Thread Creation

Thread creation is easy. Call  std::thread, and a new thread will be created. The thread gets a work package and starts it immediately. The creator of the thread (the Parent) has to take care of the created thread (the child). The parent should wait until their child is done with its task or has to detach itself from its child. The child thread can get its payload task arguments by copy or by reference.

To be honest, that was too fast. So the details will follow.

 

Modernes C++ Mentoring

Stay informed about my mentoring programs. Subscribe for the news.

• • • • • • • "Fundamentals for C++ Professionals" (open)
• • • • • • • "Design Patterns and Architectural Patterns with C++" (starts in February 2023)

 

Thank you! Please check your email and confirm the newsletter subscription.

Your subscription was updated

Email *

It seems that you have already subscribed to this list. Click here to update your profile.

Subscribe

Creation and execution of a thread

Now, a more formal approach: a thread gets a Callable and starts it immediately.

This sentence needs a few notes.

• A Callable is an entity that behaves like a function. It can be a function, a function object, or a lambda function.
• A function object is an instance of a class, for which the call operator is overloaded. The key difference between functions and function objects is, that a function object can have a state.
• A lambda function (anonymous function) is a pure function body, without a name. It can be invoked just in place. A lambda function can capture its calling context. That's why they are often called closures.

After the theory a small example.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39

// createThread.cpp

#include <iostream>
#include <thread>

void helloFunction(){
  std::cout << "Hello C++11 from function." << std::endl;
}

class HelloFunctionObject  {
  public:
    void operator()() const {
      std::cout << "Hello C++11 from a function object." << std::endl;
    }
};


int main(){

  std::cout << std::endl;

  // thread executing helloFunction
  std::thread t1(helloFunction);

  // thread executing helloFunctionObject
  HelloFunctionObject helloFunctionObject;
  std::thread t2(helloFunctionObject);

  // thread executing lambda function
  std::thread t3([]{std::cout << "Hello C++11 from lambda function." << std::endl;});

  // ensure that t1, t2 and t3 have finished before main terminates
  t1.join();
  t2.join();
  t3.join();

  std::cout << std::endl;

};

All threads -  t1, t2, and t3 - write their messages to the console. The work package of thread t2 is a function object (line 10 - 15), the work package of thread t3 is a lambda function (line 29). In lines 32 - 34 the Main thread or Parent waits, until his children are done.

Let's have a look at the output. This is more interesting.

The two programs execution results differ in two aspects. First, child threads will be executed in a different order. Second, the output is a little bit of a mess. So, in the second run, the line break of the function helloFunction happens after the lambda function call.

What's next?

The next article will be about the lifetime of a thread. (Proofreader Alexey Elymanov)