if( LAPACKE_lsame( vect, 'q' ) ) {
+          a_t = (lapack_complex_float*)LAPACKE_malloc( sizeof(lapack_complex_float) * lda_t * MAX(1,k) );
+        } else {
+          a_t = (lapack_complex_float*)LAPACKE_malloc( sizeof(lapack_complex_float) * lda_t * MAX(1,nq) );
+        }
         if( a_t == NULL ) {
             info = LAPACK_TRANSPOSE_MEMORY_ERROR;
             goto exit_level_0;

lapack-netlib/LAPACKE/src/lapacke_cunmlq_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function cunmlq
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -42,6 +42,9 @@ lapack_int LAPACKE_cunmlq_work( int matrix_layout, char side, char trans,
 {
     lapack_int info = 0;
     lapack_int r;
+    lapack_int lda_t, ldc_t;
+	lapack_complex_float* a_t = NULL;
+    lapack_complex_float* c_t = NULL;
     if( matrix_layout == LAPACK_COL_MAJOR ) {
         /* Call LAPACK function and adjust info */
         LAPACK_cunmlq( &side, &trans, &m, &n, &k, a, &lda, tau, c, &ldc, work,
@@ -51,10 +54,8 @@ lapack_int LAPACKE_cunmlq_work( int matrix_layout, char side, char trans,
         }
     } else if( matrix_layout == LAPACK_ROW_MAJOR ) {
         r = LAPACKE_lsame( side, 'l' ) ? m : n;
-        lapack_int lda_t = MAX(1,k);
-        lapack_int ldc_t = MAX(1,m);
-        lapack_complex_float* a_t = NULL;
-        lapack_complex_float* c_t = NULL;
+        lda_t = MAX(1,k);
+        ldc_t = MAX(1,m);
         /* Check leading dimension(s) */
         if( lda < r ) {
             info = -8;
@@ -73,8 +74,13 @@ lapack_int LAPACKE_cunmlq_work( int matrix_layout, char side, char trans,
             return (info < 0) ? (info - 1) : info;
         }
         /* Allocate memory for temporary array(s) */
-        a_t = (lapack_complex_float*)
-            LAPACKE_malloc( sizeof(lapack_complex_float) * lda_t * MAX(1,m) );
+        if( LAPACKE_lsame( side, 'l' ) ) {
+            a_t = (lapack_complex_float*)
+                LAPACKE_malloc( sizeof(lapack_complex_float) * lda_t * MAX(1,m) );
+        } else {
+            a_t = (lapack_complex_float*)
+                LAPACKE_malloc( sizeof(lapack_complex_float) * lda_t * MAX(1,n) );
+        }
         if( a_t == NULL ) {
             info = LAPACK_TRANSPOSE_MEMORY_ERROR;
             goto exit_level_0;

lapack-netlib/LAPACKE/src/lapacke_dbdsvdx.c

@@ -28,20 +28,20 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function dbdsvdx
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_dbdsvdx( int matrix_layout, char uplo, char jobz, char range,
                            lapack_int n, double* d, double* e,
-                           lapack_int vl, lapack_int vu,
-                           lapack_int il, lapack_int iu, lapack_int ns,
+                           double vl, double vu,
+                           lapack_int il, lapack_int iu, lapack_int* ns,
                            double* s, double* z, lapack_int ldz,
                            lapack_int* superb )
 {
     lapack_int info = 0;
-    lapack_int lwork = 14*n;
+    lapack_int lwork = MAX(14*n,1);
     double* work = NULL;
     lapack_int* iwork = NULL;
     lapack_int i;
@@ -54,7 +54,7 @@ lapack_int LAPACKE_dbdsvdx( int matrix_layout, char uplo, char jobz, char range,
     if( LAPACKE_d_nancheck( n, d, 1 ) ) {
         return -6;
     }
-    if( LAPACKE_d_nancheck( n, e, 1 ) ) {
+    if( LAPACKE_d_nancheck( n - 1, e, 1 ) ) {
         return -7;
     }
 #endif
@@ -64,14 +64,14 @@ lapack_int LAPACKE_dbdsvdx( int matrix_layout, char uplo, char jobz, char range,
         info = LAPACK_WORK_MEMORY_ERROR;
         goto exit_level_0;
     }
-    iwork = (lapack_int*)LAPACKE_malloc( sizeof(lapack_int) * (12*n) );
+    iwork = (lapack_int*)LAPACKE_malloc( sizeof(lapack_int) * MAX(12*n,1) );
     if( iwork == NULL ) {
         info = LAPACK_WORK_MEMORY_ERROR;
         goto exit_level_1;
     }
     /* Call middle-level interface */
     info = LAPACKE_dbdsvdx_work( matrix_layout, uplo, jobz,  range,
-    							 n, d, e, vl, vu, il, iu, ns, s, z,
+    				n, d, e, vl, vu, il, iu, ns, s, z,
                                 ldz, work, iwork);
     /* Backup significant data from working array(s) */
     for( i=0; i<12*n-1; i++ ) {

lapack-netlib/LAPACKE/src/lapacke_dbdsvdx_work.c

@@ -28,15 +28,15 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function dbdsvdx
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_dbdsvdx_work( int matrix_layout, char uplo, char jobz, char range,
                            		lapack_int n, double* d, double* e,
-                           		lapack_int vl, lapack_int vu,
-                           		lapack_int il, lapack_int iu, lapack_int ns,
+                           		double vl, double vu,
+                           		lapack_int il, lapack_int iu, lapack_int* ns,
                            		double* s, double* z, lapack_int ldz,	
                                 double* work, lapack_int* iwork )
 {
@@ -44,25 +44,27 @@ lapack_int LAPACKE_dbdsvdx_work( int matrix_layout, char uplo, char jobz, char r
     if( matrix_layout == LAPACK_COL_MAJOR ) {
         /* Call LAPACK function and adjust info */
         LAPACK_dbdsvdx( &uplo, &jobz,  &range, &n, d, e, &vl, &vu,
-            			&il, &iu, &ns, s, z, &ldz,
+            			&il, &iu, ns, s, z, &ldz,
                         work, iwork, &info );
         if( info < 0 ) {
             info = info - 1;
         }
     } else if( matrix_layout == LAPACK_ROW_MAJOR ) {
-        lapack_int nrows_z = ( LAPACKE_lsame( jobz, 'v' ) ) ? MAX(2, 2*n) : 1;
+        lapack_int nrows_z = ( LAPACKE_lsame( jobz, 'v' ) ) ? 2*n : 0;
+        lapack_int ncols_z = ( LAPACKE_lsame( jobz, 'v' ) ) ?
+            ( LAPACKE_lsame( range, 'i' ) ? MAX(0,iu - il + 1) : n + 1 ) : 0;
         lapack_int ldz_t = MAX(1,nrows_z);
         double* z_t = NULL;
         /* Check leading dimension(s) */
-        if( ldz < nrows_z ) {
+        if( ldz < ncols_z ) {
             info = -3;
             LAPACKE_xerbla( "LAPACKE_dbdsvdx_work", info );
             return info;
         }
         /* Allocate memory for temporary array(s) */
-        if( LAPACKE_lsame( jobz, 'n' ) ) {
+        if( LAPACKE_lsame( jobz, 'v' ) ) {
            z_t = (double*)
-              LAPACKE_malloc( sizeof(double) * ldz_t * 2*n );
+              LAPACKE_malloc( sizeof(double) * ldz_t * MAX(2*n,1) );
            if( z_t == NULL ) {
               info = LAPACK_TRANSPOSE_MEMORY_ERROR;
               goto exit_level_0;
@@ -70,17 +72,17 @@ lapack_int LAPACKE_dbdsvdx_work( int matrix_layout, char uplo, char jobz, char r
         }
         /* Call LAPACK function and adjust info */
         LAPACK_dbdsvdx( &uplo, &jobz, &range, &n, d, e, &vl, &vu,
-            			&il, &iu, &ns, s, z_t, &ldz_t, work, 
+            			&il, &iu, ns, s, z_t, &ldz_t, work,
             			iwork, &info );
         if( info < 0 ) {
             info = info - 1;
         }
         /* Transpose output matrices */
-        if( LAPACKE_lsame( jobz, 'n' ) ) {
-            LAPACKE_dge_trans( LAPACK_COL_MAJOR, nrows_z, nrows_z, z_t, ldz_t, z, ldz);
+        if( LAPACKE_lsame( jobz, 'v' ) ) {
+            LAPACKE_dge_trans( LAPACK_COL_MAJOR, nrows_z, ncols_z, z_t, ldz_t, z, ldz);
         }
         /* Release memory and exit */
-        if( LAPACKE_lsame( jobz, 'n' ) ) {
+        if( LAPACKE_lsame( jobz, 'v' ) ) {
             LAPACKE_free( z_t );
         }
 exit_level_0:

lapack-netlib/LAPACKE/src/lapacke_dgejsv.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function dgejsv
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -70,7 +70,7 @@ lapack_int LAPACKE_dgejsv( int matrix_layout, char joba, char jobu, char jobv,
                        ( LAPACKE_lsame( jobv, 'v' ) ||
                        LAPACKE_lsame( jobv, 'j' ) ) &&
                        LAPACKE_lsame( jobv, 'j' ) ? MAX(7,m+3*n+n*n) :
-                       1) ) ) ) ) );
+                       7) ) ) ) ) );
     lapack_int* iwork = NULL;
     double* work = NULL;
     lapack_int i;
@@ -86,25 +86,25 @@ lapack_int LAPACKE_dgejsv( int matrix_layout, char joba, char jobu, char jobv,
     if( LAPACKE_dge_nancheck( matrix_layout, m, n, a, lda ) ) {
         return -10;
     }
-    if( LAPACKE_lsame( jobu, 'f' ) || LAPACKE_lsame( jobu, 'u' ) ||
-        LAPACKE_lsame( jobu, 'w' ) ) {
-        if( LAPACKE_dge_nancheck( matrix_layout, nu, n, u, ldu ) ) {
-            return -13;
-        }
-    }
-    if( LAPACKE_lsame( jobv, 'j' ) || LAPACKE_lsame( jobv, 'v' ) ||
-        LAPACKE_lsame( jobv, 'w' ) ) {
-        if( LAPACKE_dge_nancheck( matrix_layout, nv, n, v, ldv ) ) {
-            return -15;
-        }
-    }
 #endif
     /* Allocate memory for working array(s) */
-    iwork = (lapack_int*)LAPACKE_malloc( sizeof(lapack_int) * MAX(1,m+3*n) );
+    iwork = (lapack_int*)LAPACKE_malloc( sizeof(lapack_int) * MAX(3,m+3*n) );
     if( iwork == NULL ) {
         info = LAPACK_WORK_MEMORY_ERROR;
         goto exit_level_0;
     }
+    lwork = MAX3( lwork, 7, 2*m+n );
+    { /* FIXUP LWORK */
+        int want_u = LAPACKE_lsame( jobu, 'u' ) || LAPACKE_lsame( jobu, 'f' );
+        int want_v = LAPACKE_lsame( jobv, 'v' ) || LAPACKE_lsame( jobv, 'j' );
+        int want_sce = LAPACKE_lsame( joba, 'e' ) || LAPACKE_lsame( joba, 'g' );
+        if( !want_u && !want_v && !want_sce )  lwork = MAX( lwork, 4*n+1 ); // 1.1
+        if( !want_u && !want_v && want_sce )   lwork = MAX( lwork, n*n+4*n ); // 1.2
+        if( !want_u && want_v ) lwork = MAX( lwork, 4*n+1 ); // 2
+        if( want_u && !want_v ) lwork = MAX( lwork, 4*n+1 ); // 3
+        if( want_u && LAPACKE_lsame( jobv, 'v' ) ) lwork = MAX( lwork, 6*n+2*n*n ); // 4.1
+        if( want_u && LAPACKE_lsame( jobv, 'j' ) ) lwork = MAX3( lwork, 4*n+n*n, 2*n+n*n+6 ); // 4.2
+    }
     work = (double*)LAPACKE_malloc( sizeof(double) * lwork );
     if( work == NULL ) {
         info = LAPACK_WORK_MEMORY_ERROR;

lapack-netlib/LAPACKE/src/lapacke_dgejsv_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function dgejsv
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -53,6 +53,8 @@ lapack_int LAPACKE_dgejsv_work( int matrix_layout, char joba, char jobu,
     } else if( matrix_layout == LAPACK_ROW_MAJOR ) {
         lapack_int nu = LAPACKE_lsame( jobu, 'n' ) ? 1 : m;
         lapack_int nv = LAPACKE_lsame( jobv, 'n' ) ? 1 : n;
+        lapack_int ncols_u = LAPACKE_lsame( jobu, 'n' ) ? 1 :
+            LAPACKE_lsame( jobu, 'f' ) ? m : n;
         lapack_int lda_t = MAX(1,m);
         lapack_int ldu_t = MAX(1,nu);
         lapack_int ldv_t = MAX(1,nv);
@@ -65,7 +67,7 @@ lapack_int LAPACKE_dgejsv_work( int matrix_layout, char joba, char jobu,
             LAPACKE_xerbla( "LAPACKE_dgejsv_work", info );
             return info;
         }
-        if( ldu < n ) {
+        if( ldu < ncols_u ) {
             info = -14;
             LAPACKE_xerbla( "LAPACKE_dgejsv_work", info );
             return info;
@@ -83,7 +85,7 @@ lapack_int LAPACKE_dgejsv_work( int matrix_layout, char joba, char jobu,
         }
         if( LAPACKE_lsame( jobu, 'f' ) || LAPACKE_lsame( jobu, 'u' ) ||
             LAPACKE_lsame( jobu, 'w' ) ) {
-            u_t = (double*)LAPACKE_malloc( sizeof(double) * ldu_t * MAX(1,n) );
+            u_t = (double*)LAPACKE_malloc( sizeof(double) * ldu_t * MAX(1,ncols_u) );
             if( u_t == NULL ) {
                 info = LAPACK_TRANSPOSE_MEMORY_ERROR;
                 goto exit_level_1;
@@ -99,14 +101,6 @@ lapack_int LAPACKE_dgejsv_work( int matrix_layout, char joba, char jobu,
         }
         /* Transpose input matrices */
         LAPACKE_dge_trans( matrix_layout, m, n, a, lda, a_t, lda_t );
-        if( LAPACKE_lsame( jobu, 'f' ) || LAPACKE_lsame( jobu, 'u' ) ||
-            LAPACKE_lsame( jobu, 'w' ) ) {
-            LAPACKE_dge_trans( matrix_layout, nu, n, u, ldu, u_t, ldu_t );
-        }
-        if( LAPACKE_lsame( jobv, 'j' ) || LAPACKE_lsame( jobv, 'v' ) ||
-            LAPACKE_lsame( jobv, 'w' ) ) {
-            LAPACKE_dge_trans( matrix_layout, nv, n, v, ldv, v_t, ldv_t );
-        }
         /* Call LAPACK function and adjust info */
         LAPACK_dgejsv( &joba, &jobu, &jobv, &jobr, &jobt, &jobp, &m, &n, a_t,
                        &lda_t, sva, u_t, &ldu_t, v_t, &ldv_t, work, &lwork,
@@ -117,7 +111,7 @@ lapack_int LAPACKE_dgejsv_work( int matrix_layout, char joba, char jobu,
         /* Transpose output matrices */
         if( LAPACKE_lsame( jobu, 'f' ) || LAPACKE_lsame( jobu, 'u' ) ||
             LAPACKE_lsame( jobu, 'w' ) ) {
-            LAPACKE_dge_trans( LAPACK_COL_MAJOR, nu, n, u_t, ldu_t, u, ldu );
+            LAPACKE_dge_trans( LAPACK_COL_MAJOR, nu, ncols_u, u_t, ldu_t, u, ldu );
         }
         if( LAPACKE_lsame( jobv, 'j' ) || LAPACKE_lsame( jobv, 'v' ) ||
             LAPACKE_lsame( jobv, 'w' ) ) {

