codecircuit

#include <cuda/runtime_api.hpp>
__device__ float pirate_value;

__global__ void pirate_float_non_assembler(char operation, int num_operations) {
	float acc{};
	switch (operation) {
		case 'a':
			for (int i = 0; i < num_operations; ++i) {
				acc += threadIdx.x;
			}

codecircuit

#include <cstdio>

// compile with: nvcc -g -G -std=c++17 test-compute-sanitizer-racecheck-with-sync-intrinsic.cu -arch sm_75 -o test-compute-sanitizer-racecheck-with-sync-intrinsic
// execute with: compute-sanitizer --tool racecheck ./test-compute-sanitizer-racecheck-with-sync-intrinsic
//
// The example at https://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html#warp-examples-broadcast
// mentions that the threads are "synchronized" in the comment behind the function call to `shfl_sync`.


__managed__ float result[1];

codecircuit

#!/usr/bin/env python

import numpy as np

import matplotlib.pyplot as plt

x0 = -10.0
x1 = 10.0
N = 1000
dx = (x1 - x0) / N

codecircuit

// compile with `nvcc -std=c++17 -arch sm_61 --expt-relaxed-constexpr main.cu -o main`
#include <cfloat>
#include <cstdio>
#include <iostream>

__global__ void kernel(float* values, int num_values) {
	const int tid = threadIdx.x;
	const int G = blockDim.x;

	float r = 0;

codecircuit

cmake_minimum_required(VERSION 3.18)

project(cuda-clang LANGUAGES CXX CUDA)

find_package(CMakeshift REQUIRED)
#find_package(gsl-lite REQUIRED)
find_package(cuda-api-wrappers REQUIRED)

add_executable(cuda-clang  "cuda-clang.cu")

codecircuit

// compile with `nvcc -std=c++17 -rdc=true --extended-lambda dynamic-parallelism-with-thrust-and-multiple-streams.cu -o dpe`
#include <iostream>
#include <thrust/device_vector.h>
#include <thrust/execution_policy.h>
#include <thrust/for_each.h>
#include <thrust/iterator/counting_iterator.h>

struct do_something {
    __device__ do_something(float* ptr) : ptr_(ptr) {}