cuda x86 compile error

Hi,I’m new to CUDA x86.When I compile the cu file with the command

“pgcpp -c -O -Mcudax86”

(Win7 64bit)
,error message is below:

“C:\PROGRA~1\PGI/win64/2015/cuda/6.0/include/surface_functions.h”, line 571: error:
no instance of overloaded function “surf2Dread” matches the
specified type
forceinline device short4 surf2Dread(surface<void, cudaSurfaceType2D>
surf, int x, int y, enum cudaSurfaceBoundaryMode mode).

The cu file code is

#include "cuConvolution.h"
int cuConvolution()
	int nRet = ERR_SUCCESS;
	return nRet;

The header file code is

#ifndef GPU_CONV_H
#define GPU_CONV_H
#define GPUEXPORT __declspec(dllexport)
#define GPUEXPORT __declspec(dllimport)

int GPUEXPORT cuConvolution();


Can anyone gives me some advices?Thanks in advance.

From searching the forum,I know that cuda-x86 is no longer supported.Bad news!

Huipeng -

You are correct. We do not support Cuda-x86 on Windows,
and pgcpp on windows is being depecated.

-Deborah Caruso