Subject: boot netbsd on a 486SLC2/66
To: None <port-i386@NetBSD.ORG>
From: Phil Knaack <>
List: port-i386
Date: 03/10/1995 14:44:21

I recently turned by 386/33 into a 486slc2/66 .. I would say that it
is great except that I seem to be having difficulties getting NetBSD
to run on it.

It boots just fine, detects all my devices and such, and even attempts
to boot, but it doesn't get very far. Starting in single-user mode, it
gives the prompt about the shell to run, but once the shell is
started, after a couple of seconds or less it traps and goes to ddb.

Here is the output, copied from my markerboard on the wall:

Enter pathname of shell or RETURN for sh:
# <pause 3 seconds> vm_fault(f816f3b0, 0, 1, 0) -> 5
  kernel: page fault trap, code = 0
stopped at 0xa7c; vm_fault(f816f3b0, 0, 1, 00 -> 5
  kernel: page fault trap, code = 0
stopped at _db_read_bytes+0x10: movb 0(%edx),%bl
db> trace
_db_read_bytes(a7c, 1, f7bffc50, a7c, 0) at _db_read_bytes+0x10
_db_get_value(a7c, 1, 0, a7c, a7c) at _db_get_value+0x17
_db_disasm(a7c, 0, f8103062, a7c, 2) at _db_disasm+0x16
_db_print_loc_and_inst(a7c, 0, f7bffd98, 0, f7bffd0c) at
_db_trap(6, 0, 1, 6, f81765f0) at _db_trap+0xa8
_kdb_trap(6, 0, f7, bffd58) at _kdb_trap+0x95
_trap() at _trap+0x190
---  trap (number 6)  ---
vm_fault(f816f3b0, 0, 1, 0) -> 5
kernel; page fault trap, code = 0
Faulted in DDB; continuing ...
db> ps
pid proc     addr     uid ppid pgrp flag  stat em comm wchan
 3  f8660000 f93d3000  0    1   3   004086  3   0  sh ttyin f82ba00
 2  f8660700 f93d0000  0    0   0   000204  3   0 pagedaemon
                                           thrd_sleep f816f388
 1  f8660a00 f93cc000  0    0   1   004084  3   0 init wait f8660a00
 0  f81765f0 f8186000  0   -1   0   000204  2   0 swapper

My system:
	486slc2/66, XC87-33 co-proc, 8mb, 64k cache
	motherboard: alaris, AMIBIOS

The other hardware isn't relevant, as I pulled out all the cards
except the video and a sIDE board and booted from a floppy for the
above test.

At first I thought it was the proc or motherboard or something, yet
going into the bios and disabling every special feature i could think
of including cache and putting all wait states at their max did not
affect it. This same motherboard runs dos, windoze/workgroups, doom, a
recent generic linux kernel and all those other no-nos just fine.. :) I
think I've pretty much ruled out hardware problems/failures.

So .. what's the next step? I'd like to get this machine up and
running, and seeing as how I've got a full week's break coming up
starting now I'm not afraid of spending a lot of time on this. 

I don't know how helpful the above trace/message is, but it seems
strange to me that it would fault again in ddb trying to figure out
what the first fault is about. For completeness I have included by
kernel config file below, and the dmesg is pretty unexciting. I can
provide it if necessary, since if I sit at the "pathname of shell or
RETURN" prompt forever with no problems.

Any ideas on how to proceed or what to check would be most
appreciated. I keep coming back to the familar phrase "It SHOULD
work.. why doesn't it?" :)



#  BLADE.VORPAL.COM -- the BEAST machine .. harbinger of evil.

machine		i386

options		I486_CPU

options		TIMEZONE=420
options		DST=1
maxusers	8

#options		DUMMY_NOPS

options		FIFO


options		FFS
# all "complicated" stuff disabled until blade works
#options		LFS,UNION
options		INET

options		DDB


config		netbsd	root on sd0 swap on sd1 dumps on sd1
npx0	at isa0 port 0xf0 irq 13

isa0 at root
vt0	at isa0 port 0x60 irq 1

#mms0	at isa0 port 0x23c irq 5

wdc0	at isa? port 0x1f0 irq 14
wd*	at wdc? drive ?
fdc0	at isa0 port 0x3f0 irq 6  drq 2
fd*	at fdc? drive ?

options	SCSI

aha0	at isa0 port 0x330 irq 11 drq 5
scsibus0 at aha0
sd*	at scsibus0 target ? lun ?

#ed0	at isa0 port 0x280 irq 7 iomem 0xcc000

#sb0	at isa0 port 0x220 irq 9 drq 1

pseudo-device 	pty		8
pseudo-device	loop		1
#pseudo-device	bpfilter	2
#pseudo-device	vnd		2