lapack-netlib/LAPACKE/src/lapacke_dgemqrt.c

@@ -28,7 +28,7 @@
 ******************************************************************************
 * Contents: Native high-level C interface to LAPACK function dgemqrt
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -39,6 +39,7 @@ lapack_int LAPACKE_dgemqrt( int matrix_layout, char side, char trans,
                             const double* t, lapack_int ldt, double* c,
                             lapack_int ldc )
 {
+    lapack_int nrows_v;
     lapack_int info = 0;
     double* work = NULL;
     if( matrix_layout != LAPACK_COL_MAJOR && matrix_layout != LAPACK_ROW_MAJOR ) {
@@ -47,13 +48,15 @@ lapack_int LAPACKE_dgemqrt( int matrix_layout, char side, char trans,
     }
 #ifndef LAPACK_DISABLE_NAN_CHECK
     /* Optionally check input matrices for NaNs */
+    nrows_v = LAPACKE_lsame( side, 'L' ) ? m :
+                         ( LAPACKE_lsame( side, 'R' ) ? n : 0 );
     if( LAPACKE_dge_nancheck( matrix_layout, m, n, c, ldc ) ) {
         return -12;
     }
-    if( LAPACKE_dge_nancheck( matrix_layout, ldt, nb, t, ldt ) ) {
+    if( LAPACKE_dge_nancheck( matrix_layout, nb, k, t, ldt ) ) {
         return -10;
     }
-    if( LAPACKE_dge_nancheck( matrix_layout, ldv, k, v, ldv ) ) {
+    if( LAPACKE_dge_nancheck( matrix_layout, nrows_v, k, v, ldv ) ) {
         return -8;
     }
 #endif

lapack-netlib/LAPACKE/src/lapacke_dgesvdx.c

@@ -28,15 +28,15 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function dgesvdx
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_dgesvdx( int matrix_layout, char jobu, char jobvt, char range,
                            lapack_int m, lapack_int n, double* a,
-                           lapack_int lda, lapack_int vl, lapack_int vu,
-                           lapack_int il, lapack_int iu, lapack_int ns,
+                           lapack_int lda, double vl, double vu,
+                           lapack_int il, lapack_int iu, lapack_int* ns,
                            double* s, double* u, lapack_int ldu,
                            double* vt, lapack_int ldvt,
                            lapack_int* superb )
@@ -71,7 +71,7 @@ lapack_int LAPACKE_dgesvdx( int matrix_layout, char jobu, char jobvt, char range
         info = LAPACK_WORK_MEMORY_ERROR;
         goto exit_level_0;
     }
-    iwork = (lapack_int*)LAPACKE_malloc( sizeof(lapack_int) * (12*MIN(m,n)) );
+    iwork = (lapack_int*)LAPACKE_malloc( sizeof(lapack_int) * MAX(1,(12*MIN(m,n))) );
     if( iwork == NULL ) {
         info = LAPACK_WORK_MEMORY_ERROR;
         goto exit_level_1;

lapack-netlib/LAPACKE/src/lapacke_dgesvdx_work.c

@@ -28,15 +28,15 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function dgesvdx
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_dgesvdx_work( int matrix_layout, char jobu, char jobvt, char range,
                            		lapack_int m, lapack_int n, double* a,
-                          		lapack_int lda, lapack_int vl, lapack_int vu,
-                           		lapack_int il, lapack_int iu, lapack_int ns,
+                          		lapack_int lda, double vl, double vu,
+                           		lapack_int il, lapack_int iu, lapack_int* ns,
                            		double* s, double* u, lapack_int ldu,
                            		double* vt, lapack_int ldvt,	
                                 double* work, lapack_int lwork, lapack_int* iwork )
@@ -45,21 +45,23 @@ lapack_int LAPACKE_dgesvdx_work( int matrix_layout, char jobu, char jobvt, char
     if( matrix_layout == LAPACK_COL_MAJOR ) {
         /* Call LAPACK function and adjust info */
         LAPACK_dgesvdx( &jobu, &jobvt,  &range, &m, &n, a, &lda, &vl, &vu,
-            			&il, &iu, &ns, s, u, &ldu, vt, &ldvt,
+            			&il, &iu, ns, s, u, &ldu, vt, &ldvt,
                         work, &lwork, iwork, &info );
         if( info < 0 ) {
             info = info - 1;
         }
     } else if( matrix_layout == LAPACK_ROW_MAJOR ) {
-        lapack_int nrows_u = ( LAPACKE_lsame( jobu, 'a' ) ||
-                             LAPACKE_lsame( jobu, 's' ) ) ? m : 1;
-        lapack_int ncols_u = LAPACKE_lsame( jobu, 'a' ) ? m :
-                             ( LAPACKE_lsame( jobu, 's' ) ? MIN(m,n) : 1);
-        lapack_int nrows_vt = LAPACKE_lsame( jobvt, 'a' ) ? n :
-                              ( LAPACKE_lsame( jobvt, 's' ) ? MIN(m,n) : 1);
+        lapack_int nrows_u = LAPACKE_lsame( jobu, 'v' ) ? m : 0;
+        lapack_int ncols_u = LAPACKE_lsame( jobu, 'v' ) ?
+                             ( LAPACKE_lsame( range, 'i' ) ? MAX(iu - il + 1, 0) : MIN(m,n)) : 0;
+        lapack_int nrows_vt = LAPACKE_lsame( jobvt, 'v' ) ?
+                             ( LAPACKE_lsame( range, 'i' ) ? MAX(iu - il + 1, 0) : MIN(m,n)) : 0;
+        lapack_int ncols_vt = LAPACKE_lsame( jobvt, 'v' ) ? n : 0;
+
         lapack_int lda_t = MAX(1,m);
         lapack_int ldu_t = MAX(1,nrows_u);
         lapack_int ldvt_t = MAX(1,nrows_vt);
+
         double* a_t = NULL;
         double* u_t = NULL;
         double* vt_t = NULL;
@@ -74,7 +76,7 @@ lapack_int LAPACKE_dgesvdx_work( int matrix_layout, char jobu, char jobvt, char
             LAPACKE_xerbla( "LAPACKE_dgesvdx_work", info );
             return info;
         }
-        if( ldvt < n ) {
+        if( ldvt < ncols_vt ) {
             info = -18;
             LAPACKE_xerbla( "LAPACKE_dgesvdx_work", info );
             return info;
@@ -82,7 +84,7 @@ lapack_int LAPACKE_dgesvdx_work( int matrix_layout, char jobu, char jobvt, char
         /* Query optimal working array(s) size if requested */
         if( lwork == -1 ) {
             LAPACK_dgesvdx( &jobu, &jobvt, &range, &m, &n, a, &lda_t, &vl, &vu,
-            				&il, &iu, &ns, s, u, &ldu_t, vt,
+            				&il, &iu, ns, s, u, &ldu_t, vt,
                             &ldvt_t, work, &lwork, iwork, &info );
             return (info < 0) ? (info - 1) : info;
         }
@@ -92,7 +94,7 @@ lapack_int LAPACKE_dgesvdx_work( int matrix_layout, char jobu, char jobvt, char
             info = LAPACK_TRANSPOSE_MEMORY_ERROR;
             goto exit_level_0;
         }
-        if( LAPACKE_lsame( jobu, 'a' ) || LAPACKE_lsame( jobu, 's' ) ) {
+        if( LAPACKE_lsame( jobu, 'v' ) ) {
             u_t = (double*)
                 LAPACKE_malloc( sizeof(double) * ldu_t * MAX(1,ncols_u) );
             if( u_t == NULL ) {
@@ -100,7 +102,7 @@ lapack_int LAPACKE_dgesvdx_work( int matrix_layout, char jobu, char jobvt, char
                 goto exit_level_1;
             }
         }
-        if( LAPACKE_lsame( jobvt, 'a' ) || LAPACKE_lsame( jobvt, 's' ) ) {
+        if( LAPACKE_lsame( jobvt, 'v' ) ) {
             vt_t = (double*)
                 LAPACKE_malloc( sizeof(double) * ldvt_t * MAX(1,n) );
             if( vt_t == NULL ) {
@@ -111,28 +113,28 @@ lapack_int LAPACKE_dgesvdx_work( int matrix_layout, char jobu, char jobvt, char
         /* Transpose input matrices */
         LAPACKE_dge_trans( matrix_layout, m, n, a, lda, a_t, lda_t );
         /* Call LAPACK function and adjust info */
-        LAPACK_dgesvdx( &jobu, &jobvt, &range, &m, &n, a, &lda_t, &vl, &vu,
-            				&il, &iu, &ns, s, u, &ldu_t, vt,
+        LAPACK_dgesvdx( &jobu, &jobvt, &range, &m, &n, a_t, &lda_t, &vl, &vu,
+            				&il, &iu, ns, s, u_t, &ldu_t, vt_t,
                             &ldvt_t, work, &lwork, iwork, &info );
         if( info < 0 ) {
             info = info - 1;
         }
         /* Transpose output matrices */
         LAPACKE_dge_trans( LAPACK_COL_MAJOR, m, n, a_t, lda_t, a, lda );
-        if( LAPACKE_lsame( jobu, 'a' ) || LAPACKE_lsame( jobu, 's' ) ) {
+        if( LAPACKE_lsame( jobu, 'v' ) ) {
             LAPACKE_dge_trans( LAPACK_COL_MAJOR, nrows_u, ncols_u, u_t, ldu_t,
                                u, ldu );
         }
-        if( LAPACKE_lsame( jobvt, 'a' ) || LAPACKE_lsame( jobvt, 's' ) ) {
+        if( LAPACKE_lsame( jobvt, 'v' ) ) {
             LAPACKE_dge_trans( LAPACK_COL_MAJOR, nrows_vt, n, vt_t, ldvt_t, vt,
                                ldvt );
         }
         /* Release memory and exit */
-        if( LAPACKE_lsame( jobvt, 'a' ) || LAPACKE_lsame( jobvt, 's' ) ) {
+        if( LAPACKE_lsame( jobvt, 'v' ) ) {
             LAPACKE_free( vt_t );
         }
 exit_level_2:
-        if( LAPACKE_lsame( jobu, 'a' ) || LAPACKE_lsame( jobu, 's' ) ) {
+        if( LAPACKE_lsame( jobu, 'v' ) ) {
             LAPACKE_free( u_t );
         }
 exit_level_1:

lapack-netlib/LAPACKE/src/lapacke_dgesvj.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function dgesvj
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -49,8 +49,8 @@ lapack_int LAPACKE_dgesvj( int matrix_layout, char joba, char jobu, char jobv,
     }
 #ifndef LAPACK_DISABLE_NAN_CHECK
     /* Optionally check input matrices for NaNs */
-    nrows_v = LAPACKE_lsame( jobv, 'v' ) ? n :
-                         ( LAPACKE_lsame( jobv, 'a' ) ? mv : 1);
+    nrows_v = LAPACKE_lsame( jobv, 'v' ) ? MAX(0,n) :
+            ( LAPACKE_lsame( jobv, 'a' ) ? MAX(0,mv) : 0);
     if( LAPACKE_dge_nancheck( matrix_layout, m, n, a, lda ) ) {
         return -7;
     }

lapack-netlib/LAPACKE/src/lapacke_dgesvj_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function dgesvj
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -48,8 +48,8 @@ lapack_int LAPACKE_dgesvj_work( int matrix_layout, char joba, char jobu,
             info = info - 1;
         }
     } else if( matrix_layout == LAPACK_ROW_MAJOR ) {
-        lapack_int nrows_v = LAPACKE_lsame( jobv, 'v' ) ? n :
-                             ( LAPACKE_lsame( jobv, 'a' ) ? mv : 1);
+        lapack_int nrows_v = LAPACKE_lsame( jobv, 'v' ) ? MAX(0,n) :
+                           ( LAPACKE_lsame( jobv, 'a' ) ? MAX(0,mv) : 0);
         lapack_int lda_t = MAX(1,m);
         lapack_int ldv_t = MAX(1,nrows_v);
         double* a_t = NULL;

lapack-netlib/LAPACKE/src/lapacke_dggsvd3_work.c

@@ -91,9 +91,9 @@ lapack_int LAPACKE_dggsvd3_work( int matrix_layout, char jobu, char jobv,
         }
         /* Query optimal working array(s) size if requested */
         if( lwork == -1 ) {
-          LAPACK_dggsvd3( &jobu, &jobv, &jobq, &m, &n, &p, k, l, a_t, &lda_t,
-                          b_t, &ldb_t, alpha, beta, u_t, &ldu_t, v_t, &ldv_t,
-                          q_t, &ldq_t, work, &lwork, iwork, &info );
+          LAPACK_dggsvd3( &jobu, &jobv, &jobq, &m, &n, &p, k, l, a, &lda_t,
+                          b, &ldb_t, alpha, beta, u, &ldu_t, v, &ldv_t,
+                          q, &ldq_t, work, &lwork, iwork, &info );
           return (info < 0) ? (info - 1) : info;
         }
         /* Allocate memory for temporary array(s) */

