Subject: DS5000/125 difficulty with monitor
To: None <port-pmax@NetBSD.ORG>
From: Anne Bennett <anne@alcor.concordia.ca>
List: port-pmax
Date: 04/30/1997 17:49:42
Hi, folks.

I was trying to install netbsd-1.2 (pmax) on a DECstation 5000/125,
using the miniroot method, but when I tried to boot the miniroot, after
telling me the sizes of the kernel and claiming at was starting at
"blah" address, it started methodically writing garbage to the top half
or so of my screen and "sort of" scrolling it (i.e. "scrolling" the top
half of the screen in a funny horizontal and vertical pattern).  After
it while this stopped the screen went blank, so, guessing that the
kernel had perhaps written its boot messages and was asking me for the
root device, I typed in a a root device, and the disk became active
briefly, confirming my guess.

I tried the disk with the miniroot on it on another machine (DS2100),
and it booted the miniroot just fine, kernel messages and all.

It looks as though the miniroot kernel it not speaking to my monitor
correctly on the DS5000, almost as though it was assuming a lower
resolution and twiddling the wrong pixels -- at least, based on the
access pattern and weird scrolling, that's what I think.

The graphics card is a PMAG-AA (according to cnfg), and the model
number on the back of the monitor is VR319-DA, which according to the
Hardware Operator's Guide for the machine has a monochrome CRT, a
resolution of 1280x1024, and refreshes at 72 HZ.  The set-up works
fine under Ultrix, fine at the boot PROM level, and after I did a hard
reset of the machine, the monitor returned to showing me garbage at
the top and my boot command and the kernel's output up to "starting
at..." near the bottom -- i.e., what had been displayed before the
screen blanked out.  So it seems that the netbsd-1.2 miniroot kernel
simply is communicating with my monitor (or the frame buffer)
incorrectly.

I did check the list of supported hardware, and support is claimed for
my system (5000/125), and PMAG-AA "1280x1024 four-bit greyscale frame
buffer".

Ideas, anyone?


Anne.
-- 
Ms. Anne Bennett, Computing Services, Concordia University, Montreal H3G 1M8
anne@alcor.concordia.ca                                       (514) 848-7606