From owner-freebsd-hackers  Sat Jul  1 08:14:06 1995
Return-Path: hackers-owner
Received: (from majordom@localhost)
          by freefall.cdrom.com (8.6.10/8.6.6) id IAA22130
          for hackers-outgoing; Sat, 1 Jul 1995 08:14:06 -0700
Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.34])
          by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id IAA22124
          for <hackers@freebsd.org>; Sat, 1 Jul 1995 08:13:57 -0700
Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.9/8.6.9) id BAA13954; Sun, 2 Jul 1995 01:09:37 +1000
Date: Sun, 2 Jul 1995 01:09:37 +1000
From: Bruce Evans <bde@zeta.org.au>
Message-Id: <199507011509.BAA13954@godzilla.zeta.org.au>
To: hackers@freebsd.org, kwong@fathergoose.net6c.io.org
Subject: Re: sio driver problem
Sender: hackers-owner@freebsd.org
Precedence: bulk

>1)
>I am not sure if someone is already discover this, found a bug in the
>serial card driver. in the sioopen function ( line 802 of sio.c ) a line
>	if (!com->active_out) {
>it should be
>	if (com->active_out) {

	if (!com->active_out) {
is correct.  The test succeeds if the callout port is not open (and the
tty is open), i.e., when the callin port is open.

>2)
>this is about the callin and callout problem: a getty is running on 
>ttyd1. a call is make to a ppp provider then do a pppd /dev/cuaa1 57600
>but the pppd complained:
>pppd[187]:ioctl(TIOCSCTTY) operation not permitted
>again this only happen when getty is running on same port. otherwise
>It works fine.

Apparently the getty has opened ttyd1 and made it a controlling
terminal.  This is supposed to stop the open of cuaa1 from succeeding,
but if you invert the test in (1) then the open works and something
breaks a little later.

ttyd1 should only be used for incoming calls.

Bruce