*BSD News Article 16610


Return to BSD News archive

Path: sserve!newshost.anu.edu.au!munnari.oz.au!news.Hawaii.Edu!ames!agate!spool.mu.edu!uunet!mcsun!sun4nl!tuegate.tue.nl!svin09.info.win.tue.nl!wzv.win.tue.nl!gvr.win.tue.nl!guido
From: guido@gvr.win.tue.nl (Guido van Rooij)
Newsgroups: comp.os.386bsd.questions
Subject: Re: Modem setup on 386BSD [and a further QUESTION]
Date: 29 May 1993 16:04:37 GMT
Organization: Eindhoven University of Technology, The Netherlands
Lines: 45
Distribution: world
Message-ID: <1u81il$o45@wzv.win.tue.nl>
References: <1993May27.043819.13711@mcshub.dcss.mcmaster.ca> <CJB.93May28115334@thrip.cs.uq.oz.au> <1993May28.195017.23712@fcom.cc.utah.edu>
NNTP-Posting-Host: gvr.win.tue.nl

terry@cs.weber.edu (A Wizard of Earth C) writes:

>In article <CJB.93May28115334@thrip.cs.uq.oz.au> cjb@cs.uq.oz.au (Christopher J Biggs) writes:
>>Question:  Has anyone got a driver that will drive a multi-IO card at IRQ 2&3
>>and an internal modem on 2or3 (it wont go anywhere else!!).

>I haven't seen one yet -- and, unless you have a very special modem and com
>card, you aren't going to be able to write one -- ever.

>Interrrupt sharing depnds on being able to ask all devices using the
>interrupt if they were the one that caused the interrupt -- basically,
>the interrupt is a "data available" flag.  Most serial hardware doesn't
>support a flag to indicate data available since last read.  Multiport

Not true. The 8250, 16440, 16450 and friends all have a bit telling if
this UART triggered the interrupt. (bit 0 in the IIR register).

The reason you can't share interrupts between cards in different slots
in an isabus system is the fact that you can break your hardware: if
one card generates an interrupt it wants to hold up, say, the isa bus's
irq 3 line. But the other card is holding the same line down. Unless you
designed your system such that this is allowed, normally you have problems.

>boards, on the other hand, *do* support flagging the interrupt source,
>*but* require a driver that knows about the flagging mechanism.  Some
>UARTS (but not the low-end ones in most PCs) support a status register
>indicating "data in FIFO".  This can be used the same way as a hardware
>specific interrupt origin flag (like most multiport boards use), but
>the hardware specific flags are not frequently in the same place between
>multiple hardware vendors, so a special driver is normally needed with
>at least a vendor- (and potentially a product-) specific piece of code
>to query interrupt origin.

>					Terry Lambert
>					terry@icarus.weber.edu
>---
>Any opinions in this posting are my own and not those of my present
>or previous employers.

-Guido
-- 
Guido van Rooij                 |  Internet: guido@gvr.win.tue.nl
Bisschopsmolen 16               |  Phone:    ++31.40.461433
5612 DS Eindhoven               |  ((12+144+20)+3*sqrt(4))/7 
The Netherlands                 |    +(5*11)=9^2+0