*BSD News Article 21227


Return to BSD News archive

Path: sserve!newshost.anu.edu.au!munnari.oz.au!hp9000.csc.cuhk.hk!saimiri.primate.wisc.edu!sdd.hp.com!elroy.jpl.nasa.gov!swrinde!cs.utexas.edu!natinst.com!hrd769.brooks.af.mil!hrd769.brooks.af.mil!not-for-mail
From: burgess@hrd769.brooks.af.mil (Dave Burgess)
Newsgroups: comp.os.386bsd.questions
Subject: pccons expert wanted...
Date: 20 Sep 1993 19:30:48 -0500
Organization: Armstrong Laboratory, Brooks AFB, TX
Lines: 63
Message-ID: <27li06$ilk@hrd769.brooks.af.mil>
NNTP-Posting-Host: hrd769.brooks.af.mil

First,  thanks to everyone that has helped me get my EGA monitor
working.  It seems as though we are at a stumped point, and I thought it
would be a good time to trawl for experts one more time before I move to
Omaha and possibly lose my net.access.

I am using an Oak Technologies VGA card configured as an EGA card with
an EGA monitor attached.  I am trying to boot NetBSD with it.  NOTE:  If
I could afford a VGA monitor, I would have already bought one and this 
question would be moot.

The bootstrap message (on the NetBSD-current disk I created) comes up in
glorious white on black.  I can read clearly as the system proceeds to
load each of the pieces of the kernel.  The boot disk then asks for a
root file system disk.

At this point, the text that was displayed by the boot strap scrolls off
the top of the screen.  The cursor turns dark blue, and the system just
sits there.  The keyboard is still kind of active, i.e I can turn on the
caps lock, scroll lock, and num lock; and turn them all off again.  If I
press enough keys, I can get the 'buffer overflow' bell to go off.  Once
in a while, I will even get a dark blue 'trap type 12:  offset 0' to
show up.

I have tried the original 386bsd 0.1 disk, a patched boot disk, a NetBSD
boot disk, a FreeBSD boot disk, and a Linux boot disk.  They all exhibit
more or less the same symptoms.  I have listed the differences below:

Original disk didn't turn the cursor blue; it turned it gray.
All the others look identical (little surprise really, they all use
about the same pccons routines)

The Linux disk was the most disconcerting...

I could read the bootstrap text, and could read the 'autoconfig'
equivalent screen, but the screen did not scroll.  The virtual consoles
were active, and I was able to change from VC to VC.  After I changed
back to the initial VC, my cursor was on the screen, as though the
scrolling for my card was not working.

Anyway.

Does anyone know what the difference is in the video handling between
the  bootstrap code and rest of the system?  If so, I would appreciate
hearing from you.  The fact that the bootstrap works with the card and
NetBSD doesn't in tantalizingly excrutiating.

I am running a 386DX with an AMI BIOS, 4 Meg of memory, and a generic
IDE hard drive controller.  Everything else is pretty much bargain
basement (as that is all my weak kneed budget can support).

I have already waded in WAY over my head in the pccons source, and I
don't see anything in there that should make this hard to do.  The
default mode for the card has the map address at B000 or B800, both of
which are valid map addresses for the card and both of which correspond
to addresses in pccons.  The card does use IRQ-2 for 'sparklie 
reduction', but I am not using a network card; nor anything else at 
IRQ-2, so I don't think this should be a problem.

-- 
------
TSgt Dave Burgess
NCOIC AL/Management Information Systems Office
Brooks AFB, TX