possible sigchld bug

[Mikulas Patocka]

> signals are not blocked during execution of handlers?

On BSD and Linux glibc-2 yes. On SysV or Linux libc-5 no. (on these
systems, signal handler is restored to default before entering handler
function).

You should use sigaction() instead of signal() if you want consistent
behaviour on all systems.

Mikulas

> On Mon, Dec 22, 2003 at 08:26:28PM +0100, Mikulas Patocka wrote:
> > Hi
> >
> > What if you have sysv signals (i.e. signal is restored when handler is
> > called) and child process exits here? Zombie will be left, because SIGCHLD
> > is ignored at that point. Shouldn't signal be before waitpid?
> >
> > Mikulas
> >
> > static void
> > main_sigchld_handler(int sig)
> > {
> >         int save_errno = errno;
> >         pid_t pid;
> >         int status;
> >
> >         while ((pid = waitpid(-1, &status, WNOHANG)) > 0 ||
> >             (pid < 0 && errno == EINTR))
> >                 ;
> >
> > 		<----- HERE
> >
> >         signal(SIGCHLD, main_sigchld_handler);
> >         errno = save_errno;
> > }
> >
> > _______________________________________________
> > openssh-unix-dev mailing list
> > openssh-unix-dev at mindrot.org
> > http://www.mindrot.org/mailman/listinfo/openssh-unix-dev
>