mirror of
https://github.com/boostorg/math.git
synced 2025-05-11 21:33:52 +00:00
bbb8eee1d6
Add SYCL testing of expint Add markers to forward decls Add CUDA testing of expint Fix static variable usage under NVRTC Add NVRTC testing Add configurable definition of complex Add function aliases Add GPU support to gegenbauer polynomials Add SYCL testing of gegenbauer Add NVCC testing of gegenbauer Add NVRTC testing of gegenbauer Add GPU support for hankel Add SYCL testing of hankel Add NVCC testing of cyl_hankel_1 Add comprehensive NVCC testing Add NVRTC testing of cyl and sph hankel Update docs Fix writing cuda::std::complex<T> to stdout Add GPU support to hermite Add SYCL testing of hermite Add CUDA testing of hermite Add NVRTC testing of hermite Add markers to hermite docs
121 lines
3.5 KiB
Plaintext
121 lines
3.5 KiB
Plaintext
|
|
// Copyright John Maddock 2016.
|
|
// Copyright Matt Borland 2024.
|
|
// Use, modification and distribution are subject to the
|
|
// Boost Software License, Version 1.0. (See accompanying file
|
|
// LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
|
|
|
|
#include <iostream>
|
|
#include <iomanip>
|
|
#include <vector>
|
|
#include <boost/math/special_functions.hpp>
|
|
#include "cuda_managed_ptr.hpp"
|
|
#include "stopwatch.hpp"
|
|
|
|
// For the CUDA runtime routines (prefixed with "cuda_")
|
|
#include <cuda_runtime.h>
|
|
|
|
typedef float float_type;
|
|
|
|
/**
|
|
* CUDA Kernel Device code
|
|
*
|
|
*/
|
|
__global__ void cuda_test(const float_type *in1, const float_type *in2, float_type *out, int numElements)
|
|
{
|
|
using std::cos;
|
|
int i = blockDim.x * blockIdx.x + threadIdx.x;
|
|
|
|
if (i < numElements)
|
|
{
|
|
out[i] = boost::math::hermite(1U, in2[i]);
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Host main routine
|
|
*/
|
|
int main(void)
|
|
{
|
|
// Error code to check return values for CUDA calls
|
|
cudaError_t err = cudaSuccess;
|
|
|
|
// Print the vector length to be used, and compute its size
|
|
int numElements = 50000;
|
|
std::cout << "[Vector operation on " << numElements << " elements]" << std::endl;
|
|
|
|
// Allocate the managed input vector A
|
|
cuda_managed_ptr<float_type> input_vector1(numElements);
|
|
|
|
// Allocate the managed input vector B
|
|
cuda_managed_ptr<float_type> input_vector2(numElements);
|
|
|
|
// Allocate the managed output vector C
|
|
cuda_managed_ptr<float_type> output_vector(numElements);
|
|
|
|
// Initialize the input vectors
|
|
for (int i = 0; i < numElements; ++i)
|
|
{
|
|
input_vector1[i] = rand()/(float_type)RAND_MAX;
|
|
input_vector2[i] = rand()/(float_type)RAND_MAX;
|
|
}
|
|
|
|
// Launch the Vector Add CUDA Kernel
|
|
int threadsPerBlock = 256;
|
|
int blocksPerGrid =(numElements + threadsPerBlock - 1) / threadsPerBlock;
|
|
std::cout << "CUDA kernel launch with " << blocksPerGrid << " blocks of " << threadsPerBlock << " threads" << std::endl;
|
|
|
|
watch w;
|
|
|
|
cuda_test<<<blocksPerGrid, threadsPerBlock>>>(input_vector1.get(), input_vector2.get(), output_vector.get(), numElements);
|
|
cudaDeviceSynchronize();
|
|
|
|
std::cout << "CUDA kernal done in: " << w.elapsed() << "s" << std::endl;
|
|
|
|
err = cudaGetLastError();
|
|
|
|
if (err != cudaSuccess)
|
|
{
|
|
std::cerr << "Failed to launch CUDA kernel (error code " << cudaGetErrorString(err) << ")!" << std::endl;
|
|
return EXIT_FAILURE;
|
|
}
|
|
|
|
// Verify that the result vector is correct
|
|
std::vector<float_type> results;
|
|
results.reserve(numElements);
|
|
w.reset();
|
|
for(int i = 0; i < numElements; ++i)
|
|
results.push_back(boost::math::hermite(1U, input_vector2[i]));
|
|
double t = w.elapsed();
|
|
// check the results
|
|
int fail_counter = 0;
|
|
for(int i = 0; i < numElements; ++i)
|
|
{
|
|
if (std::isfinite(results[i]))
|
|
{
|
|
if (boost::math::epsilon_difference(output_vector[i], results[i]) > 1000)
|
|
{
|
|
std::cerr << "Result verification failed at element " << i << "!\n"
|
|
<< "Device: " << output_vector[i] << '\n'
|
|
<< " Host: " << results[i] << '\n'
|
|
<< " Eps: " << boost::math::epsilon_difference(output_vector[i], results[i]) << std::endl;
|
|
fail_counter++;
|
|
if (fail_counter > 100)
|
|
{
|
|
break;
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
if (fail_counter > 0)
|
|
{
|
|
return EXIT_FAILURE;
|
|
}
|
|
|
|
std::cout << "Test PASSED, normal calculation time: " << t << "s" << std::endl;
|
|
std::cout << "Done\n";
|
|
|
|
return 0;
|
|
}
|