Process Duplication

Process Duplication
===================
- A new process is created using fork().
- fork() is a system call.
- fork duplicates the current process.
- It creates a new entry in the process table with many of same attributes as current process.
- The new process (known as child process) is almost identical to original process (known as parent process).
- PPID (Parent Process Identification) of new process is PID of original process.
- PID (Process Identification) of new process is lowest PID available after the PID of original process .Mostly it is 1 greater than the original process(If available).
- Prototype of fork :-
    pid_t fork(void);
- Function fork() is declared in header:
    #include<unistd.h>
- It has a return type pid_t .
- System type pid_t is declared in header :
    #include<sys/types.h>
- The Process counter of new process starts from fork() statement.
- So, fork() is executed twice.
- When fork() is executed by parent, it returns PID of child process on success.
- When fork() is executed by child, it returns 0 for success.
- If fork fails, it returns -1.
- Falure may be due to :-
    * Limit on the number of child processes that a parent may have (CHILD_MAX), in which errno will be set to "EAGAIN"

    * If process table or virtual memory is filled (not enough space), errno variable will be set to "ENOMEM".
- The return value of fork() can be used to determine whether child or parent is executing.
- For eg. :-
#include<stdio.h>
#include<unistd.h>
#include<sys/types.h>
int main()
{
    pid_t fret;
    printf("Only parent execute\n");
    fret = fork();//Fork statement
    printf("Both parent and child execute this statement.\n");
    switch(fork)
    {
        case -1 :
            perror("fork");
            exit(EXIT_FAILURE);
            break;
        case 0 :
            printf("Child executing\n");
            break;
        default :
            printf("Parent executing\n");
            break;
    }
    return 0;
}