最近在寫CUDA和LAPACK混編的程式，貼出Makefile檔案和源程式檔案，供大家參考。
Makefile檔案

# Makefile for GPU program based on SU

# Home path
SU_HOME = /home/ry/su
SU_HOME1 = /home/ry/su/src/Complex
MPI_HOME = /home/ry/openmpi
CUDA_HOME = /home/ry/cuda9

LAPACK_HOME = /home/ry/Lapack/lapack-3.8.0
# CLAPACK_HOME = /home/ry/Clapack/CLAPAC


# Which compiler
 

NVCC = nvcc 

# Include path
CFLAGS = -I$(SU_HOME)/include -I$(SU_HOME1)/include -I$(MPI_HOME)/include \
         -I$(CUDA_HOME)/include -I$(LAPACK_HOME)/LAPACKE/include \
         -I$(CLAPACK_HOME)/INCLUDE

# Lib path
LFLAGS = -L$(SU_HOME)/lib -L$(SU_HOME1)/lib -L$(MPI_HOME)/lib \
         -L$(CUDA_HOME)/lib64

LD
 
 = -lsu -lpar -lcwp -lm -lcublas -lcurand -lcufft -lcusparse -llapack

CU_APPS= lapack_test

all: ${CU_APPS}

%: %.cu
    -$(NVCC) -O2 -arch=sm_35 $(@F).cu -o [email protected] $(LD) $(LFLAGS) $(CFLAGS)

clean:
    rm -f ${CU_APPS} 

源程式檔案

#include "./common.h"
#include <stdio.h>
#include <stdlib.h>
 

#include <cusparse_v2.h>
#include <cusparse.h>
#include <cuda.h>
#include <cuda_runtime.h>

#include <stdio.h>
#include "time.h"
#include "par.h"
#include "su.h"
#include "segy.h"

#include "lapacke.h"

extern lapack_int LAPACKE_dgesv( int matrix_order, lapack_int n, lapack_int nrhs,
                          double* a, lapack_int lda, lapack_int* ipiv,
                          double* b, lapack_int ldb );

extern lapack_int LAPACKE_dgetrf(int nx, int nz  );


int main(int argc, char **argv)
{
    int A[5][5] = {3,2,0,1,0,
                   2,3,2,0,1,
                   0,2,3,2,0,
                   1,0,2,3,2,
                   0,1,0,2,3};

    int nx;
    int nz;

    nx=5;
    nz=5;



    return 0;

}