added linux build script and tweaked code to compile on Ubuntu

.gitignore

@@ -19,3 +19,4 @@ UpgradeLog.htm
 *.bak
 *.vsdoc
 *.XML
+*.so

src/NativeWrappers/Common/lapack_common.h

@@ -1,6 +1,8 @@
 #ifndef LAPACK_COMMON_H
 #define LAPACK_COMMON_H
 
+#include <string.h>
+
 void shift_ipiv_down(int m, int ipiv[]);
 inline void shift_ipiv_down(int m, int ipiv[]){
 	for(int i = 0; i < m; ++i ){
@@ -19,7 +21,7 @@ template<typename T>
 inline T* Clone(const int m, const int n, const T* a)
 {
 	T* clone = new T[m*n];
-	std::memcpy(clone, a, m*n*sizeof(T));
+	memcpy(clone, a, m*n*sizeof(T));
 	return clone;
 }
 
@@ -34,4 +36,4 @@ void copyBtoX (int m, int n, int bn, T b[], T x[]){
 	}
 }
 
-#endif
+#endif

src/NativeWrappers/Linux/build.sh

@@ -0,0 +1,6 @@
+export INTEL=/opt/intel
+export MKL=$INTEL/mkl
+
+#g++ --shared -fPIC -o ./x64/MathNET.Numerics.MKL.so -I$MKL/include -I../Common  ../MKL/vector_functions.c ../MKL/blas.c ../MKL/lapack.cpp $INTEL/lib/intel64/libiomp5.a $MKL/lib/intel64/libmkl_intel_lp64.a $MKL/lib/intel64/libmkl_intel_thread.a $MKL/lib/intel64/libmkl_core.a 
+
+g++ -m32 --shared -fPIC -o ./x86/MathNET.Numerics.MKL.so -I$MKL/include -I../Common  ../MKL/vector_functions.c ../MKL/blas.c ../MKL/lapack.cpp $INTEL/lib/ia32/libiomp5.a $MKL/lib/ia32/libmkl_intel.a $MKL/lib/ia32/libmkl_intel_thread.a $MKL/lib/ia32/libmkl_core.a 

src/NativeWrappers/MKL/blas.c

@@ -53,28 +53,28 @@ DLLEXPORT MKL_Complex16 z_dot_product(const MKL_INT n, const MKL_Complex16 x[],
 	return ret;
 }
 
-DLLEXPORT void s_matrix_multiply(const enum CBLAS_TRANSPOSE transA, const enum CBLAS_TRANSPOSE transB, const MKL_INT m, const MKL_INT n, const MKL_INT k, const float alpha, const float x[], const float y[], const float beta, float c[]){
+DLLEXPORT void s_matrix_multiply(CBLAS_TRANSPOSE transA, CBLAS_TRANSPOSE transB, const MKL_INT m, const MKL_INT n, const MKL_INT k, const float alpha, const float x[], const float y[], const float beta, float c[]){
 	MKL_INT lda = transA == CblasNoTrans ? m : k;
 	MKL_INT ldb = transB == CblasNoTrans ? k : n;
 
 	cblas_sgemm(CblasColMajor, transA, transB, m, n, k, alpha, x, lda, y, ldb, beta, c, m);
 }
 
-DLLEXPORT void d_matrix_multiply(const enum CBLAS_TRANSPOSE transA, const enum CBLAS_TRANSPOSE transB, const MKL_INT m, const MKL_INT n, const MKL_INT k, const double alpha, const double x[], const double y[], const double beta, double c[]){
+DLLEXPORT void d_matrix_multiply(CBLAS_TRANSPOSE transA, CBLAS_TRANSPOSE transB, const MKL_INT m, const MKL_INT n, const MKL_INT k, const double alpha, const double x[], const double y[], const double beta, double c[]){
 	MKL_INT lda = transA == CblasNoTrans ? m : k;
 	MKL_INT ldb = transB == CblasNoTrans ? k : n;
 
 	cblas_dgemm(CblasColMajor, transA, transB, m, n, k, alpha, x, lda, y, ldb, beta, c, m);
 }
 
-DLLEXPORT void c_matrix_multiply(const enum CBLAS_TRANSPOSE transA, const enum CBLAS_TRANSPOSE transB, const MKL_INT m, const MKL_INT n, const MKL_INT k, const MKL_Complex8 alpha, const MKL_Complex8 x[], const MKL_Complex8 y[], const MKL_Complex8 beta, MKL_Complex8 c[]){
+DLLEXPORT void c_matrix_multiply(CBLAS_TRANSPOSE transA, CBLAS_TRANSPOSE transB, const MKL_INT m, const MKL_INT n, const MKL_INT k, const MKL_Complex8 alpha, const MKL_Complex8 x[], const MKL_Complex8 y[], const MKL_Complex8 beta, MKL_Complex8 c[]){
 	MKL_INT lda = transA == CblasNoTrans ? m : k;
 	MKL_INT ldb = transB == CblasNoTrans ? k : n;
 
 	cblas_cgemm(CblasColMajor, transA, transB, m, n, k, &alpha, x, lda, y, ldb, &beta, c, m);
 }
 
-DLLEXPORT void z_matrix_multiply(const enum CBLAS_TRANSPOSE transA, const enum CBLAS_TRANSPOSE transB, const MKL_INT m, const MKL_INT n, const MKL_INT k, const MKL_Complex16 alpha, const MKL_Complex16 x[], const MKL_Complex16 y[], const MKL_Complex16 beta, MKL_Complex16 c[]){
+DLLEXPORT void z_matrix_multiply(CBLAS_TRANSPOSE transA, CBLAS_TRANSPOSE transB, const MKL_INT m, const MKL_INT n, const MKL_INT k, const MKL_Complex16 alpha, const MKL_Complex16 x[], const MKL_Complex16 y[], const MKL_Complex16 beta, MKL_Complex16 c[]){
 	MKL_INT lda = transA == CblasNoTrans ? m : k;
 	MKL_INT ldb = transB == CblasNoTrans ? k : n;