Path: utzoo!mnetor!uunet!seismo!sundc!pitstop!sun!decwrl!purdue!i.cc.purdue.edu!j.cc.purdue.edu!pur-ee!uiucdcs!uxc.cso.uiuc.edu!uxg.cso.uiuc.edu!schwager
From: schwager@uxg.cso.uiuc.edu
Newsgroups: comp.sys.amiga.tech
Subject: Weird Guru...
Message-ID: <44000001@uxg.cso.uiuc.edu>
Date: 25 Apr 88 18:23:00 GMT
Lines: 63
Nf-ID: #N:uxg.cso.uiuc.edu:44000001:000:2594
Nf-From: uxg.cso.uiuc.edu!schwager    Apr 25 13:23:00 1988



Here's a weird one:
I was rearranging my master Aztec C disk, the one I boot for doing my C
work.  I had replaced all my AmigaDos commands with Arp commands, and I
added rez also.  Cool stuff, by the way, folks.  Rez is fantastic, as is
the arp stuff.  Anyways, I had everything like I wanted so I finished
the startup-sequence file on my disk and was one happy camper... except
it took a looooonnngg time to boot and the disk gronked a lot.  So I did
a copy dfX: to dfX: to see if that helped.  Well, when I rebooted on the
newly copied disk it certainly did halve my bootup time (or so it
seemed, anyway, it was a lot faster).  But, (and here's the weird part)
I got a guru meditation error! 

Thinking that rez might be the problem, I rebooted the machine on a
(relatively) raw workbench disk... I say relatively because I'd added
vt and vc to it previously and took the endcli out of the
startup-sequence, but those were the only changes.  Now I could go in
and try to edit the startup-sequence file on the disk that guru'ed, but
whenever I tried to write the file out, another guru!  This happened if
I used ed or the microemacs from the Extras disk.  It happened even if I
booted cold from the Workbench disk (since rez is a beta version, I
kinda suspected it of messing around with my memory).  Nothing worked-
as soon as I tried to write the startup-sequence file out, guru.  So I
finally ended up copying my Aztec master disk to a newly formatted disk using
AmigaDos' version of copy.  Reboot off the newly copied master... no
problem.  I'll include a copy of my startup-sequence that caused the
problem (I'm using the same one now, error free).  The guru would happen
at the very  end, since my mclk did indeed fire up.  I never got a cli
prompt:
-Mike Schwager
-- {ihnp4,convex,pur-ee}!uiucdcs!schwager   schwager%uiuc@csnet-relay.arpa
	University of Illinois, Dept. of Computer Science
*********************Begin startup-sequence file************
mount vd0:
Dir vd0:
AddBuffers df0: 20
AddBuffers df1: 20
if NOT EXISTS vd0:common
	df0:common/mkdir vd0:common
	df0:common/mkdir vd0:include
	df0:common/copy df0:lib/c.lib vd0:
endif
df0:common/cd df0:common
rez ls more list mkdir path copy grep delete arun cd set type info
df0:common/path df0:common add
echo "Mike's Aztec C 3.4a Work Disk"
Stack 8000
set INCLUDE=vd0:include!SYS1:include!SYS2:asm
CLIB=vd0:!SYS1:lib!SYS2:lib!
set FUNCLIST=SYS2:lint/manx.c DBINIT=s:.dbinit
set CCTEMP=ram:
path ram: add 
path df1: add 
path df0:bin add 
setclock opt load
;
;cd df0:bin
;rez cc as
;
arun GOMF1.0
arun df0:c/mclk
date > SYS1:s/now