// This is the exact cblas.h header file, placed here purely in order to get
// the enums.

#include "caffe2/core/macros.h"

#ifndef CBLAS_H
#ifdef CAFFE2_USE_MKL
#include <mkl_cblas.h>
#else  // CAFFE2_USE_MKL

#ifndef CBLAS_ENUM_DEFINED_H
   #define CBLAS_ENUM_DEFINED_H
   enum CBLAS_ORDER {CblasRowMajor=101, CblasColMajor=102 };
   enum CBLAS_TRANSPOSE {CblasNoTrans=111, CblasTrans=112, CblasConjTrans=113,
                         AtlasConj=114};
   enum CBLAS_UPLO  {CblasUpper=121, CblasLower=122};
   enum CBLAS_DIAG  {CblasNonUnit=131, CblasUnit=132};
   enum CBLAS_SIDE  {CblasLeft=141, CblasRight=142};
#endif

#ifndef CBLAS_ENUM_ONLY
#define CBLAS_H
#define CBLAS_INDEX int

int cblas_errprn(int ierr, int info, char *form, ...);
void cblas_xerbla(int p, const char *rout, const char *form, ...);

/*
 * ===========================================================================
 * Prototypes for level 1 BLAS functions (complex are recast as routines)
 * ===========================================================================
 */
float  cblas_sdsdot(const int N, const float alpha, const float *X,
                    const int incX, const float *Y, const int incY);
double cblas_dsdot(const int N, const float *X, const int incX, const float *Y,
                   const int incY);
float  cblas_sdot(const int N, const float  *X, const int incX,
                  const float  *Y, const int incY);
double cblas_ddot(const int N, const double *X, const int incX,
                  const double *Y, const int incY);
/*
 * Functions having prefixes Z and C only
 */
void   cblas_cdotu_sub(const int N, const void *X, const int incX,
                       const void *Y, const int incY, void *dotu);
void   cblas_cdotc_sub(const int N, const void *X, const int incX,
                       const void *Y, const int incY, void *dotc);

void   cblas_zdotu_sub(const int N, const void *X, const int incX,
                       const void *Y, const int incY, void *dotu);
void   cblas_zdotc_sub(const int N, const void *X, const int incX,
                       const void *Y, const int incY, void *dotc);


/*
 * Functions having prefixes S D SC DZ
 */
float  cblas_snrm2(const int N, const float *X, const int incX);
float  cblas_sasum(const int N, const float *X, const int incX);

double cblas_dnrm2(const int N, const double *X, const int incX);
double cblas_dasum(const int N, const double *X, const int incX);

float  cblas_scnrm2(const int N, const void *X, const int incX);
float  cblas_scasum(const int N, const void *X, const int incX);

double cblas_dznrm2(const int N, const void *X, const int incX);
double cblas_dzasum(const int N, const void *X, const int incX);


/*
 * Functions having standard 4 prefixes (S D C Z)
 */
CBLAS_INDEX cblas_isamax(const int N, const float  *X, const int incX);
CBLAS_INDEX cblas_idamax(const int N, const double *X, const int incX);
CBLAS_INDEX cblas_icamax(const int N, const void   *X, const int incX);
CBLAS_INDEX cblas_izamax(const int N, const void   *X, const int incX);

/*
 * ===========================================================================
 * Prototypes for level 1 BLAS routines
 * ===========================================================================
 */

/*
 * Routines with standard 4 prefixes (s, d, c, z)
 */
void cblas_sswap(const int N, float *X, const int incX,
                 float *Y, const int incY);
void cblas_scopy(const int N, const float *X, const int incX,
                 float *Y, const int incY);
void cblas_saxpy(const int N, const float alpha, const float *X,
                 const int incX, float *Y, const int incY);
void catlas_saxpby(const int N, const float alpha, const float *X,
                  const int incX, const float beta, float *Y, const int incY);
void catlas_sset
   (const int N, const float alpha, float *X, const int incX);

void cblas_dswap(const int N, double *X, const int incX,
                 double *Y, const int incY);
void cblas_dcopy(const int N, const double *X, const int incX,
                 double *Y, const int incY);
void cblas_daxpy(const int N, const double alpha, const double *X,
                 const int incX, double *Y, const int incY);
void catlas_daxpby(const int N, const double alpha, const double *X,
                  const int incX, const double beta, double *Y, const int incY);
void catlas_dset
   (const int N, const double alpha, double *X, const int incX);

