From owner-freebsd-current  Sun Mar 10 22:52:02 1996
Return-Path: owner-current
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id WAA09088
          for current-outgoing; Sun, 10 Mar 1996 22:52:02 -0800 (PST)
Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19])
          by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id WAA09048
          for <current@freebsd.org>; Sun, 10 Mar 1996 22:51:55 -0800 (PST)
Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.12/8.6.9) id RAA07820; Mon, 11 Mar 1996 17:47:26 +1100
Date: Mon, 11 Mar 1996 17:47:26 +1100
From: Bruce Evans <bde@zeta.org.au>
Message-Id: <199603110647.RAA07820@godzilla.zeta.org.au>
To: current@freebsd.org, jmz@cabri.obs-besancon.fr
Subject: Re: COMCONSOLE: a patch and a problem
Sender: owner-current@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

>I played with the boot blocks and I noticed that I was unable to boot
>from a serial console (keyboard unplugged). The patch below fixes a
>bug in boot.c/probe_keyboard.c (well, at least on my machine:-).

I don't think I broke this :-).  I wonder when it last worked.

>The problem now: when booting from a serial console, there is no
>timeout at the 'Boot:' prompt. Further investigations showed that the
>bios_tick function always returns the same value! Is this related to
>the fact that the keyboard is not plugged in? 

I broke this.  It's because there are no BIOS calls which have the
side affect of allowing clock interrupts.  Add some dummy BIOS calls.
Putting them in the bios_tick macro would take too much space, so
arrange to always call ischar() and throw away the result in the
RB_SERIAL case.

Bruce