std::atomic_exchange, std::atomic_exchange_explicit

From cppreference.com
< cpp‎ | atomic
 
 
 
Defined in header <atomic>
(1) (since C++11)
template< class T >

T atomic_exchange( std::atomic<T>* obj,

                   typename std::atomic<T>::value_type desr ) noexcept;
template< class T >

T atomic_exchange( volatile std::atomic<T>* obj,

                   typename std::atomic<T>::value_type desr ) noexcept;
(2) (since C++11)
template< class T >

T atomic_exchange_explicit( std::atomic<T>* obj,
                            typename std::atomic<T>::value_type desr,

                            std::memory_order order ) noexcept;
template< class T >

T atomic_exchange_explicit( volatile std::atomic<T>* obj,
                            typename std::atomic<T>::value_type desr,

                            std::memory_order order ) noexcept;
1) Atomically replaces the value pointed to by obj with the value of desr and returns the value obj held previously, as if by obj->exchange(desr)
2) Atomically replaces the value pointed to by obj with the value of desr and returns the value obj held previously, as if by obj->exchange(desr, order)

Parameters

obj - pointer to the atomic object to modify
desr - the value to store in the atomic object
order - the memory synchronization ordering for this operation: all values are permitted.

Return value

The value held previously by the atomic object pointed to by obj

Example

A spinlock mutex can be implemented in userspace using an atomic exchange operation, similar to std::atomic_flag_test_and_set:

#include <thread>
#include <vector>
#include <iostream>
#include <atomic>
 
std::atomic<bool> lock(false); // holds true when locked
                               // holds false when unlocked
 
void f(int n)
{
    for (int cnt = 0; cnt < 100; ++cnt) {
        while(std::atomic_exchange_explicit(&lock, true, std::memory_order_acquire))
             ; // spin until acquired
        std::cout << "Output from thread " << n << '\n';
        std::atomic_store_explicit(&lock, false, std::memory_order_release);
    }
}
int main()
{
    std::vector<std::thread> v;
    for (int n = 0; n < 10; ++n) {
        v.emplace_back(f, n);
    }
    for (auto& t : v) {
        t.join();
    }
}

Output:

Output from thread 2
Output from thread 6
Output from thread 7
...<exactly 1000 lines>...

Defect reports

The following behavior-changing defect reports were applied retroactively to previously published C++ standards.

DR Applied to Behavior as published Correct behavior
P0558R1 C++11 exact type match required because T is deduced from multiple arguments T is deduced from the atomic argument only

See also

atomically replaces the value of the atomic object and obtains the value held previously
(public member function of std::atomic<T>)
atomically compares the value of the atomic object with non-atomic argument and performs atomic exchange if equal or atomic load if not
(function template)
specializes atomic operations for std::shared_ptr
(function template)