Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 29 Sep 2000 07:35:10 +0000 (GMT)
From:      Terry Lambert <tlambert@primenet.com>
To:        res03db2@gte.net (Robert Clark)
Cc:        res03db2@gte.net, tlambert@primenet.com, dot@dotat.at, freebsd-chat@FreeBSD.ORG, ragnar@sysabend.org
Subject:   Re: Ideas about network interfaces.
Message-ID:  <200009290735.AAA13895@usr08.primenet.com>
In-Reply-To: <200009290716.AAA01416@gte.net> from "Robert Clark" at Sep 29, 2000 12:16:13 AM

next in thread | previous in thread | raw e-mail | index | archive | help
> So would this PCI style of hardware pathing travel anywhere the PCI
> bus goes?

Yes.

> Its funny to see the PCI bus go so many places. I saw several
> books today on PCI. I may have to see if I can find an approachable
> one.

The "MindShare" one is pretty good, IMO.


> Why couldn't we just have one or more unique interrupts per slot, like
> on the apple II.

Because PCI doesn't cascade interrupts on all motherboards,
there's an interrupt wrap on slot 5, if there's a slot 5, and
some boards use more than one interrupt, which would make
the second device share the interrupt with the card in the
next adjacent slot (in the case of a cascade), or the first
device share with the previous adjacent slot (in the case of
a non-cascade).

Historically, Intel motherboards did not cacade (A/B/C/D for
slot one, B/C/D/A for slot two, etc.) unless they were from
their "Server Products Group".  Of course, the only group
that makes good motherboards at Intel is the SPG, and the
only groups you can buy motherboards from _aren't_ the SPG...

Suffice it to say, you can't just decide that you aren't
going to share, so long as card and motherboard designers
are still going insane from the arsenic they use to block
their hats, and can't agree with each other between all
vendors.

NB:	You'd think a "greedy" card that wanted more than
	one PCI interrupt would try to sneak a peek around,
	and try not to share an interrupt with another card,
	so that the vendors products at least appeared faster
	than their competition wo didn't peek around...


					Terry Lambert
					terry@lambert.org
---
Any opinions in this posting are my own and not those of my present
or previous employers.


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




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200009290735.AAA13895>