From owner-freebsd-current Tue Feb 18 02:27:37 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id CAA24710 for current-outgoing; Tue, 18 Feb 1997 02:27:37 -0800 (PST) Received: from hq.icb.chel.su (hq.icb.chel.su [193.125.10.33]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id CAA24567 for ; Tue, 18 Feb 1997 02:24:29 -0800 (PST) Received: (babkin@localhost) by hq.icb.chel.su (8.8.3/8.6.5) id PAA07973; Tue, 18 Feb 1997 15:15:01 +0500 (ESK) From: "Serge A. Babkin" Message-Id: <199702181015.PAA07973@hq.icb.chel.su> Subject: Re: ep0 in GENERIC To: bde@zeta.org.au (Bruce Evans) Date: Tue, 18 Feb 1997 15:15:00 +0500 (ESK) Cc: freebsd-current@freebsd.org, j@ida.interface-business.de In-Reply-To: <199702180148.MAA20191@godzilla.zeta.org.au> from "Bruce Evans" at Feb 18, 97 12:48:37 pm X-Mailer: ELM [version 2.4 PL23] Content-Type: text Sender: owner-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > > >Why don't we use the following config line for ep0 (3C509)? > > > >device ep0 at isa? port ? net irq ? vector epintr > > Because the probe doesn't honour conflicts and is invasive. > > >The autoprobing seems to work, and it's as invasive as not using > >autoprobing since it happens by reading the EEPROM via a fixed port > >address (which is done anyway as long as at least one ep device is > > No, the autoprobing isn't done if a previous driver finds something > at 0x300, because the ep0 address conflicts so ep0 isn't probed. The > probes are done in non-alphabetical order to increase the chance of > avoiding invasive probes like ep0's. I'm not shure about the current state but the history is: ep0 conflicted with some another 3COM card driver. It seems to me that this conflict was resolved but I'm not shure. But it does that independenlty of is it configured with explicit or auto-recognized address. The ep0 probe routine uses port 0x100, not 0x300 and reads from it about what cards are there. It uses port 0x300 only if there is a 3c509 card configured for that address. -SB