lapack-netlib/LAPACKE/src/lapacke_dggsvp3_work.c

@@ -84,15 +84,15 @@ lapack_int LAPACKE_dggsvp3_work( int matrix_layout, char jobu, char jobv,
             LAPACKE_xerbla( "LAPACKE_dggsvp3_work", info );
             return info;
         }
-        if( ldv < m ) {
+        if( ldv < p ) {
             info = -19;
             LAPACKE_xerbla( "LAPACKE_dggsvp3_work", info );
             return info;
         }
         if( lwork == -1 ) {
-          LAPACK_dggsvp3( &jobu, &jobv, &jobq, &m, &p, &n, a_t, &lda_t, b_t,
-                          &ldb_t, &tola, &tolb, k, l, u_t, &ldu_t, v_t,
-                          &ldv_t, q_t, &ldq_t, iwork, tau, work, &lwork,
+          LAPACK_dggsvp3( &jobu, &jobv, &jobq, &m, &p, &n, a, &lda_t, b,
+                          &ldb_t, &tola, &tolb, k, l, u, &ldu_t, v,
+                          &ldv_t, q, &ldq_t, iwork, tau, work, &lwork,
                           &info );
           return (info < 0) ? (info - 1) : info;
         }
@@ -115,7 +115,7 @@ lapack_int LAPACKE_dggsvp3_work( int matrix_layout, char jobu, char jobv,
             }
         }
         if( LAPACKE_lsame( jobv, 'v' ) ) {
-            v_t = (double*)LAPACKE_malloc( sizeof(double) * ldv_t * MAX(1,m) );
+            v_t = (double*)LAPACKE_malloc( sizeof(double) * ldv_t * MAX(1,p) );
             if( v_t == NULL ) {
                 info = LAPACK_TRANSPOSE_MEMORY_ERROR;
                 goto exit_level_3;
@@ -145,7 +145,7 @@ lapack_int LAPACKE_dggsvp3_work( int matrix_layout, char jobu, char jobv,
             LAPACKE_dge_trans( LAPACK_COL_MAJOR, m, m, u_t, ldu_t, u, ldu );
         }
         if( LAPACKE_lsame( jobv, 'v' ) ) {
-            LAPACKE_dge_trans( LAPACK_COL_MAJOR, p, m, v_t, ldv_t, v, ldv );
+            LAPACKE_dge_trans( LAPACK_COL_MAJOR, p, p, v_t, ldv_t, v, ldv );
         }
         if( LAPACKE_lsame( jobq, 'q' ) ) {
             LAPACKE_dge_trans( LAPACK_COL_MAJOR, n, n, q_t, ldq_t, q, ldq );

lapack-netlib/LAPACKE/src/lapacke_dlantr.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function dlantr
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"

lapack-netlib/LAPACKE/src/lapacke_dlantr_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function dlantr
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"

lapack-netlib/LAPACKE/src/lapacke_dlarfb_work.c

@@ -119,7 +119,7 @@ lapack_int LAPACKE_dlarfb_work( int matrix_layout, char side, char trans,
             LAPACKE_dge_trans( matrix_layout, nrows_v, ncols_v-k, &v[k], ldv,
                                &v_t[k*ldv_t], ldv_t );
         } else if( LAPACKE_lsame( storev, 'r' ) &&
-                   LAPACKE_lsame( direct, 'b' ) ) {
+                   LAPACKE_lsame( direct, 'f' ) ) {
             if( k > ncols_v ) {
                 LAPACKE_xerbla( "LAPACKE_dlarfb_work", -8 );
                 return -8;

lapack-netlib/LAPACKE/src/lapacke_dlascl.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function dlaswp
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -46,50 +46,64 @@ lapack_int LAPACKE_dlascl( int matrix_layout, char type, lapack_int kl,
     /* Optionally check input matrices for NaNs */
     switch (type) {
     case 'G':
-       if( LAPACKE_dge_nancheck( matrix_layout, lda, n, a, lda ) ) {
-           return -9;
-           }
+        if( LAPACKE_dge_nancheck( matrix_layout, m, n, a, lda ) ) {
+            return -9;
+        }
         break;
     case 'L':
-       // TYPE = 'L' - lower triangular matrix.
-       if( LAPACKE_dtr_nancheck( matrix_layout, 'L', 'N', n, a, lda ) ) {
-           return -9;
-          }
+        // TYPE = 'L' - lower triangle of general matrix
+        if( matrix_layout == LAPACK_COL_MAJOR &&
+            LAPACKE_dgb_nancheck( matrix_layout, m, n, m-1, 0, a, lda+1 ) ) {
+            return -9;
+        }
+        if( matrix_layout == LAPACK_ROW_MAJOR &&
+            LAPACKE_dgb_nancheck( LAPACK_COL_MAJOR, n, m, 0, m-1, a-m+1, lda+1 ) ) {
+            return -9;
+        }
         break;
     case 'U':
-       // TYPE = 'U' - upper triangular matrix
-       if( LAPACKE_dtr_nancheck( matrix_layout, 'U', 'N', n, a, lda ) ) {
-           return -9;
-           } 
+        // TYPE = 'U' - upper triangle of general matrix
+        if( matrix_layout == LAPACK_COL_MAJOR &&
+            LAPACKE_dgb_nancheck( matrix_layout, m, n, 0, n-1, a-n+1, lda+1 ) ) {
+            return -9;
+        }
+        if( matrix_layout == LAPACK_ROW_MAJOR &&
+            LAPACKE_dgb_nancheck( LAPACK_COL_MAJOR, n, m, n-1, 0, a, lda+1 ) ) {
+            return -9;
+        }
         break;
     case 'H':
-       // TYPE = 'H' - upper Hessenberg matrix   
-       if( LAPACKE_dhs_nancheck( matrix_layout, n, a, lda ) ) {
-           return -9;
-           }    
-        break;
+        // TYPE = 'H' - part of upper Hessenberg matrix in general matrix
+        if( matrix_layout == LAPACK_COL_MAJOR &&
+            LAPACKE_dgb_nancheck( matrix_layout, m, n, 1, n-1, a-n+1, lda+1 ) ) {
+            return -9;
+        }
+        if( matrix_layout == LAPACK_ROW_MAJOR &&
+            LAPACKE_dgb_nancheck( LAPACK_COL_MAJOR, n, m, n-1, 1, a-1, lda+1 ) ) {
+            return -9;
+        }
     case 'B':
-       // TYPE = 'B' - A is a symmetric band matrix with lower bandwidth KL
-       //             and upper bandwidth KU and with the only the lower
-       //             half stored.   
-       if( LAPACKE_dsb_nancheck( matrix_layout, 'L', n, kl, a, lda ) ) {
-           return -9;
-           }
-         break;
-   case 'Q':
-       // TYPE = 'Q' - A is a symmetric band matrix with lower bandwidth KL
-       //             and upper bandwidth KU and with the only the upper
-       //             half stored.   
-       if( LAPACKE_dsb_nancheck( matrix_layout, 'U', n, ku, a, lda ) ) {
-           return -9;
-           }
+        // TYPE = 'B' - lower part of symmetric band matrix (assume m==n)
+        if( LAPACKE_dsb_nancheck( matrix_layout, 'L', n, kl, a, lda ) ) {
+            return -9;
+        }
+        break;
+    case 'Q':
+        // TYPE = 'Q' - upper part of symmetric band matrix (assume m==n)
+        if( LAPACKE_dsb_nancheck( matrix_layout, 'U', n, ku, a, lda ) ) {
+            return -9;
+        }
         break;
     case 'Z':
-       // TYPE = 'Z' -  A is a band matrix with lower bandwidth KL and upper
-       //             bandwidth KU. See DGBTRF for storage details.        
-       if( LAPACKE_dgb_nancheck( matrix_layout, n, n, kl, kl+ku, a, lda ) ) {
-           return -6;
-           }
+        // TYPE = 'Z' -  band matrix laid out for ?GBTRF
+        if( matrix_layout == LAPACK_COL_MAJOR &&
+            LAPACKE_dgb_nancheck( matrix_layout, m, n, kl, ku, a+kl, lda ) ) {
+            return -9;
+        }
+        if( matrix_layout == LAPACK_ROW_MAJOR &&
+            LAPACKE_dgb_nancheck( matrix_layout, m, n, kl, ku, a+lda*kl, lda ) ) {
+            return -9;
+        }
         break;
     }
 #endif

lapack-netlib/LAPACKE/src/lapacke_dlascl_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function dlaswp
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -46,7 +46,10 @@ lapack_int LAPACKE_dlascl_work( int matrix_layout, char type, lapack_int kl,
             info = info - 1;
         }
     } else if( matrix_layout == LAPACK_ROW_MAJOR ) {
-        lapack_int lda_t = MAX(1,lda);
+        lapack_int nrows_a = LAPACKE_lsame(type, 'b') ? kl + 1 :
+                             LAPACKE_lsame(type, 'q') ? ku + 1 :
+                             LAPACKE_lsame(type, 'z') ? 2 * kl + ku + 1 : m;
+        lapack_int lda_t = MAX(1,nrows_a);
         double* a_t = NULL;
         /* Check leading dimension(s) */
         if( lda < n ) {
@@ -61,12 +64,14 @@ lapack_int LAPACKE_dlascl_work( int matrix_layout, char type, lapack_int kl,
             goto exit_level_0;
         }
         /* Transpose input matrices */
-        LAPACKE_dge_trans( matrix_layout, lda, n, a, lda, a_t, lda_t );
+        LAPACKE_dge_trans( matrix_layout, nrows_a, n, a, lda, a_t, lda_t );
         /* Call LAPACK function and adjust info */
         LAPACK_dlascl( &type, &kl, &ku, &cfrom, &cto, &m, &n, a_t, &lda_t, &info);
-        info = 0;  /* LAPACK call is ok! */
+        if( info < 0 ) {
+            info = info - 1;
+        }
         /* Transpose output matrices */
-        LAPACKE_dge_trans( LAPACK_COL_MAJOR, lda, n, a_t, lda_t, a, lda );
+        LAPACKE_dge_trans( LAPACK_COL_MAJOR, nrows_a, n, a_t, lda_t, a, lda );
         /* Release memory and exit */
         LAPACKE_free( a_t );
 exit_level_0:

lapack-netlib/LAPACKE/src/lapacke_dorcsd2by1_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function dorcsd2by1
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -97,8 +97,8 @@ lapack_int LAPACKE_dorcsd2by1_work( int matrix_layout, char jobu1, char jobu2,
         /* Query optimal working array(s) size if requested */
         if( lwork == -1 ) {
             LAPACK_dorcsd2by1( &jobu1, &jobu2, &jobv1t, &m, &p,
-                       &q, x11, &ldx11, x21, &ldx21, 
-                       theta, u1, &ldu1, u2, &ldu2, v1t, &ldv1t, 
+                       &q, x11, &ldx11_t, x21, &ldx21_t,
+                       theta, u1, &ldu1_t, u2, &ldu2_t, v1t, &ldv1t_t,
                        work, &lwork, iwork, &info );
             return (info < 0) ? (info - 1) : info;
         }
@@ -144,8 +144,8 @@ lapack_int LAPACKE_dorcsd2by1_work( int matrix_layout, char jobu1, char jobu2,
                            ldx21_t );
         /* Call LAPACK function and adjust info */
         LAPACK_dorcsd2by1( &jobu1, &jobu2, &jobv1t, &m, &p,
-                       &q, x11, &ldx11, x21, &ldx21, 
-                       theta, u1, &ldu1, u2, &ldu2, v1t, &ldv1t, 
+                       &q, x11_t, &ldx11_t, x21_t, &ldx21_t,
+                       theta, u1_t, &ldu1_t, u2_t, &ldu2_t, v1t_t, &ldv1t_t,
                        work, &lwork, iwork, &info );
         if( info < 0 ) {
             info = info - 1;

lapack-netlib/LAPACKE/src/lapacke_dormbr_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function dormbr
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"

lapack-netlib/LAPACKE/src/lapacke_dormlq_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function dormlq
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"

lapack-netlib/LAPACKE/src/lapacke_dsbtrd.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function dsbtrd
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -48,7 +48,7 @@ lapack_int LAPACKE_dsbtrd( int matrix_layout, char vect, char uplo, lapack_int n
     if( LAPACKE_dsb_nancheck( matrix_layout, uplo, n, kd, ab, ldab ) ) {
         return -6;
     }
-    if( LAPACKE_lsame( vect, 'u' ) || LAPACKE_lsame( vect, 'v' ) ) {
+    if( LAPACKE_lsame( vect, 'u' ) ) {
         if( LAPACKE_dge_nancheck( matrix_layout, n, n, q, ldq ) ) {
             return -10;
         }

lapack-netlib/LAPACKE/src/lapacke_dstedc.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function dstedc
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -55,7 +55,7 @@ lapack_int LAPACKE_dstedc( int matrix_layout, char compz, lapack_int n,
     if( LAPACKE_d_nancheck( n-1, e, 1 ) ) {
         return -5;
     }
-    if( LAPACKE_lsame( compz, 'i' ) || LAPACKE_lsame( compz, 'v' ) ) {
+    if( LAPACKE_lsame( compz, 'v' ) ) {
         if( LAPACKE_dge_nancheck( matrix_layout, n, n, z, ldz ) ) {
             return -6;
         }

lapack-netlib/LAPACKE/src/lapacke_dsteqr.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function dsteqr
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -52,7 +52,7 @@ lapack_int LAPACKE_dsteqr( int matrix_layout, char compz, lapack_int n,
     if( LAPACKE_d_nancheck( n-1, e, 1 ) ) {
         return -5;
     }
-    if( LAPACKE_lsame( compz, 'i' ) || LAPACKE_lsame( compz, 'v' ) ) {
+    if( LAPACKE_lsame( compz, 'v' ) ) {
         if( LAPACKE_dge_nancheck( matrix_layout, n, n, z, ldz ) ) {
             return -6;
         }

lapack-netlib/LAPACKE/src/lapacke_dsyconv.c

@@ -28,13 +28,13 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function dsyconv
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_dsyconv( int matrix_layout, char uplo, char way, lapack_int n,
-                            double* a, lapack_int lda, const lapack_int* ipiv, double* work )
+                            double* a, lapack_int lda, const lapack_int* ipiv, double* e )
 {
     lapack_int info = 0;
     if( matrix_layout != LAPACK_COL_MAJOR && matrix_layout != LAPACK_ROW_MAJOR ) {
@@ -43,14 +43,14 @@ lapack_int LAPACKE_dsyconv( int matrix_layout, char uplo, char way, lapack_int n
     }
 #ifndef LAPACK_DISABLE_NAN_CHECK
     /* Optionally check input matrices for NaNs */
-    if( LAPACKE_dge_nancheck( matrix_layout, lda, n, a, lda ) ) {
+    if( LAPACKE_dge_nancheck( matrix_layout, n, n, a, lda ) ) {
         return -5;
     }
 #endif
   
