Question

I’m new to programming with linux and I was trying to understand how fork() and exec functions work. To make it easier for myself, I created a simply Dummy executable (with gcc -o Dummy.exe …) and tried to call fork function so I can replace the child with the Dummy.exe executable I have created.

The problem I’m coming accross is that when I run the code, it shows me the contents of the Dummy.exe, however, I don’t see anything past that – meaning, I don’t see the parent process ending.

When I run my code Ex1.cpp, I get the output:

    Program is Running
    --- ****** ---
    Ended

    me@mdev>

The only way I can get the program to end is by pressing return key – you will see a blank line after word Ended.

Here is the code in my Ex1.cpp

#include <iostream>
using namespace std;

int main()
{
        pid_t retVal;
        int newStatus;

        switch(retVal = fork())
        {
                case -1:
                cout<<"Error occured with fork"<<endl;
                break;
                case 0:
                cout<<"Child forked"<<endl;
                newStatus = execl("Dummy.exe","Dummy.exe", NULL);
                break;
                default:
                cout<<"Parent has a new child: "<<retVal<<endl;
        }

        cout<<"Ended ..."<<endl;
        return 0;
}

My Dummy.cpp code is below:

#include <iostream>
using namespace std;

int main()
{
        cout<<"Program is Running"<<endl;
        cout<<"--- ****** ---"<<endl;
        cout<<"Ended"<<endl;

        return 0;
}

My background is Windows development and all this is new for me – I appreciate your help.

Answer 1

#include <iostream>

using namespace std;

int main ()
{
pid_t retVal;

    int newStatus;

    int returnStatus;

    if (retVal = fork() < 0)
    {
        cout << "Error on fork" << endl;
        exit (1);
    }   

    // Child process
    else if (retVal == 0)
    {
        // No code is executed after an exec statement
        newStatus = execl("./Dummy", "Dummy", NULL);
    }


    else
    {
        // This waits for the child to finish and clean up
        waitpid(retVal, &returnStatus, NULL);
    }

return returnStatus;

}

It is not a good idea to use a switch statement in your case since no code is executed after an exec statement. You can read up on that further by reading the man page at the command prompt by typing man 2 exec. Also it is always a good idea to wait for the child to exit successfully which is what the waitpid function does. Try this and let me know if it works and if you have any questions as well.