Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 13 Jul 1998 15:54:52 +0200 (CEST)
From:      Andre Albsmeier <andre.albsmeier@mchp.siemens.de>
To:        cmetz@inner.net (Craig Metz)
Cc:        andre.albsmeier@mchp.siemens.de, freebsd-net@FreeBSD.ORG
Subject:   Re: Looking for a 2 channel Fast Ethernet NIC
Message-ID:  <199807131354.PAA14338@internal>
In-Reply-To: <199807131024.GAA24219@itd.nrl.navy.mil> from Craig Metz at "Jul 13, 98 06:24:18 am"

next in thread | previous in thread | raw e-mail | index | archive | help
> In message <199807121637.SAA08386@internal>, you write:
> >I am looking for a 2 channel Fast Ethernet NIC. I have found
> >the Adaptec ANA-6922A and the SMC SMC9334BDT.
> ..
> >Does anyone have experience with this stuff? Or are there other
> >2 channel Fast Ethernet NICs to recommend for FreeBSD?
> 
>   We use the SMC9334BDTs rather religiously. They are great boards. SMC
> discontinued them in April or so and claimed that they'd be reintroducing them
> on June 2 and that they'd be the same thing, just redesigned for lower cost
> and to use the latest versions of the Tulip and the NS PHY chip. June 2 has
> come and gone, and neither SMC's distributors nor their salescritters has any
> idea what's going on. SMC's been pushing their own broken controller chip (the
> EPIC) lately, so it's possible that they're just bailing out of the Tulip
> market entirely.

I see. So the best would be to get some these old ones. They are using
the de-driver, right? Are you running -STABLE on them?

>   We have an ANA-6944TX (the four-port board), and it has several nontrivial
> design flaws that cause us lots of grief. Three I remember off hand are that
> the bridge IRQs are done wrong (all four Tulips and the bridge each ask PCI PnP
> for an IRQ on the motherboard PCI bus, but the bridge's IRQ is actually used
> for all four Tulips), the MII table is wrong (the values it says to send/
> expect on the MII port don't actually cause the right things to happen), and
> that the NS PHY chips on the thing, even when programmed correctly, seem to
> take a lot longer than they're supposed to in completing N-way negotiation.
> I am not sure if the two port board has the same problems as the four port, but
> it wouldn't surprise me if many of them were on that board, too.

OK, so Adaptec failed for me here :-)

>   Another manufacturer you may want to check into is Znyx (http://www.znyx.com)
> who also make 1/2/4-port Tulip boards. I have no first-hand experience with
> these, but I know many people who say they work fine.

Thanks for reminding me; I have forgotten the completely.

>   Note that, if you have an AMI BIOS, your BIOS sucks. The AMI BIOS does not
> deal correctly with PnP in the presence of PCI-PCI bridges. If you're lucky,
> your devices will get numbered backwards and you might waste some IRQs. If
> you're not, multi-port boards just plain won't work in your system (unless
> FreeBSD can be made to do PCI PnP assignments itself). People who have chatted
> with AMI about this have said that AMI is completely uninterested in fixing
> this.

I have no machine with AMI BIOS here (only Award). Thanks for the hint anyway.

Thanks for this great information.

> 
> 									-Craig
> 

	-Andre

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-net" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199807131354.PAA14338>