import CudaLib.CudaMemObj;

import static java.lang.foreign.ValueLayout.*;

import static CudaLib.CudaNumLib.*;

import static CudaLib.CudaNumLib.cudaMemcpyKind.*;

public class Main {

public static void main(String[] args) throws Throwable {

int num_elements = 1000, threadsPerBlock = 256;

CudaMemObj a_arr = hostMalloc(JAVA_DOUBLE, num_elements);

CudaMemObj b_arr = hostMalloc(JAVA_DOUBLE, num_elements);

CudaMemObj c_arr = hostMalloc(JAVA_DOUBLE, num_elements);

for (int i = 0; i < 5; i++) {

a_arr.get_ptr().setAtIndex(JAVA_DOUBLE, i, 10.0);

b_arr.get_ptr().setAtIndex(JAVA_DOUBLE, i, 10.0);

}

int byte_size_arr = (int)a_arr.get_ptr().byteSize();

CudaMemObj cu_a_arr = cudaMalloc(byte_size_arr);

CudaMemObj cu_b_arr = cudaMalloc(byte_size_arr);

CudaMemObj cu_c_arr = cudaMalloc(byte_size_arr);

cudaMemcpy(cu_a_arr, a_arr, byte_size_arr, cudaMemcpyHostToDevice);

cudaMemcpy(cu_b_arr, b_arr, byte_size_arr, cudaMemcpyHostToDevice);

cuda_add(threadsPerBlock, num_elements, cu_a_arr, cu_b_arr, cu_c_arr);

cuda__powf(threadsPerBlock, num_elements, cu_c_arr, 2.0, cu_c_arr);

cudaDeviceSynchronize();

cudaMemcpy(c_arr, cu_c_arr, byte_size_arr, cudaMemcpyDeviceToHost);

for (int i = 0; i < 5; i++) {

IO.println(c_arr.get_ptr().getAtIndex(JAVA_DOUBLE, i));

}

cudaFree(cu_a_arr);

cudaFree(cu_b_arr);

cudaFree(cu_c_arr);

}

}