Oops... forgot to include port-mac68k...

--- begin forwarded text


Mime-Version: 1.0
Date: Wed, 10 Jun 1998 21:23:04 +0200
To: Simon Raahauge DeSantis <xiamin@scdesantis.ne.mediaone.net>
From: Hauke Fath <hauke@Espresso.Rhein-Neckar.DE>
Subject: MACH_Q2 class IO mapping, was:  NetBSD/mac68k floppy driver
Cc: Hauke Fath <hauke@Espresso.Rhein-Neckar.DE>
Status:

At 3:17 Uhr +0200 10.06.1998, Simon Raahauge DeSantis wrote:
>
>I do have options LKM in my 1.3 kernel. The patch did apply cleanly. I do
>not have INSECURE in my config, but that shouldn't matter if I'm in single
>user mode (as I was). Here's the output I get from modload -dv
>/usr/lkm/IWM.o while in single-user mode:
>
>ld -A /netbsd -e _IWM_lkentry -o /usr/lkm/IWM -T 0 /usr/lkm/IWM.o
>ld -A /netbsd -e _IWM_lkentry -o /usr/lkm/IWM -T 61d000 /usr/lkm/IWM.o
>iwm: IWMBase mapped to 0x1e000 in VM
>modload: error initializing module: Input/output error
>>
>> >I'm using a Q650 running a 1.3 custom
>> >kernel with the machdep.c patch applied.

Ahhhh...  (big light bulb going on)

I have slightly different init code for the LKM than for the in-kernel
driver; it does not issue diagnostic messages unless DEBUG is defined.
Withe the GENERICIWM kernel, you'd have gotten a diagnostic along the lines
of "error -77".

Yours is a MACH_Q2 style Quadra (to use the diction of machdep.c) like the
Performa 630 that I tested. Try booting (or attaching the LKM manually)
with a 800k disk inserted, and tell me if it makes any difference. On the
P630, I saw bogus messages about "two drives attached, with an unsupported
HD disk in each of them".

I had a look at the relevant sections of the P630's ".Sony" driver code,
and it is byte for byte identical with the Quadra 700 ROM that I worked
from. So, I'd suspect  an IO mapping problem.

Could anybody with VM knowledge please speak up here?

	hauke


--
"It's never straight up and down"     (DEVO)

--- end forwarded text




--
"It's never straight up and down"     (DEVO)