void cblas_cswap(const int N, void *X, const int incX,
                 void *Y, const int incY);
void cblas_ccopy(const int N, const void *X, const int incX,
                 void *Y, const int incY);
void cblas_caxpy(const int N, const void *alpha, const void *X,
                 const int incX, void *Y, const int incY);
void catlas_caxpby(const int N, const void *alpha, const void *X,
                  const int incX, const void *beta, void *Y, const int incY);
void catlas_cset
   (const int N, const void *alpha, void *X, const int incX);

void cblas_zswap(const int N, void *X, const int incX,
                 void *Y, const int incY);
void cblas_zcopy(const int N, const void *X, const int incX,
                 void *Y, const int incY);
void cblas_zaxpy(const int N, const void *alpha, const void *X,
                 const int incX, void *Y, const int incY);
void catlas_zaxpby(const int N, const void *alpha, const void *X,
                  const int incX, const void *beta, void *Y, const int incY);
void catlas_zset
   (const int N, const void *alpha, void *X, const int incX);


/*
 * Routines with S and D prefix only
 */
void cblas_srotg(float *a, float *b, float *c, float *s);
void cblas_srotmg(float *d1, float *d2, float *b1, const float b2, float *P);
void cblas_srot(const int N, float *X, const int incX,
                float *Y, const int incY, const float c, const float s);
void cblas_srotm(const int N, float *X, const int incX,
                float *Y, const int incY, const float *P);

void cblas_drotg(double *a, double *b, double *c, double *s);
void cblas_drotmg(double *d1, double *d2, double *b1, const double b2, double *P);
void cblas_drot(const int N, double *X, const int incX,
                double *Y, const int incY, const double c, const double s);
void cblas_drotm(const int N, double *X, const int incX,
                double *Y, const int incY, const double *P);


/*
 * Routines with S D C Z CS and ZD prefixes
 */
void cblas_sscal(const int N, const float alpha, float *X, const int incX);
void cblas_dscal(const int N, const double alpha, double *X, const int incX);
void cblas_cscal(const int N, const void *alpha, void *X, const int incX);
void cblas_zscal(const int N, const void *alpha, void *X, const int incX);
void cblas_csscal(const int N, const float alpha, void *X, const int incX);
void cblas_zdscal(const int N, const double alpha, void *X, const int incX);

/*
 * Extra reference routines provided by ATLAS, but not mandated by the standard
 */
void cblas_crotg(void *a, void *b, void *c, void *s);
void cblas_zrotg(void *a, void *b, void *c, void *s);
void cblas_csrot(const int N, void *X, const int incX, void *Y, const int incY,
                 const float c, const float s);
void cblas_zdrot(const int N, void *X, const int incX, void *Y, const int incY,
                 const double c, const double s);

/*
 * ===========================================================================
 * Prototypes for level 2 BLAS
 * ===========================================================================
 */

/*
 * Routines with standard 4 prefixes (S, D, C, Z)
 */
void cblas_sgemv(const enum CBLAS_ORDER Order,
                 const enum CBLAS_TRANSPOSE TransA, const int M, const int N,
                 const float alpha, const float *A, const int lda,
                 const float *X, const int incX, const float beta,
                 float *Y, const int incY);
void cblas_sgbmv(const enum CBLAS_ORDER Order,
                 const enum CBLAS_TRANSPOSE TransA, const int M, const int N,
                 const int KL, const int KU, const float alpha,
                 const float *A, const int lda, const float *X,
                 const int incX, const float beta, float *Y, const int incY);
void cblas_strmv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const float *A, const int lda,
                 float *X, const int incX);
void cblas_stbmv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const int K, const float *A, const int lda,
                 float *X, const int incX);
void cblas_stpmv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const float *Ap, float *X, const int incX);
void cblas_strsv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const float *A, const int lda, float *X,
                 const int incX);
void cblas_stbsv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const int K, const float *A, const int lda,
                 float *X, const int incX);
void cblas_stpsv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const float *Ap, float *X, const int incX);

