add retry option to supervise-daemon

The --retry option for supervise-daemon defines how the supervisor will
attempt to stop the child process it is monitoring. It is defined when
the supervisor is started since stopping the supervisor just sends a
signal to the active supervisor.

This fixes #160.

sh/supervise-daemon.sh

@@ -23,6 +23,7 @@ supervise_start()
 	# command_args="this \"is a\" test"
 	# to work properly.
 	eval supervise-daemon --start \
+		${retry:+--retry} $retry \
 		${chroot:+--chroot} $chroot \
 		${pidfile:+--pidfile} $pidfile \
 		${respawn_delay:+--respawn-delay} $respawn_delay \