Subject: Re: sys/arch/hp300/dev/if_le.c broken
To: None <current-users@NetBSD.ORG>
From: Ty Sarna <tsarna@endicor.com>
List: current-users
Date: 07/03/1995 18:05:14
In article <199507031449.KAA12274@wheel.tiac.net>,
Gary D. Duzan <gary@wheel.tiac.net> wrote:
> In Message <9507031337.AA26640@snark.imsi.com> ,
>    "Perry E. Metzger" <perry@imsi.com> wrote:
> =>Peter Galbavy writes:
> =>> Just asking because if I run a NetBSD/sparc system, all the Ethernet
> =>> devices have the same MAC address (blame Sun).
> =>
> =>Why is that a bad thing? It follows the original intent of the way
> =>Ethernet was used at Xerox, actually. SunOS works just fine with
> =>things set up that way.
> =>
>    Couldn't it rather confuse intelligent bridges that connect the
> two nets?

Only if the bridges are broken, I would think. In fact, it seems to me
that it would be desirable to use the same ether addr on all interfaces
in such a situation, since an intelligent bridge could make use of that
information to improve performance:

bridge: "ether address X appears on a local segment A and through
another bridge on segment B... I will send to segment A since it's
closer"

or,

bridge: "ether address X appears on both segments A and B. A is more
heavily loaded [or is 10baseFOO] while B is less loaded [or is
100baseFOO]. Therefore I'll send on segment B."

Actually, this end up being an argument in support of the "ifconfig
ether..." feature, since on most systems the ether address is
per-interface, and with ifconfig one could configure all interfaces to
use the same address.