Re: fork()/exec() dumping

["Michael J. Hammel" <mjhammel graphics-muse org>]

Thus spoke Erik Mouw
> > using the regular fork() and exec() to execute the other application
> 
> This is becoming a FAQ. Never mind, here we go again:
>       /* child code */
>       system("nxterm -T 'Alloc' -n KMan -e man /usr/man/man3/Alloc.3");
>       _exit(0); /* note the use of _exit() instead of exit();

But this is different than fork'ing and exec'ing.  system() returns to the
child and leaves exiting a responsibility of the child.  exec'ing never
returns.  I haven't tried this yet, but according to the original post the
problem was with exec'ing another program. There shouldn't be any problems
with exec'ing, should there?

>        /* parent code: just wait for the child to exit */
>        /* you can also install a signal handler for SIGCHLD */
>        waitpid(pid, &status, WUNTRACED); /* anti zombie */

You don't want to wait on the child if you exec().  You should do a SIG_IGN
for children that call exec() or you'll end up with zombies.
-- 
Michael J. Hammel           |    We need somebody who can work 
The Graphics Muse           |    independently and innovatively in the
mjhammel@graphics-muse.org  |    face of unreasonable demands."
http://www.graphics-muse.org        Job posting at Stanford CS Department