I've had reports of rsyncd leaving zombies under digital unix. This
authorAndrew Tridgell <tridge@samba.org>
Wed, 1 Jul 1998 05:10:42 +0000 (05:10 +0000)
committerAndrew Tridgell <tridge@samba.org>
Wed, 1 Jul 1998 05:10:42 +0000 (05:10 +0000)
commit31f440e68b1a087093f1c34be1ad3444f911d7d9
tree8e170ed71bfed70aef412c3af9d7dbf469f3b5c9
parentc95da96a0c51c66c8cb2eff97b768a717d9e0c79
I've had reports of rsyncd leaving zombies under digital unix. This
patch tries to address the problem in two ways:

1) reinstall the SIGCHLD handler before each fork
2) reap any children not caught by the handler using waitpid with
WNOHANG.

I expect this will fix the problem.
socket.c