【新手提问】cudaMalloc报错 an illegal memory access was encountered

显卡是Quadro P1000 最新驱动,CUDA 8, 代码如下

#include "cuda_runtime.h"
#include "device_launch_parameters.h"

#include <stdio.h>
//#include <cmath>
#include<stdint.h>
//typedef BYTE  uint16_t;
//typedef int  uint16_t;
#include "CudaKernelInfo.h"
#include <iostream>
//#include <iostream>
using namespace std;

cudaError_t addWithCuda(int *c, const int *a, const int *b, unsigned int size);

__global__ void addKernel(int *c, const int *a, const int *b)
{
   int i = threadIdx.x;
   c[i] = a[i] + b[i];
}


/*__device__ inline float lerp(float v0, float v1, float t)
{
		
	return fmaf(t, v1, fmaf(-t, v0, v0));
}*/

__device__ float lerp(float v0, float v1, float t)
{

	return fmaf(t, v1, fmaf(-t, v0, v0));
}


__global__ void VolumeProcessing_resizeAndMaskKernel(
	uint16_t * out_ptr, const int out_stride,
	const float * in_ptr, const int in_stride,
	const int out_size_x, const int out_size_y,
	const int in_size_x, const int in_size_y, const int in_size_z,
	int slice_index, float resample_step, float radius_sqr,
	const int max_voxel_value)
{
	int ox = blockIdx.x*blockDim.x + threadIdx.x;
	int oy = blockIdx.y*blockDim.y + threadIdx.y;

	//std::cout << ox << " " << oy;
	if (ox >= out_size_x || oy >= out_size_y)
	{
		//cout << "exceed limit";
		return;
	}
	
	
	float dx = fmaf(0.5f, out_size_x, float(-ox) - 0.5f);
	float dy = fmaf(0.5f, out_size_y, float(-oy) - 0.5f);
	float d = fmaf(dx, dx, dy*dy);
	

	float value = 0.0f;

	if (d <= radius_sqr)
	{
		// clamp to edge
		float ix = resample_step * ox;
		float iy = resample_step * oy;
		float iz = resample_step * slice_index;

		// 0: first, 1: next voxel
		int x0 = min(int(ix) + 0, in_size_x - 1);
		int x1 = min(int(ix) + 1, in_size_x - 1);
		int y0 = min(int(iy) + 0, in_size_y - 1);
		int y1 = min(int(iy) + 1, in_size_y - 1);
		int z0 = min(int(iz) + 0, in_size_z - 1);
		int z1 = min(int(iz) + 1, in_size_z - 1);

		// weight of next voxel
		float t = min(ix - x0, 1.0f);
		float u = min(iy - y0, 1.0f);
		float v = min(iz - z0, 1.0f);



		float xy0 = lerp(
			lerp(in_ptr[x0 + (y0 + z0 * in_size_y) * in_stride]
				, in_ptr[x1 + (y0 + z0 * in_size_y) * in_stride], t),
			lerp(in_ptr[x0 + (y1 + z0 * in_size_y) * in_stride]
				, in_ptr[x1 + (y1 + z0 * in_size_y) * in_stride], t), u);

		float xy1 = lerp(
			lerp(in_ptr[x0 + (y0 + z1 * in_size_y) * in_stride]
				, in_ptr[x1 + (y0 + z1 * in_size_y) * in_stride], t),
			lerp(in_ptr[x0 + (y1 + z1 * in_size_y) * in_stride]
				, in_ptr[x1 + (y1 + z1 * in_size_y) * in_stride], t), u);

		value = lerp(xy0, xy1, v);
	}

	out_ptr[ox + oy * out_stride] = uint16_t(min(max(0.5f, value + 0.5f), 0.5f + max_voxel_value));
}



int main()
{
   /* const int arraySize = 5;
   const int a[arraySize] = { 1, 2, 3, 4, 5 };
   const int b[arraySize] = { 10, 20, 30, 40, 50 };
   int c[arraySize] = { 0 };

   // Add vectors in parallel.
   cudaError_t cudaStatus = addWithCuda(c, a, b, arraySize);
   if (cudaStatus != cudaSuccess) {
   fprintf(stderr, "addWithCuda failed!");
   return 1;
   }

   printf("{1,2,3,4,5} + {10,20,30,40,50} = {%d,%d,%d,%d,%d}\n",
   c[0], c[1], c[2], c[3], c[4]); */

   // cudaDeviceReset must be called before exiting in order for profiling and
   // tracing tools such as Nsight and Visual Profiler to show complete traces.