     /* Call middle-level interface */
     info = LAPACKE_dsyconv_work( matrix_layout, uplo, way, n, a, lda, ipiv,
-                                 work );
+                                 e );
 exit_level_0:
     if( info == LAPACK_WORK_MEMORY_ERROR ) {
         LAPACKE_xerbla( "LAPACKE_dsyconv", info );

lapack-netlib/LAPACKE/src/lapacke_dsyconv_work.c

@@ -28,19 +28,19 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function dsyconv
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_dsyconv_work( int matrix_layout, char uplo, char way,
                                  lapack_int n, double* a, lapack_int lda,
-                                 const lapack_int* ipiv, double* work )
+                                 const lapack_int* ipiv, double* e )
 {
     lapack_int info = 0;
     if( matrix_layout == LAPACK_COL_MAJOR ) {
         /* Call LAPACK function and adjust info */
-        LAPACK_dsyconv( &uplo, &way, &n, a, &lda, ipiv, work, &info );
+        LAPACK_dsyconv( &uplo, &way, &n, a, &lda, ipiv, e, &info );
         if( info < 0 ) {
             info = info - 1;
         }
@@ -62,7 +62,7 @@ lapack_int LAPACKE_dsyconv_work( int matrix_layout, char uplo, char way,
         /* Transpose input matrices */
         LAPACKE_dge_trans( matrix_layout, lda, n, a, lda, a_t, lda_t );
         /* Call LAPACK function and adjust info */
-        LAPACK_dsyconv( &uplo, &way, &n, a_t, &lda_t, ipiv, work, &info );
+        LAPACK_dsyconv( &uplo, &way, &n, a_t, &lda_t, ipiv, e, &info );
         if( info < 0 ) {
             info = info - 1;
         }

lapack-netlib/LAPACKE/src/lapacke_dsyswapr.c

@@ -28,13 +28,13 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function dsyswapr
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_dsyswapr( int matrix_layout, char uplo, lapack_int n,
-                             double* a, lapack_int i1, lapack_int i2 )
+                             double* a, lapack_int lda, lapack_int i1, lapack_int i2 )
 {
     if( matrix_layout != LAPACK_COL_MAJOR && matrix_layout != LAPACK_ROW_MAJOR ) {
         LAPACKE_xerbla( "LAPACKE_dsyswapr", -1 );
@@ -42,9 +42,9 @@ lapack_int LAPACKE_dsyswapr( int matrix_layout, char uplo, lapack_int n,
     }
 #ifndef LAPACK_DISABLE_NAN_CHECK
     /* Optionally check input matrices for NaNs */
-    if( LAPACKE_dsy_nancheck( matrix_layout, uplo, n, a, n ) ) {
+    if( LAPACKE_dsy_nancheck( matrix_layout, uplo, n, a, lda ) ) {
         return -4;
     }
 #endif
-    return LAPACKE_dsyswapr_work( matrix_layout, uplo, n, a, i1, i2 );
+    return LAPACKE_dsyswapr_work( matrix_layout, uplo, n, a, lda, i1, i2 );
 }

lapack-netlib/LAPACKE/src/lapacke_dsyswapr_work.c

@@ -28,36 +28,38 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function dsyswapr
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_dsyswapr_work( int matrix_layout, char uplo, lapack_int n,
-                                  double* a, lapack_int i1, lapack_int i2 )
+                                  double* a, lapack_int lda,
+                                  lapack_int i1, lapack_int i2 )
 {
     lapack_int info = 0;
     if( matrix_layout == LAPACK_COL_MAJOR ) {
         /* Call LAPACK function and adjust info */
-        LAPACK_dsyswapr( &uplo, &n, a, &i1, &i2 );
+        LAPACK_dsyswapr( &uplo, &n, a, &lda, &i1, &i2 );
         if( info < 0 ) {
             info = info - 1;
         }
     } else if( matrix_layout == LAPACK_ROW_MAJOR ) {
+        lapack_int lda_t = MAX(1,n);
         double* a_t = NULL;
         /* Allocate memory for temporary array(s) */
-        a_t = (double*)LAPACKE_malloc( sizeof(double) * n * MAX(1,n) );
+        a_t = (double*)LAPACKE_malloc( sizeof(double) * lda_t * MAX(1,n) );
         if( a_t == NULL ) {
             info = LAPACK_TRANSPOSE_MEMORY_ERROR;
             goto exit_level_0;
         }
         /* Transpose input matrices */
-        LAPACKE_dsy_trans( matrix_layout, uplo, n, a, n, a_t, n );
+        LAPACKE_dsy_trans( matrix_layout, uplo, n, a, lda, a_t, lda_t );
         /* Call LAPACK function and adjust info */
-        LAPACK_dsyswapr( &uplo, &n, a_t, &i1, &i2 );
+        LAPACK_dsyswapr( &uplo, &n, a_t, &lda_t, &i1, &i2 );
         info = 0;  /* LAPACK call is ok! */
         /* Transpose output matrices */
-        LAPACKE_dsy_trans( LAPACK_COL_MAJOR, uplo, n, a_t, n, a, n );
+        LAPACKE_dsy_trans( LAPACK_COL_MAJOR, uplo, n, a_t, lda_t, a, lda );
         /* Release memory and exit */
         LAPACKE_free( a_t );
 exit_level_0:

lapack-netlib/LAPACKE/src/lapacke_dsytri2.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function dsytri2
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -38,8 +38,8 @@ lapack_int LAPACKE_dsytri2( int matrix_layout, char uplo, lapack_int n,
 {
     lapack_int info = 0;
     lapack_int lwork = -1;
-    lapack_complex_double* work = NULL;
-    lapack_complex_double work_query;
+    double* work = NULL;
+    double work_query;
     if( matrix_layout != LAPACK_COL_MAJOR && matrix_layout != LAPACK_ROW_MAJOR ) {
         LAPACKE_xerbla( "LAPACKE_dsytri2", -1 );
         return -1;
@@ -58,8 +58,8 @@ lapack_int LAPACKE_dsytri2( int matrix_layout, char uplo, lapack_int n,
     }
     lwork = LAPACK_Z2INT( work_query );
     /* Allocate memory for work arrays */
-    work = (lapack_complex_double*)
-        LAPACKE_malloc( sizeof(lapack_complex_double) * lwork );
+    work = (double*)
+        LAPACKE_malloc( sizeof(double) * lwork );
     if( work == NULL ) {
         info = LAPACK_WORK_MEMORY_ERROR;
         goto exit_level_0;

lapack-netlib/LAPACKE/src/lapacke_dsytri2_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function dsytri2
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -36,7 +36,7 @@
 lapack_int LAPACKE_dsytri2_work( int matrix_layout, char uplo, lapack_int n,
                                  double* a, lapack_int lda,
                                  const lapack_int* ipiv,
-                                 lapack_complex_double* work, lapack_int lwork )
+                                 double* work, lapack_int lwork )
 {
     lapack_int info = 0;
     if( matrix_layout == LAPACK_COL_MAJOR ) {

lapack-netlib/LAPACKE/src/lapacke_dtpmqrt.c

@@ -28,7 +28,7 @@
 ******************************************************************************
 * Contents: Native high-level C interface to LAPACK function dtpmqrt
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -40,6 +40,8 @@ lapack_int LAPACKE_dtpmqrt( int matrix_layout, char side, char trans,
                             double* a, lapack_int lda, double* b,
                             lapack_int ldb )
 {
+    lapack_int ncols_a, nrows_a;
+    lapack_int nrows_v;
     lapack_int info = 0;
     double* work = NULL;
     if( matrix_layout != LAPACK_COL_MAJOR && matrix_layout != LAPACK_ROW_MAJOR ) {
@@ -48,16 +50,22 @@ lapack_int LAPACKE_dtpmqrt( int matrix_layout, char side, char trans,
     }
 #ifndef LAPACK_DISABLE_NAN_CHECK
     /* Optionally check input matrices for NaNs */
-    if( LAPACKE_dge_nancheck( matrix_layout, k, m, a, lda ) ) {
+    ncols_a = LAPACKE_lsame( side, 'L' ) ? n :
+                         ( LAPACKE_lsame( side, 'R' ) ? k : 0 );
+    nrows_a = LAPACKE_lsame( side, 'L' ) ? k :
+                         ( LAPACKE_lsame( side, 'R' ) ? m : 0 );
+    nrows_v = LAPACKE_lsame( side, 'L' ) ? m :
+                         ( LAPACKE_lsame( side, 'R' ) ? n : 0 );
+    if( LAPACKE_dge_nancheck( matrix_layout, nrows_a, ncols_a, a, lda ) ) {
         return -13;
     }
     if( LAPACKE_dge_nancheck( matrix_layout, m, n, b, ldb ) ) {
         return -15;
     }
-    if( LAPACKE_dge_nancheck( matrix_layout, ldt, nb, t, ldt ) ) {
+    if( LAPACKE_dge_nancheck( matrix_layout, nb, k, t, ldt ) ) {
         return -11;
     }
-    if( LAPACKE_dge_nancheck( matrix_layout, ldv, k, v, ldv ) ) {
+    if( LAPACKE_dge_nancheck( matrix_layout, nrows_v, k, v, ldv ) ) {
         return -9;
     }
 #endif

lapack-netlib/LAPACKE/src/lapacke_dtprfb.c

@@ -28,7 +28,7 @@
 ******************************************************************************
 * Contents: Native high-level C interface to LAPACK function dtprfb
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -39,6 +39,7 @@ lapack_int LAPACKE_dtprfb( int matrix_layout, char side, char trans, char direct
                            lapack_int ldv, const double* t, lapack_int ldt,
                            double* a, lapack_int lda, double* b, lapack_int ldb )
 {
+    lapack_int ncols_v, nrows_v;
     lapack_int info = 0;
     lapack_int ldwork;
     lapack_int work_size;
@@ -49,16 +50,28 @@ lapack_int LAPACKE_dtprfb( int matrix_layout, char side, char trans, char direct
     }
 #ifndef LAPACK_DISABLE_NAN_CHECK
     /* Optionally check input matrices for NaNs */
+    if( LAPACKE_lsame( storev, 'C' ) ) {
+        ncols_v = k;
+        nrows_v = LAPACKE_lsame( side, 'L' ) ? m :
+                             ( LAPACKE_lsame( side, 'R' ) ? n : 0 );
+    } else if( LAPACKE_lsame( storev, 'R' ) ) {
+        ncols_v = LAPACKE_lsame( side, 'L' ) ? m :
+                             ( LAPACKE_lsame( side, 'R' ) ? n : 0 );
+        nrows_v = k;
+    } else {
+        ncols_v = 0;
+        nrows_v = 0;
+    }
     if( LAPACKE_dge_nancheck( matrix_layout, k, m, a, lda ) ) {
         return -14;
     }
     if( LAPACKE_dge_nancheck( matrix_layout, m, n, b, ldb ) ) {
         return -16;
     }
-    if( LAPACKE_dge_nancheck( matrix_layout, ldt, k, t, ldt ) ) {
+    if( LAPACKE_dge_nancheck( matrix_layout, k, k, t, ldt ) ) {
         return -12;
     }
-    if( LAPACKE_dge_nancheck( matrix_layout, ldv, k, v, ldv ) ) {
+    if( LAPACKE_dge_nancheck( matrix_layout, nrows_v, ncols_v, v, ldv ) ) {
         return -10;
     }
 #endif

lapack-netlib/LAPACKE/src/lapacke_dtprfb_work.c

@@ -28,7 +28,7 @@
 ******************************************************************************
 * Contents: Native middle-level C interface to LAPACK function dtprfb
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -39,7 +39,7 @@ lapack_int LAPACKE_dtprfb_work( int matrix_layout, char side, char trans,
                                 const double* v, lapack_int ldv,
                                 const double* t, lapack_int ldt, double* a,
                                 lapack_int lda, double* b, lapack_int ldb,
-                                const double* work, lapack_int ldwork )
+                                double* work, lapack_int ldwork )
 {
     lapack_int info = 0;
     if( matrix_layout == LAPACK_COL_MAJOR ) {

lapack-netlib/LAPACKE/src/lapacke_ilaver.c

@@ -28,14 +28,14 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function dgesv
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
-void LAPACKE_ilaver( const lapack_int* vers_major,
-                     const lapack_int* vers_minor,
-                     const lapack_int* vers_patch )
+void LAPACKE_ilaver( lapack_int* vers_major,
+                     lapack_int* vers_minor,
+                     lapack_int* vers_patch )
 {
     /* Call LAPACK function */
     LAPACK_ilaver( vers_major, vers_minor, vers_patch );

lapack-netlib/LAPACKE/src/lapacke_sbdsvdx.c

@@ -28,20 +28,20 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function sbdsvdx
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_sbdsvdx( int matrix_layout, char uplo, char jobz, char range,
                            lapack_int n, float* d, float* e,
-                           lapack_int vl, lapack_int vu,
-                           lapack_int il, lapack_int iu, lapack_int ns,
+                           float vl, float vu,
+                           lapack_int il, lapack_int iu, lapack_int* ns,
                            float* s, float* z, lapack_int ldz,
                            lapack_int* superb )
 {
     lapack_int info = 0;
-    lapack_int lwork = 14*n;
+    lapack_int lwork = MAX(14*n,1);
     float* work = NULL;
     lapack_int* iwork = NULL;
     lapack_int i;
@@ -54,7 +54,7 @@ lapack_int LAPACKE_sbdsvdx( int matrix_layout, char uplo, char jobz, char range,
     if( LAPACKE_s_nancheck( n, d, 1 ) ) {
         return -6;
     }
-    if( LAPACKE_s_nancheck( n, e, 1 ) ) {
+    if( LAPACKE_s_nancheck( n - 1, e, 1 ) ) {
         return -7;
     }
 #endif
@@ -64,14 +64,14 @@ lapack_int LAPACKE_sbdsvdx( int matrix_layout, char uplo, char jobz, char range,
         info = LAPACK_WORK_MEMORY_ERROR;
         goto exit_level_0;
     }
-    iwork = (lapack_int*)LAPACKE_malloc( sizeof(lapack_int) * (12*n) );
+    iwork = (lapack_int*)LAPACKE_malloc( sizeof(lapack_int) * MAX(12*n,1) );
     if( iwork == NULL ) {
         info = LAPACK_WORK_MEMORY_ERROR;
         goto exit_level_1;
     }
     /* Call middle-level interface */
     info = LAPACKE_sbdsvdx_work( matrix_layout, uplo, jobz,  range,
-    							 n, d, e, vl, vu, il, iu, ns, s, z,
+    				n, d, e, vl, vu, il, iu, ns, s, z,
                                 ldz, work, iwork);
     /* Backup significant data from working array(s) */
     for( i=0; i<12*n-1; i++ ) {

