Interlude: Process API

Prev: the-abstraction-the-process Next: mechanism-limited-direct-execution

The fork() System Call

The Fork system call is for creating a new process. It creates a copy of its parent and then gets its own address space.

#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
 
int main(int argc, char *argv[]) {
  printf("hello world (pid:%d)\n", (int) getpid());
  int rc = fork();
  if (rc < 0) { // fork failed; exit
    fprintf(stderr, "fork failed\n");
    exit(1);
  } else if (rc == 0) { // child (new process)
    printf("hello, I am child (pid:%d)\n", (int) getpid());
  } else { // parent goes down this path (main)
    printf("hello, I am parent of %d (pid:%d)\n",
        rc, (int) getpid());
  }
  return 0;
}

This prints out:

prompt> ./p1
hello world (pid:29146)
hello, I am parent of 29147 (pid:29146)
hello, I am child (pid:29147)
prompt>

When calling fork, the parent gets the pid of the newly-created child, whereas the child receives a return code of 0, which allows you to handle the parent and child’s path.

The wait() System Call

What happens if we want either the parent or child to execute first? Add a wait() system call.

#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#include <sys/wait.h>
 
int main(int argc, char *argv[]) {
  printf("hello world (pid:%d)\n", (int) getpid());
  int rc = fork();
  if (rc < 0) { // fork failed; exit
    fprintf(stderr, "fork failed\n");
    exit(1);
  } else if (rc == 0) { // child (new process)
    printf("hello, I am child (pid:%d)\n", (int) getpid());
  } else { // parent goes down this path (main)
  int rc_wait = wait(NULL);
  printf("hello, I am parent of %d (rc_wait:%d) (pid:%d)\n",
  rc, rc_wait, (int) getpid());
  }
  return 0;
}

The exec() System Call

You can use exec() to run a different program than the parent process.

include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#include <string.h>
#include <sys/wait.h>
 
int main(int argc, char *argv[]) {
  printf("hello world (pid:%d)\n", (int) getpid());
  int rc = fork();
  if (rc < 0) { // fork failed; exit
    fprintf(stderr, "fork failed\n");
    exit(1);
  } else if (rc == 0) { // child (new process)
    printf("hello, I am child (pid:%d)\n", (int) getpid());
    char *myargs[3];
    myargs[0] = strdup("wc"); // program: "wc" (word count)
    myargs[1] = strdup("p3.c"); // argument: file to count
    myargs[2] = NULL; // marks end of array
    execvp(myargs[0], myargs); // runs word count
    printf("this shouldn’t print out");
  } else { // parent goes down this path (main)
    int rc_wait = wait(NULL);
    printf("hello, I am parent of %d (rc_wait:%d) (pid:%d)\n",
  rc, rc_wait, (int) getpid());
  }
  return 0;
}

This runs wc in an odd way. Exec loads code and static data from that executable and overwrites its current code segment with that data. Then it runs that process, transforming itself into that program.

Why? Motivating the API

With these design decisions, we can easily design a shell.

By separating fork and exec, you can easily create a shell that handles this command:

wc p3.c > newfile.txt

This is done by redirecting the output of the command wc p3.c to newfile.txt by closing stdout and opening newfile.txt.

#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#include <string.h>
#include <fcntl.h>
#include <sys/wait.h>
 
int main(int argc, char *argv[]) {
  int rc = fork();
  if (rc < 0) { // fork failed; exit
   fprintf(stderr, "fork failed\n");
   exit(1);
  } else if (rc == 0) { // child: redirect standard output to a file
   close(STDOUT_FILENO);
   open("./p4.output", O_CREAT|O_WRONLY|O_TRUNC, S_IRWXU);
 
   // now exec "wc"...
   char *myargs[3];
   myargs[0] = strdup("wc"); // program: "wc" (word count)
   myargs[1] = strdup("p4.c"); // argument: file to count
   myargs[2] = NULL; // marks end of array
   execvp(myargs[0], myargs); // runs word count
  } else { // parent goes down this path (main)
   int rc_wait = wait(NULL);
  }
  return 0;
}

Process Control and Users

Besides fork, exec, and wait, there are other useful commands, like kill, which sends signals to a process. Ctrl-C sends a SIGINT (interrupt) and Ctrl-Z sends a SIGTSTP (stop) command, pausing the process to be rerun later with fg.

A process can use the signal() system call to catch various signals, and react to those (e.g. cleanup on Ctrl-C).

Because having a process receive signals from multiple users would run into concurrency issues, we limit the users who can send signals (e.g. only root and the current user).

Useful Tools

For example ps shows which processes are running. top says which processes are using the most CPU and memory. kill and killall can send signals to processes.

Prev: the-abstraction-the-process Next: mechanism-limited-direct-execution