From owner-freebsd-hackers  Tue Jan 14 16:34:56 1997
Return-Path: <owner-hackers>
Received: (from root@localhost)
          by freefall.freebsd.org (8.8.4/8.8.4) id QAA19696
          for hackers-outgoing; Tue, 14 Jan 1997 16:34:56 -0800 (PST)
Received: from research.gate.nec.co.jp (research.gate.nec.co.jp [202.32.8.49])
          by freefall.freebsd.org (8.8.4/8.8.4) with ESMTP id QAA19691
          for <hackers@FreeBSD.ORG>; Tue, 14 Jan 1997 16:34:48 -0800 (PST)
Received: from sbl-gw.sbl.cl.nec.co.jp by research.gate.nec.co.jp (8.8.3p1+2.6Wbeta9/950912) with ESMTP id JAA04582; Wed, 15 Jan 1997 09:34:36 +0900 (JST)
Received: from sirius.sbl.cl.nec.co.jp by sbl-gw.sbl.cl.nec.co.jp (8.7.6+2.6Wbeta7/3.3W6) with ESMTP
	id JAA29512; Wed, 15 Jan 1997 09:34:32 +0900 (JST)
Received: by sirius.sbl.cl.nec.co.jp (8.7.5+2.6Wbeta6/3.3W6) with UUCP
	id JAA28430; Wed, 15 Jan 1997 09:34:28 +0900 (JST)
Date: Wed, 15 Jan 1997 09:34:28 +0900 (JST)
From: Naoki Hamada <nao@sbl.cl.nec.co.jp>
Message-Id: <199701150034.JAA28430@sirius.sbl.cl.nec.co.jp>
References: <199701142313.SAA02192@chaos.tmok.com>
To: daver@chaos.tmok.com
CC: mcgovern@spoon.beta.com, hackers@FreeBSD.ORG
In-reply-to: Dave Costantino's message of "Tue, 14 Jan 1997 18:13:43 -0500 (EST)"
	     <199701142313.SAA02192@chaos.tmok.com>
Subject: Re: 3com 3c509 / SB16 PNP Problems...
Sender: owner-hackers@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

Dave Costantino:
>Here's what I did to make it work... I don't guarantee this will work with
>anyone elses configuration, but it's working for me.

PnP feature of the 3C509 and newer 3C509B adapters is nothing to do
with this problem. 3C509 adapters use 'id port' while
initialization. Id port is the most recently accessed address among
i/o port 0x100, 0x110 ... 0x1f0. Since the driver a priori assumes
that the id port is 0x100 and some type of sound blaster cards use i/o
port 0x100, you fail to detect your 3C509 adapter.

There seems no device which use i/o port 0x110, so the exact solution is:
>2) In /usr/src/sys/i386/isa/elink.h, I changed: "#define ELINK_ID_PORT 0x100"
>   to read "#define ELINK_ID_PORT 0x110".

This must be committed. Linux way apparently seems better, but
it is known that sometimes it fails.

- nao