tf::Runtime class

tf::Runtime::~Runtime()

destroys the runtime object

Issues a tf::Runtime::corun_all to finish all spawned asynchronous tasks and then destroys the runtime object.

Executor& tf::Runtime::executor()

obtains the running executor

The running executor of a runtime task is the executor that runs the parent taskflow of that runtime task.

tf::Executor executor;
tf::Taskflow taskflow;
taskflow.emplace([&](tf::Runtime& rt){
  assert(&(rt.executor()) == &executor);
});
executor.run(taskflow).wait();

void tf::Runtime::schedule(Task task)

schedules an active task immediately to the worker's queue

This member function immediately schedules an active task to the task queue of the associated worker in the runtime task. An active task is a task in a running taskflow. The task may or may not be running, and scheduling that task will immediately put the task into the task queue of the worker that is running the runtime task. Consider the following example:

tf::Task A, B, C, D;
std::tie(A, B, C, D) = taskflow.emplace(
  [] () { return 0; },
  [&C] (tf::Runtime& rt) {  // C must be captured by reference
    std::cout << "B\n";
    rt.schedule(C);
  },
  [] () { std::cout << "C\n"; },
  [] () { std::cout << "D\n"; }
);
A.precede(B, C, D);
executor.run(taskflow).wait();

The executor will first run the condition task A which returns 0 to inform the scheduler to go to the runtime task B. During the execution of B, it directly schedules task C without going through the normal taskflow graph scheduling process. At this moment, task C is active because its parent taskflow is running. When the taskflow finishes, we will see both B and C in the output.

template<typename F>

auto tf::Runtime::async(F&& f)

runs the given callable asynchronously

The method creates an asynchronous task to launch the given function on the given arguments. The difference to tf::Executor::async is that the created asynchronous task pertains to the runtime object. Applications can explicitly issue tf::Runtime::corun_all to wait for all spawned asynchronous tasks to finish. For example:

std::atomic<int> counter(0);
taskflow.emplace([&](tf::Runtime& rt){
  auto fu1 = rt.async([&](){ counter++; });
  auto fu2 = rt.async([&](){ counter++; });
  fu1.get();
  fu2.get();
  assert(counter == 2);
  
  // spawn 100 asynchronous tasks from the worker of the runtime
  for(int i=0; i<100; i++) {
    rt.async([&](){ counter++; });
  }
  
  // wait for the 100 asynchronous tasks to finish
  rt.corun_all();
  assert(counter == 102);
});

This method is thread-safe and can be called by multiple workers that hold the reference to the runtime. For example, the code below spawns 100 tasks from the worker of a runtime, and each of the 100 tasks spawns another task that will be run by another worker.

std::atomic<int> counter(0);
taskflow.emplace([&](tf::Runtime& rt){
  // worker of the runtime spawns 100 tasks each spawning another task
  // that will be run by another worker
  for(int i=0; i<100; i++) {
    rt.async([&](){ 
      counter++; 
      rt.async([](){ counter++; });
    });
  }
  
  // wait for the 200 asynchronous tasks to finish
  rt.corun_all();
  assert(counter == 200);
});

template<typename P, typename F>

auto tf::Runtime::async(P&& params, F&& f)

runs the given callable asynchronously

taskflow.emplace([&](tf::Runtime& rt){
  auto future = rt.async("my task", [](){});
  future.get();
});

template<typename F>

void tf::Runtime::silent_async(F&& f)

runs the given function asynchronously without returning any future object

This member function is more efficient than tf::Runtime::async and is encouraged to use when there is no data returned.

std::atomic<int> counter(0);
taskflow.emplace([&](tf::Runtime& rt){
  for(int i=0; i<100; i++) {
    rt.silent_async([&](){ counter++; });
  }
  rt.corun_all();
  assert(counter == 100);
});

This member function is thread-safe.

template<typename P, typename F>

void tf::Runtime::silent_async(P&& params, F&& f)

runs the given function asynchronously without returning any future object

taskflow.emplace([&](tf::Runtime& rt){
  rt.silent_async("my task", [](){});
  rt.corun_all();
});

template<typename F>

void tf::Runtime::silent_async_unchecked(F&& f)

similar to tf::Runtime::silent_async but the caller must be the worker of the runtime

The method bypass the check of the caller worker from the executor and thus can only called by the worker of this runtime.

taskflow.emplace([&](tf::Runtime& rt){
  // running by the worker of this runtime
  rt.silent_async_unchecked([](){});
  rt.corun_all();
});

template<typename P, typename F>

void tf::Runtime::silent_async_unchecked(P&& params, F&& f)

similar to tf::Runtime::silent_async but the caller must be the worker of the runtime

The method bypass the check of the caller worker from the executor and thus can only called by the worker of this runtime.

taskflow.emplace([&](tf::Runtime& rt){
  // running by the worker of this runtime
  rt.silent_async_unchecked("my task", [](){});
  rt.corun_all();
});

template<typename T>

void tf::Runtime::corun(T&& target)

co-runs the given target and waits until it completes

A target can be one of the following forms:

• a subflow task to spawn a subflow or
• a composable graph object with tf::Graph& T::graph() defined

// co-run a subflow and wait until all tasks complete
taskflow.emplace([](tf::Runtime& rt){
  rt.corun([](tf::Subflow& sf){
    tf::Task A = sf.emplace([](){});
    tf::Task B = sf.emplace([](){});
  }); 
});

// co-run a taskflow and wait until all tasks complete
tf::Taskflow taskflow1, taskflow2;
taskflow1.emplace([](){ std::cout << "running taskflow1\n"; });
taskflow2.emplace([&](tf::Runtime& rt){
  std::cout << "running taskflow2\n";
  rt.corun(taskflow1);
});
executor.run(taskflow2).wait();

Although tf::Runtime::corun blocks until the operation completes, the caller thread (worker) is not blocked (e.g., sleeping or holding any lock). Instead, the caller thread joins the work-stealing loop of the executor and returns when all tasks in the target completes.

template<typename P>

void tf::Runtime::corun_until(P&& predicate)

keeps running the work-stealing loop until the predicate becomes true

The method keeps the caller worker running in the work-stealing loop until the stop predicate becomes true.

void tf::Runtime::corun_all()

corun all asynchronous tasks spawned by this runtime with other workers

Coruns all asynchronous tasks (tf::Runtime::async, tf::Runtime::silent_async) with other workers until all those asynchronous tasks finish.

std::atomic<size_t> counter{0};
taskflow.emplace([&](tf::Runtime& rt){
  // spawn 100 async tasks and wait
  for(int i=0; i<100; i++) {
    rt.silent_async([&](){ counter++; });
  }
  rt.corun_all();
  assert(counter == 100);
  
  // spawn another 100 async tasks and wait
  for(int i=0; i<100; i++) {
    rt.silent_async([&](){ counter++; });
  }
  rt.corun_all();
  assert(counter == 200);
});