*BSD News Article 35698


Return to BSD News archive

Path: sserve!newshost.anu.edu.au!harbinger.cc.monash.edu.au!msuinfo!agate!howland.reston.ans.net!swrinde!news.uh.edu!uuneo.neosoft.com!Starbase.NeoSoft.COM!georgel
From: georgel@starbase.neosoft.com (George Livsey)
Newsgroups: comp.os.386bsd.development
Subject: Configuring Internal Modem on COM1
Date: 11 Sep 1994 17:39:31 GMT
Organization: NeoSoft Internet Services   +1 713 684 5969
Lines: 35
Message-ID: <34vfcj$hbp@uuneo.neosoft.com>
NNTP-Posting-Host: starbase.neosoft.com
X-Newsreader: TIN [version 1.2 PL2]



     I am having trouble with an internal modem on COM1.  The modem has a 
16250/16450 UART.  It is a flakey little 14.4 Kbps modem.  I was told 
repeatedly that I couldn't use an internal modem under SCO Unix and 
expect it to perform well.  That was wrong!  It was a real pain to 
configure but it did work.  The source of the problem with this one is 
that if the bus is probed by sending out hard interrupts, the silly thing 
not only doesn't respond; it locks up.  I had the same problem with comm 
programs under DOS.  The bios finds the modem and as long as the thing is 
accessed through the bios, there is no problem at all.  
 
    Well, now I'm having the same sort of problem under FreeBSD 1.1.5.1 
and I suspect that there must be a workaround in this case also.  I know 
-- I should get a modem with a true 16550 UART ... and I will soon, but 
in the meanwhile....   
 
    I recompiled the source and saw that the 
/usr/src/sys/i386/conf/(systemname) file had an entry for sio0 which probes 
the isa bus with interrupt 4.  When I boot, the kernel complains that it 
can't find a device at 0x3f8.  When I then access the modem under DOS the 
modem complains that its onboard vram is not functioning.  Do a hard boot 
and the problem is resolved.  I guess that the solution to the problem is 
in giving the kernel an exact address for the modem, bypassing the hard 
interrupt probe, and thereby not locking the modem.  I can't seem to find 
anything anywhere that details the format of the sio0 entry.  I surmise 
that the entry `at isa?` is probing the at bus but I don't know how to 
bypass this hard interrupt.  If anyone knows,  I would greatly appreciate 
any answers.
 
George Livsey
georgel@starbase.neosoft.com