test

src/life.cu

@@ -7,7 +7,7 @@ __device__ int get_val_if_in(int* src,int x,int y,int i, int j){
     if(i>=x || i<0 || j>=y || j<0){
         return 0;
     }
-    return src[i*(y-1)+j];
+    return src[(j-1)*x+i];
 }
 
 __device__ int life_destiny(int* src, int x,int y,int i, int j){
@@ -18,7 +18,7 @@ __device__ int life_destiny(int* src, int x,int y,int i, int j){
             destiny += get_val_if_in(src,x,y,a,b);
         }
     }
-    destiny -= src[i*(y-1)+j];
+    destiny -= src[(j-1)*x+i];
     
     if(src[i*(y-1)+j]){
         if(destiny > 3 || destiny < 2){
@@ -74,8 +74,8 @@ void life(int* domain, int H, int W, int n_iter, dim3 grid_dim, dim3 block_dim){
     CUDA_CALL(cudaMalloc(&d_domain1, n_bytes));
     CUDA_CALL(cudaMalloc(&d_domain2, n_bytes));
 
-    cudaMemset(d_domain2, 0, n_bytes);
-    cudaMemcpy(d_domain1,domain,n_bytes,cudaMemcpyHostToDevice);
+    CUDA_CALL(cudaMemset(d_domain2, 0, n_bytes));
+    CUDA_CALL(cudaMemcpy(d_domain1,domain,n_bytes,cudaMemcpyHostToDevice));
 
     
     for(int i = 0; i<n_iter; i++){
@@ -85,12 +85,12 @@ void life(int* domain, int H, int W, int n_iter, dim3 grid_dim, dim3 block_dim){
         else{
             dummy_kernel<<<grid_dim, block_dim>>>(d_domain2,d_domain1,H,W);
         }
-        cudaDeviceSynchronize();
+        CUDA_CALL(cudaDeviceSynchronize());
     }
 
-    cudaMemcpy(domain,d_domain2,n_bytes,cudaMemcpyDeviceToHost);
+    CUDA_CALL(cudaMemcpy(domain,d_domain2,n_bytes,cudaMemcpyDeviceToHost));
 
-    cudaFree(d_domain1);
-    cudaFree(d_domain2);
+    CUDA_CALL(cudaFree(d_domain1));
+    CUDA_CALL(cudaFree(d_domain2));
 
 }
\ No newline at end of file