mirror of
https://github.com/NGSolve/netgen.git
synced 2024-11-14 18:08:33 +05:00
68 lines
1.6 KiB
C++
68 lines
1.6 KiB
C++
|
|
#include "catch.hpp"
|
|
#include <core/ngcore.hpp>
|
|
using namespace ngcore;
|
|
using namespace std;
|
|
|
|
|
|
uint64_t shuffle(uint64_t N, uint64_t i) {
|
|
// Shuffle the numbers using multiplication with a prime number to force many updates of min, max
|
|
constexpr uint64_t P = 101;
|
|
return (N/2 + i*P) % N;
|
|
}
|
|
|
|
void testThreading(int n_threads)
|
|
{
|
|
TaskManager::SetNumThreads(n_threads);
|
|
n_threads = EnterTaskManager();
|
|
|
|
constexpr uint64_t N = 100000;
|
|
|
|
|
|
SECTION( "atomic operations" ) {
|
|
uint64_t i_min = 2*N;
|
|
uint64_t i_max = 0;
|
|
uint64_t i_sum = 0;
|
|
|
|
double d_min = 1e100;
|
|
double d_max = 0.0;
|
|
double d_sum = 0.0;
|
|
|
|
ParallelFor( Range(N), [&] (uint64_t i) {
|
|
AtomicMin(i_min, shuffle(N,i));
|
|
});
|
|
REQUIRE( i_min==0 );
|
|
|
|
ParallelFor( Range(N), [&] (uint64_t i) {
|
|
AtomicMax(i_max, shuffle(N,i));
|
|
});
|
|
REQUIRE( i_max==N-1 );
|
|
|
|
ParallelFor( Range(N), [&] (uint64_t i) {
|
|
AsAtomic(i_sum) += i;
|
|
});
|
|
REQUIRE( i_sum==N*(N-1)/2 );
|
|
|
|
ParallelFor( Range(N), [&] (double i) {
|
|
AtomicMin(d_min, static_cast<double>(shuffle(N,i)));
|
|
});
|
|
REQUIRE( d_min==0 );
|
|
|
|
ParallelFor( Range(N), [&] (double i) {
|
|
AtomicMax(d_max, static_cast<double>(shuffle(N,i)));
|
|
});
|
|
REQUIRE( d_max==N-1 );
|
|
|
|
ParallelFor( Range(N), [&] (double i) {
|
|
AtomicAdd(d_sum, i);
|
|
});
|
|
REQUIRE( d_sum==N*(N-1)/2 );
|
|
|
|
}
|
|
ExitTaskManager(n_threads);
|
|
}
|
|
|
|
TEST_CASE("Threading - 1 Thread") { testThreading(1); }
|
|
TEST_CASE("Threading - 2 Thread") { testThreading(2); }
|
|
TEST_CASE("Threading - 8 Thread") { testThreading(8); }
|