lapack-netlib/LAPACKE/src/lapacke_sbdsvdx_work.c

@@ -28,15 +28,15 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function sbdsvdx
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_sbdsvdx_work( int matrix_layout, char uplo, char jobz, char range,
                            		lapack_int n, float* d, float* e,
-                           		lapack_int vl, lapack_int vu,
-                           		lapack_int il, lapack_int iu, lapack_int ns,
+                           		float vl, float vu,
+                           		lapack_int il, lapack_int iu, lapack_int* ns,
                            		float* s, float* z, lapack_int ldz,	
                                 float* work, lapack_int* iwork )
 {
@@ -44,25 +44,27 @@ lapack_int LAPACKE_sbdsvdx_work( int matrix_layout, char uplo, char jobz, char r
     if( matrix_layout == LAPACK_COL_MAJOR ) {
         /* Call LAPACK function and adjust info */
         LAPACK_sbdsvdx( &uplo, &jobz,  &range, &n, d, e, &vl, &vu,
-            			&il, &iu, &ns, s, z, &ldz,
+            			&il, &iu, ns, s, z, &ldz,
                         work, iwork, &info );
         if( info < 0 ) {
             info = info - 1;
         }
     } else if( matrix_layout == LAPACK_ROW_MAJOR ) {
-        lapack_int nrows_z = ( LAPACKE_lsame( jobz, 'v' ) ) ? MAX(2, 2*n) : 1;
+        lapack_int nrows_z = ( LAPACKE_lsame( jobz, 'v' ) ) ? 2*n : 0;
+        lapack_int ncols_z = ( LAPACKE_lsame( jobz, 'v' ) ) ?
+            ( LAPACKE_lsame( range, 'i' ) ? MAX(0,iu - il + 1) : n + 1 ) : 0;
         lapack_int ldz_t = MAX(1,nrows_z);
         float* z_t = NULL;
         /* Check leading dimension(s) */
-        if( ldz < nrows_z ) {
+        if( ldz < ncols_z ) {
             info = -3;
             LAPACKE_xerbla( "LAPACKE_sbdsvdx_work", info );
             return info;
         }
         /* Allocate memory for temporary array(s) */
-        if( LAPACKE_lsame( jobz, 'n' ) ) {
+        if( LAPACKE_lsame( jobz, 'v' ) ) {
            z_t = (float*)
-              LAPACKE_malloc( sizeof(float) * ldz_t * 2*n );
+               LAPACKE_malloc( sizeof(float) * ldz_t * MAX(2*n,1) );
            if( z_t == NULL ) {
               info = LAPACK_TRANSPOSE_MEMORY_ERROR;
               goto exit_level_0;
@@ -70,17 +72,17 @@ lapack_int LAPACKE_sbdsvdx_work( int matrix_layout, char uplo, char jobz, char r
         }
         /* Call LAPACK function and adjust info */
         LAPACK_sbdsvdx( &uplo, &jobz, &range, &n, d, e, &vl, &vu,
-            			&il, &iu, &ns, s, z_t, &ldz_t, work, 
+            			&il, &iu, ns, s, z_t, &ldz_t, work,
             			iwork, &info );
         if( info < 0 ) {
             info = info - 1;
         }
         /* Transpose output matrices */
-        if( LAPACKE_lsame( jobz, 'n' ) ) {
-            LAPACKE_sge_trans( LAPACK_COL_MAJOR, nrows_z, nrows_z, z_t, ldz_t, z, ldz);
+        if( LAPACKE_lsame( jobz, 'v' ) ) {
+            LAPACKE_sge_trans( LAPACK_COL_MAJOR, nrows_z, ncols_z, z_t, ldz_t, z, ldz);
         }
         /* Release memory and exit */
-        if( LAPACKE_lsame( jobz, 'n' ) ) {
+        if( LAPACKE_lsame( jobz, 'v' ) ) {
             LAPACKE_free( z_t );
         }
 exit_level_0:

lapack-netlib/LAPACKE/src/lapacke_sgejsv.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function sgejsv
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -70,7 +70,7 @@ lapack_int LAPACKE_sgejsv( int matrix_layout, char joba, char jobu, char jobv,
                        ( LAPACKE_lsame( jobv, 'v' ) ||
                        LAPACKE_lsame( jobv, 'j' ) ) &&
                        LAPACKE_lsame( jobv, 'j' ) ? MAX(7,m+3*n+n*n) :
-                       1) ) ) ) ) );
+                       7) ) ) ) ) );
     lapack_int* iwork = NULL;
     float* work = NULL;
     lapack_int i;
@@ -86,25 +86,25 @@ lapack_int LAPACKE_sgejsv( int matrix_layout, char joba, char jobu, char jobv,
     if( LAPACKE_sge_nancheck( matrix_layout, m, n, a, lda ) ) {
         return -10;
     }
-    if( LAPACKE_lsame( jobu, 'f' ) || LAPACKE_lsame( jobu, 'u' ) ||
-        LAPACKE_lsame( jobu, 'w' ) ) {
-        if( LAPACKE_sge_nancheck( matrix_layout, nu, n, u, ldu ) ) {
-            return -13;
-        }
-    }
-    if( LAPACKE_lsame( jobv, 'j' ) || LAPACKE_lsame( jobv, 'v' ) ||
-        LAPACKE_lsame( jobv, 'w' ) ) {
-        if( LAPACKE_sge_nancheck( matrix_layout, nv, n, v, ldv ) ) {
-            return -15;
-        }
-    }
 #endif
     /* Allocate memory for working array(s) */
-    iwork = (lapack_int*)LAPACKE_malloc( sizeof(lapack_int) * MAX(1,m+3*n) );
+    iwork = (lapack_int*)LAPACKE_malloc( sizeof(lapack_int) * MAX(3,m+3*n) );
     if( iwork == NULL ) {
         info = LAPACK_WORK_MEMORY_ERROR;
         goto exit_level_0;
     }
+    lwork = MAX3( lwork, 7, 2*m+n );
+    { /* FIXUP LWORK */
+        int want_u = LAPACKE_lsame( jobu, 'u' ) || LAPACKE_lsame( jobu, 'f' );
+        int want_v = LAPACKE_lsame( jobv, 'v' ) || LAPACKE_lsame( jobv, 'j' );
+        int want_sce = LAPACKE_lsame( joba, 'e' ) || LAPACKE_lsame( joba, 'g' );
+        if( !want_u && !want_v && !want_sce )  lwork = MAX( lwork, 4*n+1 ); // 1.1
+        if( !want_u && !want_v && want_sce )   lwork = MAX( lwork, n*n+4*n ); // 1.2
+        if( !want_u && want_v ) lwork = MAX( lwork, 4*n+1 ); // 2
+        if( want_u && !want_v ) lwork = MAX( lwork, 4*n+1 ); // 3
+        if( want_u && LAPACKE_lsame( jobv, 'v' ) ) lwork = MAX( lwork, 6*n+2*n*n ); // 4.1
+        if( want_u && LAPACKE_lsame( jobv, 'j' ) ) lwork = MAX3( lwork, 4*n+n*n, 2*n+n*n+6 ); // 4.2
+    }
     work = (float*)LAPACKE_malloc( sizeof(float) * lwork );
     if( work == NULL ) {
         info = LAPACK_WORK_MEMORY_ERROR;

lapack-netlib/LAPACKE/src/lapacke_sgejsv_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function sgejsv
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -53,6 +53,8 @@ lapack_int LAPACKE_sgejsv_work( int matrix_layout, char joba, char jobu,
     } else if( matrix_layout == LAPACK_ROW_MAJOR ) {
         lapack_int nu = LAPACKE_lsame( jobu, 'n' ) ? 1 : m;
         lapack_int nv = LAPACKE_lsame( jobv, 'n' ) ? 1 : n;
+        lapack_int ncols_u = LAPACKE_lsame( jobu, 'n' ) ? 1 :
+            LAPACKE_lsame( jobu, 'f' ) ? m : n;
         lapack_int lda_t = MAX(1,m);
         lapack_int ldu_t = MAX(1,nu);
         lapack_int ldv_t = MAX(1,nv);
@@ -65,7 +67,7 @@ lapack_int LAPACKE_sgejsv_work( int matrix_layout, char joba, char jobu,
             LAPACKE_xerbla( "LAPACKE_sgejsv_work", info );
             return info;
         }
-        if( ldu < n ) {
+        if( ldu < ncols_u ) {
             info = -14;
             LAPACKE_xerbla( "LAPACKE_sgejsv_work", info );
             return info;
@@ -83,7 +85,7 @@ lapack_int LAPACKE_sgejsv_work( int matrix_layout, char joba, char jobu,
         }
         if( LAPACKE_lsame( jobu, 'f' ) || LAPACKE_lsame( jobu, 'u' ) ||
             LAPACKE_lsame( jobu, 'w' ) ) {
-            u_t = (float*)LAPACKE_malloc( sizeof(float) * ldu_t * MAX(1,n) );
+            u_t = (float*)LAPACKE_malloc( sizeof(float) * ldu_t * MAX(1,ncols_u) );
             if( u_t == NULL ) {
                 info = LAPACK_TRANSPOSE_MEMORY_ERROR;
                 goto exit_level_1;
@@ -99,14 +101,6 @@ lapack_int LAPACKE_sgejsv_work( int matrix_layout, char joba, char jobu,
         }
         /* Transpose input matrices */
         LAPACKE_sge_trans( matrix_layout, m, n, a, lda, a_t, lda_t );
-        if( LAPACKE_lsame( jobu, 'f' ) || LAPACKE_lsame( jobu, 'u' ) ||
-            LAPACKE_lsame( jobu, 'w' ) ) {
-            LAPACKE_sge_trans( matrix_layout, nu, n, u, ldu, u_t, ldu_t );
-        }
-        if( LAPACKE_lsame( jobv, 'j' ) || LAPACKE_lsame( jobv, 'v' ) ||
-            LAPACKE_lsame( jobv, 'w' ) ) {
-            LAPACKE_sge_trans( matrix_layout, nv, n, v, ldv, v_t, ldv_t );
-        }
         /* Call LAPACK function and adjust info */
         LAPACK_sgejsv( &joba, &jobu, &jobv, &jobr, &jobt, &jobp, &m, &n, a_t,
                        &lda_t, sva, u_t, &ldu_t, v_t, &ldv_t, work, &lwork,
@@ -117,7 +111,7 @@ lapack_int LAPACKE_sgejsv_work( int matrix_layout, char joba, char jobu,
         /* Transpose output matrices */
         if( LAPACKE_lsame( jobu, 'f' ) || LAPACKE_lsame( jobu, 'u' ) ||
             LAPACKE_lsame( jobu, 'w' ) ) {
-            LAPACKE_sge_trans( LAPACK_COL_MAJOR, nu, n, u_t, ldu_t, u, ldu );
+            LAPACKE_sge_trans( LAPACK_COL_MAJOR, nu, ncols_u, u_t, ldu_t, u, ldu );
         }
         if( LAPACKE_lsame( jobv, 'j' ) || LAPACKE_lsame( jobv, 'v' ) ||
             LAPACKE_lsame( jobv, 'w' ) ) {

lapack-netlib/LAPACKE/src/lapacke_sgemqrt.c

@@ -28,7 +28,7 @@
 ******************************************************************************
 * Contents: Native high-level C interface to LAPACK function sgemqrt
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -39,6 +39,7 @@ lapack_int LAPACKE_sgemqrt( int matrix_layout, char side, char trans,
                             const float* t, lapack_int ldt, float* c,
                             lapack_int ldc )
 {
+    lapack_int nrows_v;
     lapack_int info = 0;
     float* work = NULL;
     if( matrix_layout != LAPACK_COL_MAJOR && matrix_layout != LAPACK_ROW_MAJOR ) {
@@ -47,13 +48,15 @@ lapack_int LAPACKE_sgemqrt( int matrix_layout, char side, char trans,
     }
 #ifndef LAPACK_DISABLE_NAN_CHECK
     /* Optionally check input matrices for NaNs */
+    nrows_v = LAPACKE_lsame( side, 'L' ) ? m :
+                         ( LAPACKE_lsame( side, 'R' ) ? n : 0 );
     if( LAPACKE_sge_nancheck( matrix_layout, m, n, c, ldc ) ) {
         return -12;
     }
-    if( LAPACKE_sge_nancheck( matrix_layout, ldt, nb, t, ldt ) ) {
+    if( LAPACKE_sge_nancheck( matrix_layout, nb, k, t, ldt ) ) {
         return -10;
     }
-    if( LAPACKE_sge_nancheck( matrix_layout, ldv, k, v, ldv ) ) {
+    if( LAPACKE_sge_nancheck( matrix_layout, nrows_v, k, v, ldv ) ) {
         return -8;
     }
 #endif

lapack-netlib/LAPACKE/src/lapacke_sgesvdx.c

@@ -28,15 +28,15 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function sgesvdx
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_sgesvdx( int matrix_layout, char jobu, char jobvt, char range,
                            lapack_int m, lapack_int n, float* a,
-                           lapack_int lda, lapack_int vl, lapack_int vu,
-                           lapack_int il, lapack_int iu, lapack_int ns,
+                           lapack_int lda, float vl, float vu,
+                           lapack_int il, lapack_int iu, lapack_int* ns,
                            float* s, float* u, lapack_int ldu,
                            float* vt, lapack_int ldvt,
                            lapack_int* superb )
@@ -71,7 +71,7 @@ lapack_int LAPACKE_sgesvdx( int matrix_layout, char jobu, char jobvt, char range
         info = LAPACK_WORK_MEMORY_ERROR;
         goto exit_level_0;
     }
-    iwork = (lapack_int*)LAPACKE_malloc( sizeof(lapack_int) * (12*MIN(m,n)) );
+    iwork = (lapack_int*)LAPACKE_malloc( sizeof(lapack_int) * MAX(1,(12*MIN(m,n))) );
     if( iwork == NULL ) {
         info = LAPACK_WORK_MEMORY_ERROR;
         goto exit_level_1;

