Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sun-barr!apple!bionet!uwm.edu!spool.mu.edu!sdd.hp.com!uakari.primate.wisc.edu!aplcen!boingo.med.jhu.edu!haven!mimsy!nocusuhs!nmrdc1!minixug!uwalt!waltje
From: waltje@uwalt.nl.mugnet.org (Fred 'The Rebel' van Kempen)
Newsgroups: comp.os.minix
Subject: Re: modems and carrier detect under minix 1.5.10 IBM/XT
Message-ID: <9102287337@uwalt.nl.mugnet.org>
Date: 28 Feb 91 15:52:38 GMT
References: <62684@bbn.BBN.COM>
Organization: MicroWalt Corporation, for MINIX Software Development
Lines: 61

In article <62684@bbn.BBN.COM>, grossman@bbn.com (Martin Grossman) wrote:

> I've been fiddling with getting my modem on COM1 to work properly!!!
Humm.  It DOES work, but... :

> 1) Minix doesn't examine the DCD line from the modem
Correct.  That is: _standard_ MINIX doesn't.  Our ("NLMUG") does.

> 	So I added some code in getty to do a port_in(0x03f8+6,&val)
> 	and then examine bit 7.  Now I can see DCD.
Yuck!  This won't work on 80286,386,486 and 680X0 machines !
It should be in the RS-232 driver in the kernel!

> 2) Minix does enable all interrupts from the UART
> 	When the DCD line changes state it causes an interrupt
> 	that minix gets real fast and doesn't do anything with it,
> 	thus I can never see the other bit (I think its bit 3)
> 	that states that DCD changed states.
No, bit 7 (3 is the DELTA DCD bit).  I enabled this a long time ago,
and, upon the request of many, I just finished the cleanup of my
RS232 (and TTY) driver of NLMUG MINIX.  I sent it to Bruce Evans
for checking, and I will post the driver as soon as I get it back.

> 3) I've hacked up getty and login as follows:
> 	a) added some code before call to do_carier() to:
> 		set -ECHO & RAW
> 		output at\r
> 		delay
> 		output +++
> 		delay
> 		output ath\r
Brrr...
You should use the DTR line to put the modem in OFFLINE state (by
dropping the DTR line).  This is usually done by setting the line's
speed to 0.  I implemented this a long time ago.  See above.

> 	b) defined CHOISE_A to examine DCD via doing a port_in()
		       ^ C ?				^^^^^^^
						Won't work on protected-mode
						CPU's !

> 	c) added timeouts (via sig #14) to correctly timeout both getty
> 	   and login if idle for too long.
Already done.

To all people raeding this: what version is currently being used by you?
I am using version 3.17 here, and I will post it if many people are using
an old version...

> 4) TODO
> 	fix the kernel to
> 	a) do carrier correctly
> 	b) send signal to process if its not getty and DCD went down.
Already done, see above.
You also need to modify MM to do (b) correctly.

Cheers-
	The Rebel
--
MicroWalt Corporation, for MINIX Development	waltje@uwalt.nl.mugnet.org
Tel (+31) 252 230 205, Hoefbladhof  27, 2215 DV  VOORHOUT, The Netherlands