> > [A]fter a year of pfutzing around, I finally got a working NetBSD up
> > on one of my VME crates (3/260).  It is NetBSD-1.1A.  Everything
> > after that dies in the scsi area.
> 
> That sounds...wrong.  I know I've run 1.2 on my -3/260s.
> 
> I'll have to try to dig one out and see if I can bring it up.

Maybe it is my rotten luck, but 1.2 blew up on me last night.  That is
the reason I went all the way back to 1.1, so I could check what did work
and where the problems started to creep in.  I was browsing through some
of the ancient historical mail in the NetBSD archives and there was specific
mention of a 3/260 working on 1.1, so that is why I inquired a couple of
days back, on the list, for the earliest builds.  When the polling drivers
were changed to DMA, that is where my scsi interrupt problems start.
I sense it has to do with changes in the scsi drivers between 1.1 and 1.2.
I did debug the 1.1 kernel from scsi level 0 to 3 and that was still
working on my machine.  I did not try the higher levels of DMA speed, yet.
I had one other build around 1.3.x up for a couple of days, but that
proved to be unstable on my hardware. IF there is something I am missing,
please do holler.  I have NetBSD current and 1.4.2 up on DECstations,
VAXen of odd flavors, and sparcs (IPX and SS1), but have had abysmal
luck on my sun3/4 boxes with vme controllers.  I had it up on a 3/80
a couple of releases back, and that did fine.  It is the VME controllers
and NetBSD DMA drivers that seem to give me fits. IFF you can get a 3/260
box up, do tell me what happens and how you get it up.  One other fellow
on the list said he had a 3/260 up on 1.4.  My boxes just don't run it.
James Birdsall (the sun3 faq guru) was having similar problems, so I
am not alone..... (a little wierd and cantankerous, maybe, but not
alone).  Anyway, I will do what I can to load it and see what it can
do.

Thanks

Bob