lapack-netlib/LAPACKE/src/lapacke_sgesvdx_work.c

@@ -28,15 +28,15 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function sgesvdx
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_sgesvdx_work( int matrix_layout, char jobu, char jobvt, char range,
                            		lapack_int m, lapack_int n, float* a,
-                          		lapack_int lda, lapack_int vl, lapack_int vu,
-                           		lapack_int il, lapack_int iu, lapack_int ns,
+                          		lapack_int lda, float vl, float vu,
+                           		lapack_int il, lapack_int iu, lapack_int* ns,
                            		float* s, float* u, lapack_int ldu,
                            		float* vt, lapack_int ldvt,	
                                 float* work, lapack_int lwork, lapack_int* iwork )
@@ -45,21 +45,23 @@ lapack_int LAPACKE_sgesvdx_work( int matrix_layout, char jobu, char jobvt, char
     if( matrix_layout == LAPACK_COL_MAJOR ) {
         /* Call LAPACK function and adjust info */
         LAPACK_sgesvdx( &jobu, &jobvt,  &range, &m, &n, a, &lda, &vl, &vu,
-            			&il, &iu, &ns, s, u, &ldu, vt, &ldvt,
+            			&il, &iu, ns, s, u, &ldu, vt, &ldvt,
                         work, &lwork, iwork, &info );
         if( info < 0 ) {
             info = info - 1;
         }
     } else if( matrix_layout == LAPACK_ROW_MAJOR ) {
-        lapack_int nrows_u = ( LAPACKE_lsame( jobu, 'a' ) ||
-                             LAPACKE_lsame( jobu, 's' ) ) ? m : 1;
-        lapack_int ncols_u = LAPACKE_lsame( jobu, 'a' ) ? m :
-                             ( LAPACKE_lsame( jobu, 's' ) ? MIN(m,n) : 1);
-        lapack_int nrows_vt = LAPACKE_lsame( jobvt, 'a' ) ? n :
-                              ( LAPACKE_lsame( jobvt, 's' ) ? MIN(m,n) : 1);
+        lapack_int nrows_u = LAPACKE_lsame( jobu, 'v' ) ? m : 0;
+        lapack_int ncols_u = LAPACKE_lsame( jobu, 'v' ) ?
+                             ( LAPACKE_lsame( range, 'i' ) ? MAX(iu - il + 1, 0) : MIN(m,n)) : 0;
+        lapack_int nrows_vt = LAPACKE_lsame( jobvt, 'v' ) ?
+                             ( LAPACKE_lsame( range, 'i' ) ? MAX(iu - il + 1, 0) : MIN(m,n)) : 0;
+        lapack_int ncols_vt = LAPACKE_lsame( jobvt, 'v' ) ? n : 0;
+
         lapack_int lda_t = MAX(1,m);
         lapack_int ldu_t = MAX(1,nrows_u);
         lapack_int ldvt_t = MAX(1,nrows_vt);
+
         float* a_t = NULL;
         float* u_t = NULL;
         float* vt_t = NULL;
@@ -74,7 +76,7 @@ lapack_int LAPACKE_sgesvdx_work( int matrix_layout, char jobu, char jobvt, char
             LAPACKE_xerbla( "LAPACKE_sgesvdx_work", info );
             return info;
         }
-        if( ldvt < n ) {
+        if( ldvt < ncols_vt ) {
             info = -18;
             LAPACKE_xerbla( "LAPACKE_sgesvdx_work", info );
             return info;
@@ -82,7 +84,7 @@ lapack_int LAPACKE_sgesvdx_work( int matrix_layout, char jobu, char jobvt, char
         /* Query optimal working array(s) size if requested */
         if( lwork == -1 ) {
             LAPACK_sgesvdx( &jobu, &jobvt, &range, &m, &n, a, &lda_t, &vl, &vu,
-            				&il, &iu, &ns, s, u, &ldu_t, vt,
+            				&il, &iu, ns, s, u, &ldu_t, vt,
                             &ldvt_t, work, &lwork, iwork, &info );
             return (info < 0) ? (info - 1) : info;
         }
@@ -92,7 +94,7 @@ lapack_int LAPACKE_sgesvdx_work( int matrix_layout, char jobu, char jobvt, char
             info = LAPACK_TRANSPOSE_MEMORY_ERROR;
             goto exit_level_0;
         }
-        if( LAPACKE_lsame( jobu, 'a' ) || LAPACKE_lsame( jobu, 's' ) ) {
+        if( LAPACKE_lsame( jobu, 'v' ) ) { 
             u_t = (float*)
                 LAPACKE_malloc( sizeof(float) * ldu_t * MAX(1,ncols_u) );
             if( u_t == NULL ) {
@@ -100,7 +102,7 @@ lapack_int LAPACKE_sgesvdx_work( int matrix_layout, char jobu, char jobvt, char
                 goto exit_level_1;
             }
         }
-        if( LAPACKE_lsame( jobvt, 'a' ) || LAPACKE_lsame( jobvt, 's' ) ) {
+        if( LAPACKE_lsame( jobvt, 'v' ) ) { 
             vt_t = (float*)
                 LAPACKE_malloc( sizeof(float) * ldvt_t * MAX(1,n) );
             if( vt_t == NULL ) {
@@ -111,28 +113,28 @@ lapack_int LAPACKE_sgesvdx_work( int matrix_layout, char jobu, char jobvt, char
         /* Transpose input matrices */
         LAPACKE_sge_trans( matrix_layout, m, n, a, lda, a_t, lda_t );
         /* Call LAPACK function and adjust info */
-        LAPACK_sgesvdx( &jobu, &jobvt, &range, &m, &n, a, &lda_t, &vl, &vu,
-            				&il, &iu, &ns, s, u, &ldu_t, vt,
+        LAPACK_sgesvdx( &jobu, &jobvt, &range, &m, &n, a_t, &lda_t, &vl, &vu,
+            				&il, &iu, ns, s, u_t, &ldu_t, vt_t,
                             &ldvt_t, work, &lwork, iwork, &info );
         if( info < 0 ) {
             info = info - 1;
         }
         /* Transpose output matrices */
         LAPACKE_sge_trans( LAPACK_COL_MAJOR, m, n, a_t, lda_t, a, lda );
-        if( LAPACKE_lsame( jobu, 'a' ) || LAPACKE_lsame( jobu, 's' ) ) {
+        if( LAPACKE_lsame( jobu, 'v' ) ) {
             LAPACKE_sge_trans( LAPACK_COL_MAJOR, nrows_u, ncols_u, u_t, ldu_t,
                                u, ldu );
         }
-        if( LAPACKE_lsame( jobvt, 'a' ) || LAPACKE_lsame( jobvt, 's' ) ) {
+        if( LAPACKE_lsame( jobvt, 'v' ) ) { 
             LAPACKE_sge_trans( LAPACK_COL_MAJOR, nrows_vt, n, vt_t, ldvt_t, vt,
                                ldvt );
         }
         /* Release memory and exit */
-        if( LAPACKE_lsame( jobvt, 'a' ) || LAPACKE_lsame( jobvt, 's' ) ) {
+        if( LAPACKE_lsame( jobvt, 'v' ) ) { 
             LAPACKE_free( vt_t );
         }
 exit_level_2:
-        if( LAPACKE_lsame( jobu, 'a' ) || LAPACKE_lsame( jobu, 's' ) ) {
+        if( LAPACKE_lsame( jobu, 'v' ) ) {
             LAPACKE_free( u_t );
         }
 exit_level_1:

lapack-netlib/LAPACKE/src/lapacke_sgesvj.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function sgesvj
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -49,8 +49,8 @@ lapack_int LAPACKE_sgesvj( int matrix_layout, char joba, char jobu, char jobv,
     }
 #ifndef LAPACK_DISABLE_NAN_CHECK
     /* Optionally check input matrices for NaNs */
-    nrows_v = LAPACKE_lsame( jobv, 'v' ) ? n :
-                         ( LAPACKE_lsame( jobv, 'a' ) ? mv : 1);
+    nrows_v = LAPACKE_lsame( jobv, 'v' ) ? MAX(0,n) :
+            ( LAPACKE_lsame( jobv, 'a' ) ? MAX(0,mv) : 0 );
     if( LAPACKE_sge_nancheck( matrix_layout, m, n, a, lda ) ) {
         return -7;
     }

lapack-netlib/LAPACKE/src/lapacke_sgesvj_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function sgesvj
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -48,8 +48,8 @@ lapack_int LAPACKE_sgesvj_work( int matrix_layout, char joba, char jobu,
             info = info - 1;
         }
     } else if( matrix_layout == LAPACK_ROW_MAJOR ) {
-        lapack_int nrows_v = LAPACKE_lsame( jobv, 'v' ) ? n :
-                             ( LAPACKE_lsame( jobv, 'a' ) ? mv : 1);
+        lapack_int nrows_v = LAPACKE_lsame( jobv, 'v' ) ? MAX(0,n) :
+                           ( LAPACKE_lsame( jobv, 'a' ) ? MAX(0,mv) : 0);
         lapack_int lda_t = MAX(1,m);
         lapack_int ldv_t = MAX(1,nrows_v);
         float* a_t = NULL;

lapack-netlib/LAPACKE/src/lapacke_sggsvd3_work.c

@@ -91,9 +91,9 @@ lapack_int LAPACKE_sggsvd3_work( int matrix_layout, char jobu, char jobv,
         }
         /* Query optimal working array(s) size if requested */
         if( lwork == -1 ) {
-          LAPACK_sggsvd3( &jobu, &jobv, &jobq, &m, &n, &p, k, l, a_t, &lda_t,
-                          b_t, &ldb_t, alpha, beta, u_t, &ldu_t, v_t, &ldv_t,
-                          q_t, &ldq_t, work, &lwork, iwork, &info );
+          LAPACK_sggsvd3( &jobu, &jobv, &jobq, &m, &n, &p, k, l, a, &lda_t,
+                          b, &ldb_t, alpha, beta, u, &ldu_t, v, &ldv_t,
+                          q, &ldq_t, work, &lwork, iwork, &info );
           return (info < 0) ? (info - 1) : info;
         }
 

lapack-netlib/LAPACKE/src/lapacke_sggsvp3_work.c

@@ -84,16 +84,16 @@ lapack_int LAPACKE_sggsvp3_work( int matrix_layout, char jobu, char jobv,
             LAPACKE_xerbla( "LAPACKE_sggsvp3_work", info );
             return info;
         }
-        if( ldv < m ) {
+        if( ldv < p ) {
             info = -19;
             LAPACKE_xerbla( "LAPACKE_sggsvp3_work", info );
             return info;
         }
         /* Query optimal working array(s) size if requested */
         if( lwork == -1 ) {
-          LAPACK_sggsvp3( &jobu, &jobv, &jobq, &m, &p, &n, a_t, &lda_t, b_t,
-                          &ldb_t, &tola, &tolb, k, l, u_t, &ldu_t,
-                          v_t, &ldv_t, q_t, &ldq_t, iwork, tau, work, &lwork,
+          LAPACK_sggsvp3( &jobu, &jobv, &jobq, &m, &p, &n, a, &lda_t, b,
+                          &ldb_t, &tola, &tolb, k, l, u, &ldu_t,
+                          v, &ldv_t, q, &ldq_t, iwork, tau, work, &lwork,
                           &info );
           return (info < 0) ? (info - 1) : info;
         }
@@ -116,7 +116,7 @@ lapack_int LAPACKE_sggsvp3_work( int matrix_layout, char jobu, char jobv,
             }
         }
         if( LAPACKE_lsame( jobv, 'v' ) ) {
-            v_t = (float*)LAPACKE_malloc( sizeof(float) * ldv_t * MAX(1,m) );
+            v_t = (float*)LAPACKE_malloc( sizeof(float) * ldv_t * MAX(1,p) );
             if( v_t == NULL ) {
                 info = LAPACK_TRANSPOSE_MEMORY_ERROR;
                 goto exit_level_3;
@@ -146,7 +146,7 @@ lapack_int LAPACKE_sggsvp3_work( int matrix_layout, char jobu, char jobv,
             LAPACKE_sge_trans( LAPACK_COL_MAJOR, m, m, u_t, ldu_t, u, ldu );
         }
         if( LAPACKE_lsame( jobv, 'v' ) ) {
-            LAPACKE_sge_trans( LAPACK_COL_MAJOR, p, m, v_t, ldv_t, v, ldv );
+            LAPACKE_sge_trans( LAPACK_COL_MAJOR, p, p, v_t, ldv_t, v, ldv );
         }
         if( LAPACKE_lsame( jobq, 'q' ) ) {
             LAPACKE_sge_trans( LAPACK_COL_MAJOR, n, n, q_t, ldq_t, q, ldq );

lapack-netlib/LAPACKE/src/lapacke_slantr.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function slantr
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"

lapack-netlib/LAPACKE/src/lapacke_slantr_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function slantr
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"

lapack-netlib/LAPACKE/src/lapacke_slarfb_work.c

@@ -118,7 +118,7 @@ lapack_int LAPACKE_slarfb_work( int matrix_layout, char side, char trans,
             LAPACKE_sge_trans( matrix_layout, nrows_v, ncols_v-k, &v[k], ldv,
                                &v_t[k*ldv_t], ldv_t );
         } else if( LAPACKE_lsame( storev, 'r' ) &&
-                   LAPACKE_lsame( direct, 'b' ) ) {
+                   LAPACKE_lsame( direct, 'f' ) ) {
             if( k > ncols_v ) {
                 LAPACKE_xerbla( "LAPACKE_slarfb_work", -8 );
                 return -8;

lapack-netlib/LAPACKE/src/lapacke_slascl.c

@@ -28,14 +28,14 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function slaswp
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_slascl( int matrix_layout, char type, lapack_int kl,
-                           lapack_int ku, float cfrom, float cto, 
-                           lapack_int m, lapack_int n, float* a, 
+                           lapack_int ku, float cfrom, float cto,
+                           lapack_int m, lapack_int n, float* a,
                            lapack_int lda )
 {
     if( matrix_layout != LAPACK_COL_MAJOR && matrix_layout != LAPACK_ROW_MAJOR ) {
@@ -46,50 +46,64 @@ lapack_int LAPACKE_slascl( int matrix_layout, char type, lapack_int kl,
     /* Optionally check input matrices for NaNs */
     switch (type) {
     case 'G':
-       if( LAPACKE_sge_nancheck( matrix_layout, lda, n, a, lda ) ) {
-           return -9;
-           }
+        if( LAPACKE_sge_nancheck( matrix_layout, m, n, a, lda ) ) {
+            return -9;
+        }
         break;
     case 'L':
-       // TYPE = 'L' - lower triangular matrix.
-       if( LAPACKE_str_nancheck( matrix_layout, 'L', 'N', n, a, lda ) ) {
-           return -9;
-          }
+        // TYPE = 'L' - lower triangle of general matrix
+        if( matrix_layout == LAPACK_COL_MAJOR &&
+            LAPACKE_sgb_nancheck( matrix_layout, m, n, m-1, 0, a, lda+1 ) ) {
+            return -9;
+        }
+        if( matrix_layout == LAPACK_ROW_MAJOR &&
+            LAPACKE_sgb_nancheck( LAPACK_COL_MAJOR, n, m, 0, m-1, a-m+1, lda+1 ) ) {
+            return -9;
+        }
         break;
     case 'U':
-       // TYPE = 'U' - upper triangular matrix
-       if( LAPACKE_str_nancheck( matrix_layout, 'U', 'N', n, a, lda ) ) {
-           return -9;
-           } 
+        // TYPE = 'U' - upper triangle of general matrix
+        if( matrix_layout == LAPACK_COL_MAJOR &&
+            LAPACKE_sgb_nancheck( matrix_layout, m, n, 0, n-1, a-n+1, lda+1 ) ) {
+            return -9;
+        }
+        if( matrix_layout == LAPACK_ROW_MAJOR &&
+            LAPACKE_sgb_nancheck( LAPACK_COL_MAJOR, n, m, n-1, 0, a, lda+1 ) ) {
+            return -9;
+        }
         break;
     case 'H':
-       // TYPE = 'H' - upper Hessenberg matrix   
-       if( LAPACKE_shs_nancheck( matrix_layout, n, a, lda ) ) {
-           return -9;
-           }    
-        break;
+        // TYPE = 'H' - part of upper Hessenberg matrix in general matrix
+        if( matrix_layout == LAPACK_COL_MAJOR &&
+            LAPACKE_sgb_nancheck( matrix_layout, m, n, 1, n-1, a-n+1, lda+1 ) ) {
+            return -9;
+        }
+        if( matrix_layout == LAPACK_ROW_MAJOR &&
+            LAPACKE_sgb_nancheck( LAPACK_COL_MAJOR, n, m, n-1, 1, a-1, lda+1 ) ) {
+            return -9;
+        }
     case 'B':
-       // TYPE = 'B' - A is a symmetric band matrix with lower bandwidth KL
-       //             and upper bandwidth KU and with the only the lower
-       //             half stored.   
-       if( LAPACKE_ssb_nancheck( matrix_layout, 'L', n, kl, a, lda ) ) {
-           return -9;
-           }
-         break;
-   case 'Q':
-       // TYPE = 'Q' - A is a symmetric band matrix with lower bandwidth KL
-       //             and upper bandwidth KU and with the only the upper
-       //             half stored.   
-       if( LAPACKE_ssb_nancheck( matrix_layout, 'U', n, ku, a, lda ) ) {
-           return -9;
-           }
+        // TYPE = 'B' - lower part of symmetric band matrix (assume m==n)
+        if( LAPACKE_ssb_nancheck( matrix_layout, 'L', n, kl, a, lda ) ) {
+            return -9;
+        }
+        break;
+    case 'Q':
+        // TYPE = 'Q' - upper part of symmetric band matrix (assume m==n)
+        if( LAPACKE_ssb_nancheck( matrix_layout, 'U', n, ku, a, lda ) ) {
+            return -9;
+        }
         break;
     case 'Z':
-       // TYPE = 'Z' -  A is a band matrix with lower bandwidth KL and upper
-       //             bandwidth KU. See DGBTRF for storage details.        
-       if( LAPACKE_sgb_nancheck( matrix_layout, n, n, kl, kl+ku, a, lda ) ) {
-           return -6;
-           }
+        // TYPE = 'Z' -  band matrix laid out for ?GBTRF
+        if( matrix_layout == LAPACK_COL_MAJOR &&
+            LAPACKE_sgb_nancheck( matrix_layout, m, n, kl, ku, a+kl, lda ) ) {
+            return -9;
+        }
+        if( matrix_layout == LAPACK_ROW_MAJOR &&
+            LAPACKE_sgb_nancheck( matrix_layout, m, n, kl, ku, a+lda*kl, lda ) ) {
+            return -9;
+        }
         break;
     }
 #endif

