GPUで同時にランダムに生成された数値を使用して、これらの各並列配列の100個の要素を初期化しようとしています。しかし、私のルーチンはさまざまな乱数を生成していません。Visual Studioでコードをデバッグすると、配列内の要素ごとに1つの番号が表示されます。このコードの目的は、CImg FilledTrianglesルーチンを最適化して、可能な場合はGPUを使用することです。
何が間違っているので、どうすれば修正できますか?これが私のコードです:
__global__ void initCurand(curandState* state, unsigned long seed)
int idx = threadIdx.x + blockIdx.x * blockDim.x;
curand_init(seed, idx, 0, &state[idx]);
__syncthreads();
}
/*
* CUDA kernel that will execute 100 threads in parallel
*/
__global__ void initializeArrays(float* posx, float* posy,float* rayon, float* veloc, float* opacity
,float * angle, unsigned char** color, int height, int width, curandState* state){
int idx = threadIdx.x + blockIdx.x * blockDim.x;
curandState localState = state[idx];
__syncthreads();
posx[idx] = (float)(curand_uniform(&localState)*width);
posy[idx] = (float)(curand_uniform(&localState)*height);
rayon[idx] = (float)(10 + curand_uniform(&localState)*50);
angle[idx] = (float)(curand_uniform(&localState)*360);
veloc[idx] = (float)(curand_uniform(&localState)*20 - 10);
color[idx][0] = (unsigned char)(curand_uniform(&localState)*255);
color[idx][1] = (unsigned char)(curand_uniform(&localState)*255);
color[idx][2] = (unsigned char)(curand_uniform(&localState)*255);
opacity[idx] = (float)(0.3 + 1.5*curand_uniform(&localState));
}
これらのカーネルを準備して呼び出すホストコードは次のとおりです。グリッド内の1つのブロックに(要素ごとに)100個のスレッドを作成しようとしています。
// launch grid of threads
dim3 dimBlock(100);
dim3 dimGrid(1);
initCurand<<<dimBlock,dimGrid>>>(devState, unsigned(time(nullptr)));
// synchronize the device and the host
cudaDeviceSynchronize();
initializeArrays<<<dimBlock, dimGrid>>>(d_posx, d_posy, d_rayon, d_veloc, d_opacity, d_angle,d_color, img0.height(), img0.width(), devState);
予備知識:
// Define random properties (pos, size, colors, ..) for all triangles that will be displayed.
float posx[100], posy[100], rayon[100], angle[100], veloc[100], opacity[100];
// Define the same properties but for the device
float* d_posx;
float* d_posy;
float* d_rayon;
float* d_angle;
float* d_veloc;
float* d_opacity;
//unsigned char d_color[100][3];
unsigned char** d_color;
curandState* devState;
cudaError_t err;
// allocate memory on the device for the device arrays
err = cudaMalloc((void**)&d_posx, 100 * sizeof(float));
err = cudaMalloc((void**)&d_posy, 100 * sizeof(float));
err = cudaMalloc((void**)&d_rayon, 100 * sizeof(float));
err = cudaMalloc((void**)&d_angle, 100 * sizeof(float));
err = cudaMalloc((void**)&d_veloc, 100 * sizeof(float));
err = cudaMalloc((void**)&d_opacity, 100 * sizeof(float));
err = cudaMalloc((void**)&devState, 100*sizeof(curandState));
errCheck(err);
size_t pitch;
//allocated the device memory for source array
err = cudaMallocPitch(&d_color, &pitch, 3 * sizeof(unsigned char),100);
結果の取得:
// get the populated arrays back to the host for use
err = cudaMemcpy(posx,d_posx, 100 * sizeof(float), cudaMemcpyDeviceToHost);
err = cudaMemcpy(posy,d_posy, 100 * sizeof(float), cudaMemcpyDeviceToHost);
err = cudaMemcpy(rayon,d_rayon, 100 * sizeof(float), cudaMemcpyDeviceToHost);
err = cudaMemcpy(veloc,d_veloc, 100 * sizeof(float), cudaMemcpyDeviceToHost);
err = cudaMemcpy(opacity,d_opacity, 100 * sizeof(float), cudaMemcpyDeviceToHost);
err = cudaMemcpy(angle,d_angle, 100 * sizeof(float), cudaMemcpyDeviceToHost);
err = cudaMemcpy2D(color,pitch,d_color,100, 100 *sizeof(unsigned char),3, cudaMemcpyDeviceToHost);