fork(S)                        6 January 1993                        fork(S)


 Name

    fork - create a new process

 Syntax


    cc  . . .  -lc


    int fork ()


 Description

    The fork system call creates a new process.  The new process (child pro-
    cess) is an exact copy of the calling process (parent process).  The
    child process inherits the following attributes from the parent process:

    +  environment: see putenv(S), getenv(S), environ(S), exec(S)

    +  close-on-exec flag: see fcntl(S), exec(S)

    +  signal-handling settings (that is, SIGDFL, SIGIGN, SIGHOLD, func-
       tion address): see sigaction(S), signal(S)

    +  signal mask: see sigprocmask(S)

    +  profiling on/off status: see profil(S)

    +  nice value: see nice(S)

    +  all attached shared memory segments: see shmop(S)

    +  process group ID: see setpgid(S), setsid(S), setpgrp(S), getpgid(S),
       getpgrp(S)

    +  tty group ID: see tcsetpgrp(S), tcgetpgrp(S), exit(S)

    +  current working directory: see chdir(S), getcwd(S)

    +  root directory: see chroot(S)

    +  file mode creation mask: see umask(S)

    +  file size limit: see ulimit(S)

    +  real, effective, and saved user IDs: see setuid(S), getuid(S),
       geteuid(S), exec(S)

    +  real, effective and saved group IDs: see setgid(S), getgid(S),
       getegid(S), exec(S)

    +  login user ID: see setluid(S), getluid(S)

    +  supplemental groups: see setgroups(S), getgroups(S)

    +  controlling terminal

    The child process differs from the parent process in the following ways:

    +  The child process has a unique process ID: see getpid(S)

    +  The child process has a different parent process ID: see getppid(S)

    +  The child process has its own copy of the parent's file descriptors.
       Each of the child's file descriptors shares a common file pointer with
       the corresponding file descriptor of the parent.

    +  No signals are pending in the child: see sigpending(S)

    +  All semadj values are cleared: see semop(S)

    +  Process locks, text locks, and data locks are not inherited by the
       child: see plock(S), lock(S)

    +  The child process' tms structure is cleared (setting utime, stime,
       cutime, and cstime to 0): see times(S)

    +  The time left until an alarm clock or itimer signal is reset to 0: see
       alarm(S), setitimer(S), getitimer(S)

    +  Record locks set by the parent are not inherited by the child: see
       fcntl(S), locking(S), lockf(S)

    The fork system call fails and no child process is created if one or more
    of the following is true:


    [EAGAIN]


                +  The system-imposed limit on the total number of processes
                   under execution would be exceeded.

                +  The system-imposed limit on the total number of processes
                   under execution by a single user would be exceeded.

                +  The total amount of system memory available when reading
                   via raw I/O is temporarily insufficient.

    [EMFILE]    The shared data table would overflow.

    [ENOMEM]    The process requires more space than the system can supply.


 See also

    alarm(S), exec(S), getitimer(S), nice(S), plock(S), ptrace(S), semop(S),
    setgid(S), setgroups(S), setluid(S), setuid(S), shmop(S), signal(S),
    sigset(S), times(S), ulimit(S), umask(S), wait(S)

 Diagnostics

    Upon successful completion, fork returns a value of 0 to the child pro-
    cess and returns the process ID of the child process to the parent pro-
    cess.  Otherwise, a value of -1 is returned to the parent process; no
    child process is created, and errno is set to indicate the error.

 Standards conformance

    fork is conformant with:
    AT&T SVID Issue 2;
    X/Open Portability Guide, Issue 3, 1989;
    IEEE POSIX Std 1003.1-1990 System Application Program Interface (API) [C
    Language] (ISO/IEC 9945-1);
    and NIST FIPS 151-1.