Subject: Re: trouble booting SE/30
To: None <allanlee@viola.com>
From: Colin Wood <cwood@ichips.intel.com>
List: port-mac68k
Date: 07/17/1997 18:24:17
Allan M Lee wrote:
> 
> This is perhaps a common problem, I don't know.
> 
> Machine: SE/30
> MacOS: System 7.5.3
> Control Panels and extensions: all off except Mode32

Sounds good so far, but (and this may sound wacky), is the monitors
control panel in color or black&white?  I think one user reported having
to put it in "color" mode on an se/30 to get it to boot (even tho the
machine is monochrome!)

> When I tried to boot with "extra debugging info" box checked, it reads
> 
> >>>
> [snip]
> 
> Set _boothowto to 0x2.
> 
> Bye-bye
> [snip]
> to...
> 
> Set _mac68k_vrsrc_cnt to 0x1
> 
> <<<
> 
> and froze!  I tried a number of times with the same result.  What am I
> doing wrong?

Hmmmmm...how did you go about installing the distribution?  It may be that
the kernel file was corrupted on the install.  If not, try the monitors
thing above.

What other boxes do you have checked?  What is the name of the kernel you
are booting?  Have you looked through the FAQ and checked all of the items
on the checklist there?

> Misc info:
> 
> 1)
> After building devices, I was not successful in this command: "mount
> /dev/sd0a/ /usr"
> 
> output:
> 
> "/usr:  device is busy" (or something like that). 

This is because your root partition is /dev/sd0a.  The Installer mounts
this one by default, so it's already in use.  If you do in fact have a
separate Usr style partition made, it will most likely be /dev/sd0g
(although depending on your drive layout, it might also be /dev/sd0[efh]).
Try a 'disklabel' command in the Installer's mini-shell (I think this'll
work), and see what partitions you have available to find your /usr
partition.

Also, the above command should be:

mount /dev/sd0g /usr

Note that there's no trailing slash on the device name (although this just
might not make a difference, I think it might break something).

> 2)
> When using the installer, I got an "Error on SCSIRead(), # 5"; and in Mini
> shell, when I used the ls command, after .profile was the line "fstat : No
> such file or directory" then continued with netbsd.  (netbsd didn't show
> up the first time I installed it, or thought I did.)

This is fairly common.  I think that the first error message is documented
in the FAQ.  A type #5 error is a SCSI Phase error and is most often
caused by having non-harddrive SCSI devices (i.e. CD-ROM's) on the device
chain during the install, although it may be caused by other conditions.
At any rate, you can safely ignore it unless you are just being flooded by
the messages.

As for the second error, I see this one fairly often, and I'm not too sure
what it means.  I assume that there might some kind of stuck inode, but
fsck doesn't pick it up either.  Does the Installer maintainer have any
clue on this one????  (I haven't groveled through the source to find out
where it's coming from.)

> 3)
> The three set base121, netbsd, etc121 were ungzipped using stuffit
> expander with enhancer before using installer to install them.  I didn't
> know the installer could decompress the sets.

Ahh...this might be the problem.  SE tends to do linefeed conversions on
files (so that they will be Mac-readable).  This tends to make them rather
unusable for NetBSD purposes, tho :-(  Go ahead and reinstall using the
Installer (and make sure that you transferred them in binary mode to your
machine in the first place).

I hope this helps.

Good luck.

Colin