std::atomic<T>::exchange
From cppreference.com
T exchange( T desired, std::memory_order order = std::memory_order_seq_cst ) noexcept; |
(1) | (since C++11) |
T exchange( T desired, std::memory_order order = std::memory_order_seq_cst ) volatile noexcept; |
(2) | (since C++11) |
Atomically replaces the underlying value with desired (a read-modify-write operation). Memory is affected according to the value of order.
It is deprecated if std::atomic<T>::is_always_lock_free is false and overload (2) participates in overload resolution. |
(since C++20) |
Contents |
[edit] Parameters
desired | - | value to assign |
order | - | memory order constraints to enforce |
[edit] Return value
The value of the atomic variable before the call.
[edit] Example
Run this code
#include <algorithm> #include <atomic> #include <cstddef> #include <iostream> #include <syncstream> #include <thread> #include <vector> int main() { constexpr int thread_count{5}; constexpr int sum{5}; std::atomic<int> atom{0}; std::atomic<int> counter{0}; auto increment_to_sum = [&](const int id) { for (int next = 0; next < sum;) { // each thread is writing a value from its own knowledge const int current = atom.exchange(next); counter++; // sync writing to prevent from interrupting by other threads std::osyncstream(std::cout) << "Thread #" << id << " (id=" << std::this_thread::get_id() << ") wrote " << next << " replacing the old value " << current << ".\n"; next = std::max(current, next) + 1; } }; std::vector<std::thread> v; for (std::size_t i = 0; i < thread_count; ++i) v.emplace_back(increment_to_sum, i); for (auto& tr : v) tr.join(); std::cout << thread_count << " threads take " << counter << " times in total to " << "increment 0 to " << sum << ".\n"; }
Possible output:
Thread #1 (id=139722332333824) wrote 0 replacing the old value 0. Thread #2 (id=139722323941120) wrote 0 replacing the old value 0. Thread #1 (id=139722332333824) wrote 1 replacing the old value 0. Thread #1 (id=139722332333824) wrote 2 replacing the old value 1. Thread #1 (id=139722332333824) wrote 3 replacing the old value 2. Thread #1 (id=139722332333824) wrote 4 replacing the old value 3. Thread #0 (id=139722340726528) wrote 0 replacing the old value 0. Thread #3 (id=139722315548416) wrote 0 replacing the old value 0. Thread #3 (id=139722315548416) wrote 1 replacing the old value 4. Thread #0 (id=139722340726528) wrote 1 replacing the old value 1. Thread #4 (id=139722307155712) wrote 0 replacing the old value 1. Thread #4 (id=139722307155712) wrote 2 replacing the old value 2. Thread #4 (id=139722307155712) wrote 3 replacing the old value 2. Thread #4 (id=139722307155712) wrote 4 replacing the old value 3. Thread #2 (id=139722323941120) wrote 1 replacing the old value 0. Thread #0 (id=139722340726528) wrote 2 replacing the old value 1. Thread #2 (id=139722323941120) wrote 2 replacing the old value 4. Thread #0 (id=139722340726528) wrote 3 replacing the old value 2. Thread #0 (id=139722340726528) wrote 4 replacing the old value 3. 5 threads take 19 times in total to increment 0 to 5.
[edit] See also
(C++11)(C++11) |
atomically replaces the value of the atomic object with non-atomic argument and returns the old value of the atomic (function template) |
(C++14) |
replaces the argument with a new value and returns its previous value (function template) |