(OpenMP) How to start each function 1 time in 1 thread

Question

I have 4 threads and I want to start in one time 3 functions. Each function takes one thread.
But this code is starting every function 4 times

#pragma omp parallel
    {
        Func1();
        Func2();
        Func3();
    }

and I have this result:

Func*1* started

Func*1* finished

Func*2* started

Func*2* finished

...

How should I change the code to display some like this:

Func*1* started

Func*2* started

Func*3* started

Func*2* finished

Func*1* finished

Func*3* finished

You learn how parallelism works. You're creating subprocesses that call func1, func2, and func3, and then starting them all at effectively the same time. Each process is going to run func1 first, as soon as it can. If that's not what you want, then stop trying to disguise one process as four. — cHao, Oct 14 '12 at 16:25
K, i misread. Problem is, you have three functions doing three different things. That's generally not what OMP does. It's for massively parallel stuff, where you have a bunch of subprocesses doing the same thing. — cHao, Oct 14 '12 at 16:32
Does this answer your question? [Execute functions in parallel using openmp](https://stackoverflow.com/questions/25431821/execute-functions-in-parallel-using-openmp) — user202729, Jan 22 '21 at 08:51

score 5 · Accepted Answer · answered Oct 14 '12 at 16:32

5

What you are searching for is the sections worksharing construct. The syntax should be the following:

#pragma omp parallel
{
#pragma omp sections
  {
#pragma omp section
    Func1();
#pragma omp section
    Func2();
#pragma omp section
    Func3();
  }
}

I suggest you to refer to the specifications and the examples therein for more information on how this construct works.

answered Oct 14 '12 at 16:32

Massimiliano

7,842
2
47
62

One could also use the combined `parallel sections` construct for brevity. – Hristo Iliev Oct 14 '12 at 20:50

Hristo Iliev · Answer 2 · 2012-10-15T08:29:51.047

As already answered by Massimiliano, the easiest way to achieve what you want is to use the sections construct:

#pragma omp parallel sections
{
    #pragma omp section
    Func1();
    #pragma omp section
    Func2();
    #pragma omp section
    Func3();
}

(when the sections construct is the only thing nested inside a parallel region, both pragmas can be combined as shown)

Each section scopes the statement or block immediately following it. E.g. if you'd like to include some more code in a section, you should put it inside a block:

#pragma omp section
{
    Func1();
    SomeOtherFunc1();
}

A more general way to control what each thread does is to examine the thread ID as returned by omp_get_thread_num() and to branch in accordance. This resembles the way MPI programs are written:

#pragma omp parallel
{
    switch (omp_get_thread_num())
    {
        case 0:
            Func1();
            break;
        case 1:
            Func2();
            break;
        case 2:
            Func3();
            break;
        default:
            // Do nothing
     }
}

One obvious disadvantage is that it would only work unless there are at least 3 threads in the team that executes the parallel region, i.e. code that relies on thread IDs becomes somewhat non-portable. In contrast, the sections directive works with any number of threads, e.g. if there is only one thread, all sections would be executed sequentially.

(OpenMP) How to start each function 1 time in 1 thread

2 Answers2