*BSD News Article 30508


Return to BSD News archive

Newsgroups: comp.os.386bsd.questions
Path: sserve!newshost.anu.edu.au!harbinger.cc.monash.edu.au!msuinfo!netnews.upenn.edu!dsinc!spool.mu.edu!howland.reston.ans.net!vixen.cso.uiuc.edu!sdd.hp.com!hpscit.sc.hp.com!news.dtc.hp.com!col.hp.com!csn!carbon!mercury.cair.du.edu!troby
From: troby@du.edu (Thorn Roby)
Subject: Multiport sio without "master port" interrupt? (1.1 RELEASE)
Message-ID: <Cq0DBI.Lp@du.edu>
Sender: news@du.edu (USENET News System)
Organization: University of Denver
Date: Wed, 18 May 1994 17:16:29 GMT
Lines: 24

I posted this query a few weeks ago, asking specifically about the
BOCA IOAT66 card. I suspect the problem may apply to other multiport
cards without a designated "master port" as well. Does anyone know the
trick to getting these to work? I've compiled a MULTIPORT but
non-BIDIRECTIONAL kernel (these ports are for outgoing fax modems, no
login required). All ports are recognized at boot time, and they do
function on both transmit and receive, but operations seem to be
delayed - for example, if I send one AT, I get no response, but if I
send it again I get "OK". There seems to be no loss of characters 
within a given transmission, and the ports work fine if set up as
standard non-multiport sio0/sio1. This card does have an I/O address assigned
to a shared interrupt, but I don't think it corresponds to an actual
I/O port address. Any hints appreciated. If it's relevant, this problem
occurs even when only one port is accessed at a time.

Looking at the "-current" sio.c, there appears to be code that is
relevant to this problem. Anyone know if I can use that
version of sio.c together with a RELEASE kernel? The changes looked
fairly significant.
-- 

Thorn Roby                              troby@diana.cair.du.edu
CARL Systems, Inc.                               troby@carl.org   
3801 E. Florida Ave.,Suite 300, Denver, CO 80210 (303) 758-3030