lapack-netlib/LAPACKE/src/lapacke_slascl_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function slaswp
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -46,7 +46,10 @@ lapack_int LAPACKE_slascl_work( int matrix_layout, char type, lapack_int kl,
             info = info - 1;
         }
     } else if( matrix_layout == LAPACK_ROW_MAJOR ) {
-        lapack_int lda_t = MAX(1,lda);
+        lapack_int nrows_a = LAPACKE_lsame(type, 'b') ? kl + 1 :
+                             LAPACKE_lsame(type, 'q') ? ku + 1 :
+                             LAPACKE_lsame(type, 'z') ? 2 * kl + ku + 1 : m;
+        lapack_int lda_t = MAX(1,nrows_a);
         float* a_t = NULL;
         /* Check leading dimension(s) */
         if( lda < n ) {
@@ -61,12 +64,14 @@ lapack_int LAPACKE_slascl_work( int matrix_layout, char type, lapack_int kl,
             goto exit_level_0;
         }
         /* Transpose input matrices */
-        LAPACKE_sge_trans( matrix_layout, lda, n, a, lda, a_t, lda_t );
+        LAPACKE_sge_trans( matrix_layout, nrows_a, n, a, lda, a_t, lda_t );
         /* Call LAPACK function and adjust info */
         LAPACK_slascl( &type, &kl, &ku, &cfrom, &cto, &m, &n, a_t, &lda_t, &info);
-        info = 0;  /* LAPACK call is ok! */
+        if( info < 0 ) {
+            info = info - 1;
+        }
         /* Transpose output matrices */
-        LAPACKE_sge_trans( LAPACK_COL_MAJOR, lda, n, a_t, lda_t, a, lda );
+        LAPACKE_sge_trans( LAPACK_COL_MAJOR, nrows_a, n, a_t, lda_t, a, lda );
         /* Release memory and exit */
         LAPACKE_free( a_t );
 exit_level_0:

lapack-netlib/LAPACKE/src/lapacke_sorcsd2by1_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function sorcsd2by1
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -97,8 +97,8 @@ lapack_int LAPACKE_sorcsd2by1_work( int matrix_layout, char jobu1, char jobu2,
         /* Query optimal working array(s) size if requested */
         if( lwork == -1 ) {
             LAPACK_sorcsd2by1( &jobu1, &jobu2, &jobv1t, &m, &p,
-                       &q, x11, &ldx11, x21, &ldx21, 
-                       theta, u1, &ldu1, u2, &ldu2, v1t, &ldv1t, 
+                       &q, x11, &ldx11_t, x21, &ldx21_t,
+                       theta, u1, &ldu1_t, u2, &ldu2_t, v1t, &ldv1t_t,
                        work, &lwork, iwork, &info );
             return (info < 0) ? (info - 1) : info;
         }
@@ -144,8 +144,8 @@ lapack_int LAPACKE_sorcsd2by1_work( int matrix_layout, char jobu1, char jobu2,
                            ldx21_t );
         /* Call LAPACK function and adjust info */
         LAPACK_sorcsd2by1( &jobu1, &jobu2, &jobv1t, &m, &p,
-                       &q, x11, &ldx11, x21, &ldx21, 
-                       theta, u1, &ldu1, u2, &ldu2, v1t, &ldv1t, 
+                       &q, x11_t, &ldx11_t, x21_t, &ldx21_t,
+                       theta, u1_t, &ldu1_t, u2_t, &ldu2_t, v1t_t, &ldv1t_t,
                        work, &lwork, iwork, &info );
         if( info < 0 ) {
             info = info - 1;

lapack-netlib/LAPACKE/src/lapacke_sormbr_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function sormbr
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"

lapack-netlib/LAPACKE/src/lapacke_sormlq_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function sormlq
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"

lapack-netlib/LAPACKE/src/lapacke_ssbtrd.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function ssbtrd
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -48,7 +48,7 @@ lapack_int LAPACKE_ssbtrd( int matrix_layout, char vect, char uplo, lapack_int n
     if( LAPACKE_ssb_nancheck( matrix_layout, uplo, n, kd, ab, ldab ) ) {
         return -6;
     }
-    if( LAPACKE_lsame( vect, 'u' ) || LAPACKE_lsame( vect, 'v' ) ) {
+    if( LAPACKE_lsame( vect, 'u' ) ) {
         if( LAPACKE_sge_nancheck( matrix_layout, n, n, q, ldq ) ) {
             return -10;
         }

lapack-netlib/LAPACKE/src/lapacke_sstedc.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function sstedc
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -55,7 +55,7 @@ lapack_int LAPACKE_sstedc( int matrix_layout, char compz, lapack_int n, float* d
     if( LAPACKE_s_nancheck( n-1, e, 1 ) ) {
         return -5;
     }
-    if( LAPACKE_lsame( compz, 'i' ) || LAPACKE_lsame( compz, 'v' ) ) {
+    if( LAPACKE_lsame( compz, 'v' ) ) {
         if( LAPACKE_sge_nancheck( matrix_layout, n, n, z, ldz ) ) {
             return -6;
         }

lapack-netlib/LAPACKE/src/lapacke_ssteqr.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function ssteqr
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -52,7 +52,7 @@ lapack_int LAPACKE_ssteqr( int matrix_layout, char compz, lapack_int n, float* d
     if( LAPACKE_s_nancheck( n-1, e, 1 ) ) {
         return -5;
     }
-    if( LAPACKE_lsame( compz, 'i' ) || LAPACKE_lsame( compz, 'v' ) ) {
+    if( LAPACKE_lsame( compz, 'v' ) ) {
         if( LAPACKE_sge_nancheck( matrix_layout, n, n, z, ldz ) ) {
             return -6;
         }

lapack-netlib/LAPACKE/src/lapacke_ssyconv.c

@@ -28,13 +28,13 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function ssyconv
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_ssyconv( int matrix_layout, char uplo, char way, lapack_int n,
-                            float* a, lapack_int lda, const lapack_int* ipiv, float* work )
+                            float* a, lapack_int lda, const lapack_int* ipiv, float* e )
 {
     lapack_int info = 0;
     if( matrix_layout != LAPACK_COL_MAJOR && matrix_layout != LAPACK_ROW_MAJOR ) {
@@ -43,14 +43,14 @@ lapack_int LAPACKE_ssyconv( int matrix_layout, char uplo, char way, lapack_int n
     }
 #ifndef LAPACK_DISABLE_NAN_CHECK
     /* Optionally check input matrices for NaNs */
-    if( LAPACKE_sge_nancheck( matrix_layout, lda, n, a, lda ) ) {
+    if( LAPACKE_sge_nancheck( matrix_layout, n, n, a, lda ) ) {
         return -5;
     }
 #endif
    
     /* Call middle-level interface */
     info = LAPACKE_ssyconv_work( matrix_layout, uplo, way, n, a, lda, ipiv,
-                                 work );
+                                 e );
 exit_level_0:
     if( info == LAPACK_WORK_MEMORY_ERROR ) {
         LAPACKE_xerbla( "LAPACKE_ssyconv", info );

lapack-netlib/LAPACKE/src/lapacke_ssyconv_work.c

@@ -28,19 +28,19 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function ssyconv
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_ssyconv_work( int matrix_layout, char uplo, char way,
                                  lapack_int n, float* a, lapack_int lda,
-                                 const lapack_int* ipiv, float* work )
+                                 const lapack_int* ipiv, float* e )
 {
     lapack_int info = 0;
     if( matrix_layout == LAPACK_COL_MAJOR ) {
         /* Call LAPACK function and adjust info */
-        LAPACK_ssyconv( &uplo, &way, &n, a, &lda, ipiv, work, &info );
+        LAPACK_ssyconv( &uplo, &way, &n, a, &lda, ipiv, e, &info );
         if( info < 0 ) {
             info = info - 1;
         }
@@ -62,7 +62,7 @@ lapack_int LAPACKE_ssyconv_work( int matrix_layout, char uplo, char way,
         /* Transpose input matrices */
         LAPACKE_sge_trans( matrix_layout, lda, n, a, lda, a_t, lda_t );
         /* Call LAPACK function and adjust info */
-        LAPACK_ssyconv( &uplo, &way, &n, a_t, &lda_t, ipiv, work, &info );
+        LAPACK_ssyconv( &uplo, &way, &n, a_t, &lda_t, ipiv, e, &info );
         if( info < 0 ) {
             info = info - 1;
         }

lapack-netlib/LAPACKE/src/lapacke_ssyswapr.c

@@ -28,13 +28,13 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function ssyswapr
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_ssyswapr( int matrix_layout, char uplo, lapack_int n,
-                             float* a, lapack_int i1, lapack_int i2 )
+                             float* a, lapack_int lda, lapack_int i1, lapack_int i2 )
 {
     if( matrix_layout != LAPACK_COL_MAJOR && matrix_layout != LAPACK_ROW_MAJOR ) {
         LAPACKE_xerbla( "LAPACKE_ssyswapr", -1 );
@@ -42,9 +42,9 @@ lapack_int LAPACKE_ssyswapr( int matrix_layout, char uplo, lapack_int n,
     }
 #ifndef LAPACK_DISABLE_NAN_CHECK
     /* Optionally check input matrices for NaNs */
-    if( LAPACKE_ssy_nancheck( matrix_layout, uplo, n, a, n ) ) {
+    if( LAPACKE_ssy_nancheck( matrix_layout, uplo, n, a, lda ) ) {
         return -4;
     }
 #endif
-    return LAPACKE_ssyswapr_work( matrix_layout, uplo, n, a, i1, i2 );
+    return LAPACKE_ssyswapr_work( matrix_layout, uplo, n, a, lda, i1, i2 );
 }

lapack-netlib/LAPACKE/src/lapacke_ssyswapr_work.c

@@ -28,36 +28,38 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function ssyswapr
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
 
 lapack_int LAPACKE_ssyswapr_work( int matrix_layout, char uplo, lapack_int n,
-                                  float* a, lapack_int i1, lapack_int i2 )
+                                  float* a, lapack_int lda,
+                                  lapack_int i1, lapack_int i2 )
 {
     lapack_int info = 0;
     if( matrix_layout == LAPACK_COL_MAJOR ) {
         /* Call LAPACK function and adjust info */
-        LAPACK_ssyswapr( &uplo, &n, a, &i1, &i2 );
+        LAPACK_ssyswapr( &uplo, &n, a, &lda, &i1, &i2 );
         if( info < 0 ) {
             info = info - 1;
         }
     } else if( matrix_layout == LAPACK_ROW_MAJOR ) {
+        lapack_int lda_t = MAX(1,n);
         float* a_t = NULL;
         /* Allocate memory for temporary array(s) */
-        a_t = (float*)LAPACKE_malloc( sizeof(float) * n * MAX(1,n) );
+        a_t = (float*)LAPACKE_malloc( sizeof(float) * lda_t * MAX(1,n) );
         if( a_t == NULL ) {
             info = LAPACK_TRANSPOSE_MEMORY_ERROR;
             goto exit_level_0;
         }
         /* Transpose input matrices */
-        LAPACKE_ssy_trans( matrix_layout, uplo, n, a, n, a_t, n );
+        LAPACKE_ssy_trans( matrix_layout, uplo, n, a, lda, a_t, lda_t );
         /* Call LAPACK function and adjust info */
-        LAPACK_ssyswapr( &uplo, &n, a_t, &i1, &i2 );
+        LAPACK_ssyswapr( &uplo, &n, a_t, &lda_t, &i1, &i2 );
         info = 0;  /* LAPACK call is ok! */
         /* Transpose output matrices */
-        LAPACKE_ssy_trans( LAPACK_COL_MAJOR, uplo, n, a_t, n, a, n );
+        LAPACKE_ssy_trans( LAPACK_COL_MAJOR, uplo, n, a_t, lda_t, a, lda );
         /* Release memory and exit */
         LAPACKE_free( a_t );
 exit_level_0:

lapack-netlib/LAPACKE/src/lapacke_ssytri2.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function ssytri2
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -38,8 +38,8 @@ lapack_int LAPACKE_ssytri2( int matrix_layout, char uplo, lapack_int n, float* a
 {
     lapack_int info = 0;
     lapack_int lwork = -1;
-    lapack_complex_float* work = NULL;
-    lapack_complex_float work_query;
+    float* work = NULL;
+    float work_query;
     if( matrix_layout != LAPACK_COL_MAJOR && matrix_layout != LAPACK_ROW_MAJOR ) {
         LAPACKE_xerbla( "LAPACKE_ssytri2", -1 );
         return -1;
@@ -58,8 +58,8 @@ lapack_int LAPACKE_ssytri2( int matrix_layout, char uplo, lapack_int n, float* a
     }
     lwork = LAPACK_C2INT( work_query );
     /* Allocate memory for work arrays */
-    work = (lapack_complex_float*)
-        LAPACKE_malloc( sizeof(lapack_complex_float) * lwork );
+    work = (float*)
+        LAPACKE_malloc( sizeof(float) * lwork );
     if( work == NULL ) {
         info = LAPACK_WORK_MEMORY_ERROR;
         goto exit_level_0;

lapack-netlib/LAPACKE/src/lapacke_ssytri2_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function ssytri2
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -36,7 +36,7 @@
 lapack_int LAPACKE_ssytri2_work( int matrix_layout, char uplo, lapack_int n,
                                  float* a, lapack_int lda,
                                  const lapack_int* ipiv,
-                                 lapack_complex_float* work, lapack_int lwork )
+                                 float* work, lapack_int lwork )
 {
     lapack_int info = 0;
     if( matrix_layout == LAPACK_COL_MAJOR ) {

