W3cubDocs

/C++

std::condition_variable_any::wait_for

template< class Lock, class Rep, class Period >
std::cv_status wait_for( Lock& lock,
                         const std::chrono::duration<Rep, Period>& rel_time);

(1)

(since C++11)

template< class Lock, class Rep, class Period, class Predicate >
bool wait_for( Lock& lock,
               const std::chrono::duration<Rep, Period>& rel_time,
               Predicate stop_waiting);

(2)

(since C++11)

template< class Lock, class Rep, class Period, class Predicate >
bool wait_for( Lock& lock,
               std::stop_token stoken,
               const std::chrono::duration<Rep, Period>& rel_time,
               Predicate stop_waiting);

(3)

(since C++20)

1) Atomically releases lock, blocks the current executing thread, and adds it to the list of threads waiting on *this. The thread will be unblocked when notify_all() or notify_one() is executed, or when the relative timeout rel_time expires. It may also be unblocked spuriously. When unblocked, regardless of the reason, lock is reacquired and wait_for() exits.

2) Equivalent to return wait_until(lock, std::chrono::steady_clock::now() + rel_time, std::move(stop_waiting));. This overload may be used to ignore spurious awakenings by looping until some predicate is satisfied (bool(stop_waiting()) == true).

3) Equivalent to return wait_until(lock, std::move(stoken), std::chrono::steady_clock::now() + rel_time, std::move(stop_waiting));

The standard recommends that a steady clock be used to measure the duration. This function may block for longer than timeout_duration due to scheduling or resource contention delays.

If these functions fail to meet the postcondition (lock is locked by the calling thread), std::terminate is called. For example, this could happen if relocking the mutex throws an exception.

Parameters

lock	-	an object of type `Lock` that meets the BasicLockable requirements, which must be locked by the current thread
stoken	-	a `std::stop_token` to register interruption for
rel_time	-	an object of type `std::chrono::duration` representing the maximum time to spend waiting. Note that rel_time must be small enough not to overflow when added to `std::chrono::steady_clock::now()`.
stop_waiting	-	predicate which returns `false` if the waiting should be continued (`bool(stop_waiting()) == false`). The signature of the predicate function should be equivalent to the following: `bool pred();`

Return value

1) std::cv_status::timeout if the relative timeout specified by rel_time expired, std::cv_status::no_timeout otherwise.

2) false if the predicate stop_waiting still evaluates to false after the rel_time timeout expired, otherwise true.

3) stop_waiting(), regardless of whether the timeout was met or stop was requested.

Exceptions

1) Any exception thrown by clock, time_point, or duration during the execution (clocks, time points, and durations provided by the standard library never throw)

2) Same as (1) but may also propagate exceptions thrown by stop_waiting

3) Same as (2)

Notes

Even if notified under lock, overload (1) makes no guarantees about the state of the associated predicate when returning due to timeout.

The effects of notify_one()/notify_all() and each of the three atomic parts of wait()/wait_for()/wait_until() (unlock+wait, wakeup, and lock) take place in a single total order that can be viewed as modification order of an atomic variable: the order is specific to this individual condition variable. This makes it impossible for notify_one() to, for example, be delayed and unblock a thread that started waiting just after the call to notify_one() was made.

Example

#include <iostream>
#include <atomic>
#include <condition_variable>
#include <thread>
#include <chrono>
using namespace std::chrono_literals;
 
std::condition_variable_any cv;
std::mutex cv_m;
int i;
 
void waits(int idx)
{
    std::unique_lock<std::mutex> lk(cv_m);
    if(cv.wait_for(lk, idx*100ms, []{return i == 1;})) 
        std::cerr << "Thread " << idx << " finished waiting. i == " << i << '\n';
    else
        std::cerr << "Thread " << idx << " timed out. i == " << i << '\n';
}
 
void signals()
{
    std::this_thread::sleep_for(120ms);
    std::cerr << "Notifying...\n";
    cv.notify_all();
    std::this_thread::sleep_for(100ms);
    {
        std::lock_guard<std::mutex> lk(cv_m);
        i = 1;
    }
    std::cerr << "Notifying again...\n";
    cv.notify_all();
}
 
int main()
{
    std::thread t1(waits, 1), t2(waits, 2), t3(waits, 3), t4(signals);
    t1.join();
    t2.join();
    t3.join();
    t4.join();
}

Output:

Thread 1 timed out. i == 0
Notifying...
Thread 2 timed out. i == 0
Notifying again...
Thread 3 finished waiting. i == 1

Defect reports

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

DR	Applied to	Behavior as published	Correct behavior
LWG 2093	C++11	timeout-related exceptions were missing in the specification	mentioned
LWG 2135	C++11	`wait_for` threw an exception on unlocking/relocking failure	calls `std::terminate`

wait	blocks the current thread until the condition variable is awakened (public member function)
wait_until	blocks the current thread until the condition variable is awakened or until specified time point has been reached (public member function)