Subject: Re: 604e vs. 604ev (was Re: results of the IRC debug patch)
To: Tim Kelly <hockey@dialectronics.com>
From: Michael <macallan18@earthlink.net>
List: port-macppc
Date: 12/05/2004 12:04:22
Hello,

> >> Probably not too much I can help with there. I'm going to try to get
> >> somewhere on the mace issue.
> >I just thought the problem with the 3rd mc interrupt might be related.
> 
> I'm not sure about this, since it has been unregistered since the mc code
> was committed, and my later review of the interrupt stuff showed that my
> concern regarding the third interrupt was misdirected.
Hmm, what is this interrupt doing anyway? Obviously it's not really needed. I thought it's commented out because it's shared with another interrupt with a different priority, just like in my case. There must be a race condition hidden in extintr.c somewhere :/

> >> Yes, if I understand your previous email regarding this, your card does not
> >> contain any OF, in which case OF will only assign a node and basic
> >> addresses.
> >No. The card contains an OF ROM ( but not enough to boot anything besides
> >MacOS and only for the SCSI part )
> >OF doesn't see the ethernet part at all, if there was a generic node
> >everything would be fine, but there's nothing at all. Apparently OF
> >doesn't expect a device 17 on the first bandit.
> >
> 
> Does the node show up in its bandit's tree? I think we went over this a
> while back but I've piled too much onto my brain since then to remember.
There is no node for the Ethernet part. If there were one we wouldn't have a problem with it - as it is I have to assign it an interrupt by hand.

> >> I had to write the code because I have a XLR8 ZIF Carrier card that lets me
> >> swap in raw CPU chips, but it doesn't have any L2 information in OF with
> >> it, obviously.
> >Yep, as far as I can tell the backside cache on my CPU card doesn't end up
> >as an OF node either, it only >knows about the mainboard cache.
> 
> Actually, if you're running a backside cache on an aftermarket card, you
> should remove the motherboard L2 cache. At least, that what the docs I've
> seen have said.
The NewerTech config tool allows to have all caches enabled - utterly stable here. Maybe I should run some benchmarks to see if it has any effect at all.

have fun
Michael