FLA_Gemm_external_gpu.c File Reference

(r12600)

Functions
FLA_Error	FLA_Gemm_external_gpu (FLA_Trans transa, FLA_Trans transb, FLA_Obj alpha, FLA_Obj A, void *A_gpu, FLA_Obj B, void *B_gpu, FLA_Obj beta, FLA_Obj C, void *C_gpu)

---

Function Documentation

FLA_Error FLA_Gemm_external_gpu	(	FLA_Trans	transa,
		FLA_Trans	transb,
		FLA_Obj	alpha,
		FLA_Obj	A,
		void *	A_gpu,
		FLA_Obj	B,
		void *	B_gpu,
		FLA_Obj	beta,
		FLA_Obj	C,
		void *	C_gpu
	)

References FLA_Check_error_level(), FLA_Gemm_check(), FLA_Obj_datatype(), FLA_Obj_has_zero_dim(), FLA_Obj_length(), FLA_Obj_width(), FLA_Param_map_flame_to_netlib_trans(), and FLA_Scal_external_gpu().

Referenced by FLASH_Queue_exec_task_gpu().

{
  FLA_Datatype datatype;
  int          k_AB;
  int          m_A, n_A;
  int          m_C, n_C;
  int          ldim_A;
  int          ldim_B;
  int          ldim_C;
  char         blas_transa;
  char         blas_transb;

  if ( FLA_Check_error_level() == FLA_FULL_ERROR_CHECKING ) 
    FLA_Gemm_check( transa, transb, alpha, A, B, beta, C );

  if ( FLA_Obj_has_zero_dim( C ) ) return FLA_SUCCESS;

  if ( FLA_Obj_has_zero_dim( A ) || FLA_Obj_has_zero_dim( B ) )
  {
    FLA_Scal_external_gpu( beta, C, C_gpu );
    return FLA_SUCCESS;
  }

  datatype = FLA_Obj_datatype( A );

  m_A      = FLA_Obj_length( A );
  n_A      = FLA_Obj_width( A );
  ldim_A   = FLA_Obj_length( A );

  ldim_B   = FLA_Obj_length( B );

  m_C      = FLA_Obj_length( C );
  n_C      = FLA_Obj_width( C );
  ldim_C   = FLA_Obj_length( C );

  if ( transa == FLA_NO_TRANSPOSE || transa == FLA_CONJ_NO_TRANSPOSE )
    k_AB = n_A;
  else
    k_AB = m_A;

  FLA_Param_map_flame_to_netlib_trans( transa, &blas_transa );
  FLA_Param_map_flame_to_netlib_trans( transb, &blas_transb );


  switch( datatype ){

  case FLA_FLOAT:
  {
    float *buff_alpha = ( float * ) FLA_FLOAT_PTR( alpha );
    float *buff_beta  = ( float * ) FLA_FLOAT_PTR( beta );

    cublasSgemm( blas_transa,
                 blas_transb,
                 m_C,
                 n_C,
                 k_AB,
                 *buff_alpha,
                 ( float * ) A_gpu, ldim_A,
                 ( float * ) B_gpu, ldim_B,
                 *buff_beta,
                 ( float * ) C_gpu, ldim_C );
    
    break;
  }

  case FLA_DOUBLE:
  {
    double *buff_alpha = ( double * ) FLA_DOUBLE_PTR( alpha );
    double *buff_beta  = ( double * ) FLA_DOUBLE_PTR( beta );

    cublasDgemm( blas_transa,
                 blas_transb,
                 m_C,
                 n_C,
                 k_AB,
                 *buff_alpha,
                 ( double * ) A_gpu, ldim_A,
                 ( double * ) B_gpu, ldim_B,
                 *buff_beta,
                 ( double * ) C_gpu, ldim_C );
    
    break;
  }

  case FLA_COMPLEX:
  {
    cuComplex *buff_alpha = ( cuComplex * ) FLA_COMPLEX_PTR( alpha );
    cuComplex *buff_beta  = ( cuComplex * ) FLA_COMPLEX_PTR( beta );

    cublasCgemm( blas_transa,
                 blas_transb,
                 m_C,
                 n_C,
                 k_AB,
                 *buff_alpha,
                 ( cuComplex * ) A_gpu, ldim_A,
                 ( cuComplex * ) B_gpu, ldim_B,
                 *buff_beta,
                 ( cuComplex * ) C_gpu, ldim_C );
    
    break;
  }

  case FLA_DOUBLE_COMPLEX:
  {
    cuDoubleComplex *buff_alpha = ( cuDoubleComplex * ) FLA_DOUBLE_COMPLEX_PTR( alpha );
    cuDoubleComplex *buff_beta  = ( cuDoubleComplex * ) FLA_DOUBLE_COMPLEX_PTR( beta );

    cublasZgemm( blas_transa,
                 blas_transb,
                 m_C,
                 n_C,
                 k_AB,
                 *buff_alpha,
                 ( cuDoubleComplex * ) A_gpu, ldim_A,
                 ( cuDoubleComplex * ) B_gpu, ldim_B,
                 *buff_beta,
                 ( cuDoubleComplex * ) C_gpu, ldim_C );
    
    break;
  }

  }

  return FLA_SUCCESS;
}