void cblas_dgemv(const enum CBLAS_ORDER Order,
                 const enum CBLAS_TRANSPOSE TransA, const int M, const int N,
                 const double alpha, const double *A, const int lda,
                 const double *X, const int incX, const double beta,
                 double *Y, const int incY);
void cblas_dgbmv(const enum CBLAS_ORDER Order,
                 const enum CBLAS_TRANSPOSE TransA, const int M, const int N,
                 const int KL, const int KU, const double alpha,
                 const double *A, const int lda, const double *X,
                 const int incX, const double beta, double *Y, const int incY);
void cblas_dtrmv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const double *A, const int lda,
                 double *X, const int incX);
void cblas_dtbmv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const int K, const double *A, const int lda,
                 double *X, const int incX);
void cblas_dtpmv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const double *Ap, double *X, const int incX);
void cblas_dtrsv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const double *A, const int lda, double *X,
                 const int incX);
void cblas_dtbsv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const int K, const double *A, const int lda,
                 double *X, const int incX);
void cblas_dtpsv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const double *Ap, double *X, const int incX);

void cblas_cgemv(const enum CBLAS_ORDER Order,
                 const enum CBLAS_TRANSPOSE TransA, const int M, const int N,
                 const void *alpha, const void *A, const int lda,
                 const void *X, const int incX, const void *beta,
                 void *Y, const int incY);
void cblas_cgbmv(const enum CBLAS_ORDER Order,
                 const enum CBLAS_TRANSPOSE TransA, const int M, const int N,
                 const int KL, const int KU, const void *alpha,
                 const void *A, const int lda, const void *X,
                 const int incX, const void *beta, void *Y, const int incY);
void cblas_ctrmv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const void *A, const int lda,
                 void *X, const int incX);
void cblas_ctbmv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const int K, const void *A, const int lda,
                 void *X, const int incX);
void cblas_ctpmv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const void *Ap, void *X, const int incX);
void cblas_ctrsv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const void *A, const int lda, void *X,
                 const int incX);
void cblas_ctbsv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const int K, const void *A, const int lda,
                 void *X, const int incX);
void cblas_ctpsv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const void *Ap, void *X, const int incX);

void cblas_zgemv(const enum CBLAS_ORDER Order,
                 const enum CBLAS_TRANSPOSE TransA, const int M, const int N,
                 const void *alpha, const void *A, const int lda,
                 const void *X, const int incX, const void *beta,
                 void *Y, const int incY);
void cblas_zgbmv(const enum CBLAS_ORDER Order,
                 const enum CBLAS_TRANSPOSE TransA, const int M, const int N,
                 const int KL, const int KU, const void *alpha,
                 const void *A, const int lda, const void *X,
                 const int incX, const void *beta, void *Y, const int incY);
void cblas_ztrmv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const void *A, const int lda,
                 void *X, const int incX);
void cblas_ztbmv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const int K, const void *A, const int lda,
                 void *X, const int incX);
void cblas_ztpmv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const void *Ap, void *X, const int incX);
void cblas_ztrsv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const void *A, const int lda, void *X,
                 const int incX);
void cblas_ztbsv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const int K, const void *A, const int lda,
                 void *X, const int incX);
void cblas_ztpsv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const enum CBLAS_TRANSPOSE TransA, const enum CBLAS_DIAG Diag,
                 const int N, const void *Ap, void *X, const int incX);


/*
 * Routines with S and D prefixes only
 */
void cblas_ssymv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const int N, const float alpha, const float *A,
                 const int lda, const float *X, const int incX,
                 const float beta, float *Y, const int incY);
void cblas_ssbmv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const int N, const int K, const float alpha, const float *A,
                 const int lda, const float *X, const int incX,
                 const float beta, float *Y, const int incY);
void cblas_sspmv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                 const int N, const float alpha, const float *Ap,
                 const float *X, const int incX,
                 const float beta, float *Y, const int incY);
void cblas_sger(const enum CBLAS_ORDER Order, const int M, const int N,
                const float alpha, const float *X, const int incX,
                const float *Y, const int incY, float *A, const int lda);
void cblas_ssyr(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                const int N, const float alpha, const float *X,
                const int incX, float *A, const int lda);
void cblas_sspr(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                const int N, const float alpha, const float *X,
                const int incX, float *Ap);
void cblas_ssyr2(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                const int N, const float alpha, const float *X,
                const int incX, const float *Y, const int incY, float *A,
                const int lda);
void cblas_sspr2(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,
                const int N, const float alpha, const float *X,
                const int incX, const float *Y, const int incY, float *A);

void cblas_dsymv(const enum CBLAS_ORDER Order, const enum CBLAS_UPLO Uplo,