	// test resample
	// add by yyy
	
	cudaError_t cudaStatus;

	cudaStatus = cudaSetDevice(0);
	if (cudaStatus != cudaSuccess) {
		fprintf(stderr, "cudaSetDevice failed!  Do you have a CUDA-capable GPU installed?");
		
	}

	uint16_t *out_ptr;
	const int out_stride = 448;
	//const float *in_ptr;
	float *in_ptr;
	//void *v_in_ptr;
	const int in_stride = 480;
	const int out_size_x = 420;
	const int out_size_y = 420;
	const int in_size_x = 420;
	const int in_size_y = 420;
	const int in_size_z = 250;
	int slice_index = 0;
	float resample_step = 1;
	float radius_sqr = 209.5;
	const int max_voxel_value = 8191;

	//CUDACHECK
	//cudaMemset
	out_ptr = new uint16_t[420 * 420];
	//float *t1 = new float[420 * 420 * 250];

    // 执行到这里报错 an illegal memory access was encountered
	cudaStatus=cudaMalloc((void**)&in_ptr, 420 * 420 * 250 * sizeof(float));
	
	if (cudaStatus != cudaSuccess) {
		fprintf(stderr, "addKernel launch failed: %s\n", cudaGetErrorString(cudaStatus));
	}

	
	//cudaMemcpy(v_in_ptr, t1, 420 * 420 * 250 * sizeof(float), cudaMemcpyHostToDevice);
	//cudaMemcpy()
	//CudaKernelInfo launch(420, 420);
	//CudaKernelInfo launch(slice.sizeX(), slice.sizeY());
	
	// block size
	CudaKernelInfo launch(420, 420);
	
	//cout << launch.gridSize() << " " << launch.threadBlockSize();
	
   VolumeProcessing_resizeAndMaskKernel<<<launch.gridSize(),launch.threadBlockSize()>>>(out_ptr, out_stride, in_ptr, in_stride,
		out_size_x, out_size_y, in_size_x, in_size_y, in_size_z, slice_index, resample_step, radius_sqr, max_voxel_value);


	cudaStatus = cudaDeviceSynchronize();
	if (cudaStatus != cudaSuccess) {
		fprintf(stderr, "cudaDeviceSynchronize returned error code %d after launching addKernel!\n", cudaStatus);
	}

   /*cudaStatus = cudaDeviceReset();
   if (cudaStatus != cudaSuccess) {
   fprintf(stderr, "cudaDeviceReset failed!");
   return 1;
   }*/
	//cudaFreeArray(in_ptr);
	delete[]out_ptr;
	cudaFree(in_ptr); 
	
   return 0;
}

// Helper function for using CUDA to add vectors in parallel.
cudaError_t addWithCuda(int *c, const int *a, const int *b, unsigned int size)
{
   int *dev_a = 0;
   int *dev_b = 0;
   int *dev_c = 0;
   cudaError_t cudaStatus;

   // Choose which GPU to run on, change this on a multi-GPU system.
   cudaStatus = cudaSetDevice(0);
   if (cudaStatus != cudaSuccess) {
   fprintf(stderr, "cudaSetDevice failed!  Do you have a CUDA-capable GPU installed?");
   goto Error;
   }

   // Allocate GPU buffers for three vectors (two input, one output)    .
   cudaStatus = cudaMalloc((void**)&dev_c, size * sizeof(int));
   if (cudaStatus != cudaSuccess) {
   fprintf(stderr, "cudaMalloc failed!");
   goto Error;
   }

   cudaStatus = cudaMalloc((void**)&dev_a, size * sizeof(int));
   if (cudaStatus != cudaSuccess) {
   fprintf(stderr, "cudaMalloc failed!");
   goto Error;
   }

   cudaStatus = cudaMalloc((void**)&dev_b, size * sizeof(int));
   if (cudaStatus != cudaSuccess) {
   fprintf(stderr, "cudaMalloc failed!");
   goto Error;
   }

   // Copy input vectors from host memory to GPU buffers.
   cudaStatus = cudaMemcpy(dev_a, a, size * sizeof(int), cudaMemcpyHostToDevice);
   if (cudaStatus != cudaSuccess) {
   fprintf(stderr, "cudaMemcpy failed!");
   goto Error;
   }

   cudaStatus = cudaMemcpy(dev_b, b, size * sizeof(int), cudaMemcpyHostToDevice);
   if (cudaStatus != cudaSuccess) {
   fprintf(stderr, "cudaMemcpy failed!");
   goto Error;
   }

   // Launch a kernel on the GPU with one thread for each element.
   addKernel<<<1, size>>>(dev_c, dev_a, dev_b);

   // Check for any errors launching the kernel
   cudaStatus = cudaGetLastError();
   if (cudaStatus != cudaSuccess) {
   fprintf(stderr, "addKernel launch failed: %s\n", cudaGetErrorString(cudaStatus));
   goto Error;
   }
   
   // cudaDeviceSynchronize waits for the kernel to finish, and returns
   // any errors encountered during the launch.
   cudaStatus = cudaDeviceSynchronize();
   if (cudaStatus != cudaSuccess) {
   fprintf(stderr, "cudaDeviceSynchronize returned error code %d after launching addKernel!\n", cudaStatus);
   goto Error;
   }

   // Copy output vector from GPU buffer to host memory.
   cudaStatus = cudaMemcpy(c, dev_c, size * sizeof(int), cudaMemcpyDeviceToHost);
   if (cudaStatus != cudaSuccess) {
   fprintf(stderr, "cudaMemcpy failed!");
   goto Error;
   }

Error:
   cudaFree(dev_c);
   cudaFree(dev_a);
   cudaFree(dev_b);
   
   return cudaStatus;
}