lapack-netlib/LAPACKE/src/lapacke_stpmqrt.c

@@ -28,7 +28,7 @@
 ******************************************************************************
 * Contents: Native high-level C interface to LAPACK function stpmqrt
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -39,6 +39,8 @@ lapack_int LAPACKE_stpmqrt( int matrix_layout, char side, char trans,
                             lapack_int ldv, const float* t, lapack_int ldt,
                             float* a, lapack_int lda, float* b, lapack_int ldb )
 {
+    lapack_int ncols_a, nrows_a;
+    lapack_int nrows_v;
     lapack_int info = 0;
     float* work = NULL;
     if( matrix_layout != LAPACK_COL_MAJOR && matrix_layout != LAPACK_ROW_MAJOR ) {
@@ -47,16 +49,22 @@ lapack_int LAPACKE_stpmqrt( int matrix_layout, char side, char trans,
     }
 #ifndef LAPACK_DISABLE_NAN_CHECK
     /* Optionally check input matrices for NaNs */
-    if( LAPACKE_sge_nancheck( matrix_layout, k, m, a, lda ) ) {
+    ncols_a = LAPACKE_lsame( side, 'L' ) ? n :
+                         ( LAPACKE_lsame( side, 'R' ) ? k : 0 );
+    nrows_a = LAPACKE_lsame( side, 'L' ) ? k :
+                         ( LAPACKE_lsame( side, 'R' ) ? m : 0 );
+    nrows_v = LAPACKE_lsame( side, 'L' ) ? m :
+                         ( LAPACKE_lsame( side, 'R' ) ? n : 0 );
+    if( LAPACKE_sge_nancheck( matrix_layout, nrows_a, ncols_a, a, lda ) ) {
         return -13;
     }
     if( LAPACKE_sge_nancheck( matrix_layout, m, n, b, ldb ) ) {
         return -15;
     }
-    if( LAPACKE_sge_nancheck( matrix_layout, ldt, nb, t, ldt ) ) {
+    if( LAPACKE_sge_nancheck( matrix_layout, nb, k, t, ldt ) ) {
         return -11;
     }
-    if( LAPACKE_sge_nancheck( matrix_layout, ldv, k, v, ldv ) ) {
+    if( LAPACKE_sge_nancheck( matrix_layout, nrows_v, k, v, ldv ) ) {
         return -9;
     }
 #endif

lapack-netlib/LAPACKE/src/lapacke_stprfb.c

@@ -28,7 +28,7 @@
 ******************************************************************************
 * Contents: Native high-level C interface to LAPACK function stprfb
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -39,6 +39,7 @@ lapack_int LAPACKE_stprfb( int matrix_layout, char side, char trans, char direct
                            lapack_int ldv, const float* t, lapack_int ldt,
                            float* a, lapack_int lda, float* b, lapack_int ldb)
 {
+    lapack_int ncols_v, nrows_v;
     lapack_int info = 0;
     lapack_int ldwork;
     lapack_int work_size;
@@ -49,16 +50,28 @@ lapack_int LAPACKE_stprfb( int matrix_layout, char side, char trans, char direct
     }
 #ifndef LAPACK_DISABLE_NAN_CHECK
     /* Optionally check input matrices for NaNs */
+    if( LAPACKE_lsame( storev, 'C' ) ) {
+        ncols_v = k;
+        nrows_v = LAPACKE_lsame( side, 'L' ) ? m :
+                             ( LAPACKE_lsame( side, 'R' ) ? n : 0 );
+    } else if( LAPACKE_lsame( storev, 'R' ) ) {
+        ncols_v = LAPACKE_lsame( side, 'L' ) ? m :
+                             ( LAPACKE_lsame( side, 'R' ) ? n : 0 );
+        nrows_v = k;
+    } else {
+        ncols_v = 0;
+        nrows_v = 0;
+    }
     if( LAPACKE_sge_nancheck( matrix_layout, k, m, a, lda ) ) {
         return -14;
     }
     if( LAPACKE_sge_nancheck( matrix_layout, m, n, b, ldb ) ) {
         return -16;
     }
-    if( LAPACKE_sge_nancheck( matrix_layout, ldt, k, t, ldt ) ) {
+    if( LAPACKE_sge_nancheck( matrix_layout, k, k, t, ldt ) ) {
         return -12;
     }
-    if( LAPACKE_sge_nancheck( matrix_layout, ldv, k, v, ldv ) ) {
+    if( LAPACKE_sge_nancheck( matrix_layout, nrows_v, ncols_v, v, ldv ) ) {
         return -10;
     }
 #endif

lapack-netlib/LAPACKE/src/lapacke_stprfb_work.c

@@ -28,7 +28,7 @@
 ******************************************************************************
 * Contents: Native middle-level C interface to LAPACK function stprfb
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -38,7 +38,7 @@ lapack_int LAPACKE_stprfb_work( int matrix_layout, char side, char trans,
                                 lapack_int n, lapack_int k, lapack_int l,
                                 const float* v, lapack_int ldv, const float* t,
                                 lapack_int ldt, float* a, lapack_int lda,
-                                float* b, lapack_int ldb, const float* work,
+                                float* b, lapack_int ldb, float* work,
                                 lapack_int ldwork )
 {
     lapack_int info = 0;

lapack-netlib/LAPACKE/src/lapacke_zgejsv.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native high-level C interface to LAPACK function zgejsv
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -119,7 +119,7 @@ lapack_int LAPACKE_zgejsv( int matrix_layout, char joba, char jobu, char jobv,
      (  (       ( LAPACKE_lsame( jobu, 'u' ) ||  LAPACKE_lsame( jobu, 'f' ) ) &&
                ( LAPACKE_lsame( jobv, 'v' ) ||  LAPACKE_lsame( jobv, 'j' ) ) &&
                ( LAPACKE_lsame( jobt, 't' ) ||  LAPACKE_lsame( joba, 'f' ) || LAPACKE_lsame( joba, 'g' ) ))? MAX(7,2*n) :
-                       1) ) ) ) ) ) ) ) );  
+                       7) ) ) ) ) ) ) ) );
     lapack_int* iwork = NULL;
     double* rwork = NULL;
     lapack_complex_double* cwork = NULL;
@@ -136,30 +136,29 @@ lapack_int LAPACKE_zgejsv( int matrix_layout, char joba, char jobu, char jobv,
     if( LAPACKE_zge_nancheck( matrix_layout, m, n, a, lda ) ) {
         return -10;
     }
-    if( LAPACKE_lsame( jobu, 'f' ) || LAPACKE_lsame( jobu, 'u' ) ||
-        LAPACKE_lsame( jobu, 'w' ) ) {
-        if( LAPACKE_zge_nancheck( matrix_layout, nu, n, u, ldu ) ) {
-            return -13;
-        }
-    }
-    if( LAPACKE_lsame( jobv, 'j' ) || LAPACKE_lsame( jobv, 'v' ) ||
-        LAPACKE_lsame( jobv, 'w' ) ) {
-        if( LAPACKE_zge_nancheck( matrix_layout, nv, n, v, ldv ) ) {
-            return -15;
-        }
-    }
 #endif
     /* Allocate memory for working array(s) */
-    iwork = (lapack_int*)LAPACKE_malloc( sizeof(lapack_int) * MAX(1,m+3*n) );
+    iwork = (lapack_int*)LAPACKE_malloc( sizeof(lapack_int) * MAX(3,m+2*n) );
     if( iwork == NULL ) {
         info = LAPACK_WORK_MEMORY_ERROR;
         goto exit_level_0;
     }
+    lwork = MAX( lwork, 1 );
+    { /* FIXUP LWORK */
+        int want_u = LAPACKE_lsame( jobu, 'u' ) || LAPACKE_lsame( jobu, 'f' );
+        int want_v = LAPACKE_lsame( jobv, 'v' ) || LAPACKE_lsame( jobv, 'j' );
+        int want_sce = LAPACKE_lsame( joba, 'e' ) || LAPACKE_lsame( joba, 'g' );
+        if( !want_u && !want_v && !want_sce )  lwork = MAX( lwork, 2*n+1 ); // 1.1
+        if( !want_u && !want_v && want_sce )   lwork = MAX( lwork, n*n+3*n ); // 1.2
+        if( want_u && LAPACKE_lsame( jobv, 'v' ) ) lwork = MAX( lwork, 5*n+2*n*n ); // 4.1
+        if( want_u && LAPACKE_lsame( jobv, 'j' ) ) lwork = MAX( lwork, 4*n+n*n ); // 4.2
+    }
     cwork = (lapack_complex_double*)LAPACKE_malloc( sizeof(lapack_complex_double) * lwork );
     if( cwork == NULL ) {
         info = LAPACK_WORK_MEMORY_ERROR;
         goto exit_level_1;
     }
+    lrwork = MAX3( lrwork, 7, n+2*m );
     rwork = (double*)LAPACKE_malloc( sizeof(double) * lrwork );
     if( rwork == NULL ) {
         info = LAPACK_WORK_MEMORY_ERROR;

lapack-netlib/LAPACKE/src/lapacke_zgejsv_work.c

@@ -28,7 +28,7 @@
 *****************************************************************************
 * Contents: Native middle-level C interface to LAPACK function zgejsv
 * Author: Intel Corporation
-* Generated November, 2011
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -46,7 +46,7 @@ lapack_int LAPACKE_zgejsv_work( int matrix_layout, char joba, char jobu,
     if( matrix_layout == LAPACK_COL_MAJOR ) {
         /* Call LAPACK function and adjust info */
         LAPACK_zgejsv( &joba, &jobu, &jobv, &jobr, &jobt, &jobp, &m, &n, a,
-                       &lda, sva, u, &ldu, v, &ldv, cwork, &lwork, rwork, &lwork, 
+                       &lda, sva, u, &ldu, v, &ldv, cwork, &lwork, rwork, &lrwork,
                        iwork, &info );
         if( info < 0 ) {
             info = info - 1;
@@ -54,6 +54,8 @@ lapack_int LAPACKE_zgejsv_work( int matrix_layout, char joba, char jobu,
     } else if( matrix_layout == LAPACK_ROW_MAJOR ) {
         lapack_int nu = LAPACKE_lsame( jobu, 'n' ) ? 1 : m;
         lapack_int nv = LAPACKE_lsame( jobv, 'n' ) ? 1 : n;
+        lapack_int ncols_u = LAPACKE_lsame( jobu, 'n' ) ? 1 :
+            LAPACKE_lsame( jobu, 'f' ) ? m : n;
         lapack_int lda_t = MAX(1,m);
         lapack_int ldu_t = MAX(1,nu);
         lapack_int ldv_t = MAX(1,nv);
@@ -66,7 +68,7 @@ lapack_int LAPACKE_zgejsv_work( int matrix_layout, char joba, char jobu,
             LAPACKE_xerbla( "LAPACKE_zgejsv_work", info );
             return info;
         }
-        if( ldu < n ) {
+        if( ldu < ncols_u ) {
             info = -14;
             LAPACKE_xerbla( "LAPACKE_zgejsv_work", info );
             return info;
@@ -86,7 +88,7 @@ lapack_int LAPACKE_zgejsv_work( int matrix_layout, char joba, char jobu,
         if( LAPACKE_lsame( jobu, 'f' ) || LAPACKE_lsame( jobu, 'u' ) ||
             LAPACKE_lsame( jobu, 'w' ) ) {
             u_t = (lapack_complex_double*)
-               LAPACKE_malloc( sizeof(lapack_complex_double) * ldu_t * MAX(1,n) );
+               LAPACKE_malloc( sizeof(lapack_complex_double) * ldu_t * MAX(1,ncols_u) );
             if( u_t == NULL ) {
                 info = LAPACK_TRANSPOSE_MEMORY_ERROR;
                 goto exit_level_1;
@@ -103,14 +105,6 @@ lapack_int LAPACKE_zgejsv_work( int matrix_layout, char joba, char jobu,
         }
         /* Transpose input matrices */
         LAPACKE_zge_trans( matrix_layout, m, n, a, lda, a_t, lda_t );
-        if( LAPACKE_lsame( jobu, 'f' ) || LAPACKE_lsame( jobu, 'u' ) ||
-            LAPACKE_lsame( jobu, 'w' ) ) {
-            LAPACKE_zge_trans( matrix_layout, nu, n, u, ldu, u_t, ldu_t );
-        }
-        if( LAPACKE_lsame( jobv, 'j' ) || LAPACKE_lsame( jobv, 'v' ) ||
-            LAPACKE_lsame( jobv, 'w' ) ) {
-            LAPACKE_zge_trans( matrix_layout, nv, n, v, ldv, v_t, ldv_t );
-        }
         /* Call LAPACK function and adjust info */
         LAPACK_zgejsv( &joba, &jobu, &jobv, &jobr, &jobt, &jobp, &m, &n, a_t,
                        &lda_t, sva, u_t, &ldu_t, v_t, &ldv_t, cwork, &lwork,
@@ -121,7 +115,7 @@ lapack_int LAPACKE_zgejsv_work( int matrix_layout, char joba, char jobu,
         /* Transpose output matrices */
         if( LAPACKE_lsame( jobu, 'f' ) || LAPACKE_lsame( jobu, 'u' ) ||
             LAPACKE_lsame( jobu, 'w' ) ) {
-            LAPACKE_zge_trans( LAPACK_COL_MAJOR, nu, n, u_t, ldu_t, u, ldu );
+            LAPACKE_zge_trans( LAPACK_COL_MAJOR, nu, ncols_u, u_t, ldu_t, u, ldu );
         }
         if( LAPACKE_lsame( jobv, 'j' ) || LAPACKE_lsame( jobv, 'v' ) ||
             LAPACKE_lsame( jobv, 'w' ) ) {

lapack-netlib/LAPACKE/src/lapacke_zgemqrt.c

@@ -28,7 +28,7 @@
 ******************************************************************************
 * Contents: Native high-level C interface to LAPACK function zgemqrt
 * Author: Intel Corporation
-* Generated November 2015
+* Generated June 2016
 *****************************************************************************/
 
 #include "lapacke_utils.h"
@@ -40,6 +40,7 @@ lapack_int LAPACKE_zgemqrt( int matrix_layout, char side, char trans,
                             lapack_int ldt, lapack_complex_double* c,
                             lapack_int ldc )
 {
+    lapack_int nrows_v;
     lapack_int info = 0;
     lapack_complex_double* work = NULL;
     if( matrix_layout != LAPACK_COL_MAJOR && matrix_layout != LAPACK_ROW_MAJOR ) {
@@ -48,13 +49,15 @@ lapack_int LAPACKE_zgemqrt( int matrix_layout, char side, char trans,
     }
 #ifndef LAPACK_DISABLE_NAN_CHECK
     /* Optionally check input matrices for NaNs */
+    nrows_v = LAPACKE_lsame( side, 'L' ) ? m :
+                         ( LAPACKE_lsame( side, 'R' ) ? n : 0 );
     if( LAPACKE_zge_nancheck( matrix_layout, m, n, c, ldc ) ) {
         return -12;
     }
-    if( LAPACKE_zge_nancheck( matrix_layout, ldt, nb, t, ldt ) ) {
+    if( LAPACKE_zge_nancheck( matrix_layout, nb, k, t, ldt ) ) {
         return -10;
     }
-    if( LAPACKE_zge_nancheck( matrix_layout, ldv, k, v, ldv ) ) {
+    if( LAPACKE_zge_nancheck( matrix_layout, nrows_v, k, v, ldv ) ) {
         return -8;
     }
 #endif