Debugging messages to trace startnew failures

    my $pid2;

    if(not defined $child) {

        logmsg "fork() failure detected\n";

        logmsg "startnew: fork() failure detected\n";

        return (-1,-1);

    }

        die "error: exec() has returned";

    }

    # Ugly hack but ssh doesn't support pid files

    if ($fake) {

        logmsg "$pidfile faked with pid=$child\n" if($verbose);

        open(OUT, ">$pidfile");

        if(open(OUT, ">$pidfile")) {

            print OUT $child . "\n";

            close(OUT);

            logmsg "startnew: $pidfile faked with pid=$child\n" if($verbose);

        }

        else {

            logmsg "startnew: failed to write fake $pidfile with pid=$child\n";

        }

        # could/should do a while connect fails sleep a bit and loop

        sleep 5;

        if (checkdied($child)) {

            logmsg "startnew: Warning: child process has failed to start\n" if($verbose);

            logmsg "startnew: child process has failed to start\n";

            return (-1,-1);

        }

    }

    my $count=12;

    while($count--) {

        if(-f $pidfile) {

            open(PID, "<$pidfile");

            $pid2 = 0 + <PID>;

            close(PID);

            if($pid2 && kill(0, $pid2)) {

            if(($pid2 > 0) && kill(0, $pid2)) {

                # if $pid2 is valid, then make sure this pid is alive, as

                # otherwise it is just likely to be the _previous_ pidfile or

                # similar!

            }

        }

        if (checkdied($child)) {

            logmsg "startnew: Warning: child process has died\n" if($verbose);

            logmsg "startnew: child process has died, but server might start up\n";

            # We can't just abort waiting for the server with a

            # return (-1,-1);

            # because the server might have forked and could still start