[Coco] System gen problem

[Steven Hirsch]

On Sun, 7 Feb 2010, Robert Gault wrote:

> Steven Hirsch wrote:
>
>> I actually think I'm not even getting far enough to worry about the
>> descriptors. The server is not even getting a request from the CoCo. The
>> Coco hangs immediately when I try to access the dw drive, but clearly no
>> communication with the server is actually taking place. Something more
>> fundamental is wrong, but what?
>> 
>
>> When I boot over DW using the pre-made image, things march along as you'd 
> expect.
>
>> Steve
>> 
>> 
>
> OK. So, if you boot from a Drivewire distribution disk you can access the PC 
> from the Coco? That must mean Drivewire is working as is the cable. It also 
> eliminates just about everything except incorrect modules in the os9boot 
> file.

> You can get more data from Drivewire if you check both Logging and 
> Statistics.

The Linux DW client shows that by default.  Nothing to log - no 
communication takes place.

> The first obvious question is are you using the correct rbdw3 and dw3 modules 
> for the version of NitrOS-9 you are booting? Not mixing up a 6809 version 
> with a 6309 system or vice versa?

I don't believe so, no.

> There really isn't much else possible if you can boot a disk mounted in 
> Drivewire but can boot on the Coco side and access Drivewire.

Now you understand the flat-spot on my forehead.

Something else occurs to me:  The Nitros-9 version.  My hard disk install 
uses 3.2.8, built Oct. 9, 2008.  When I boot over DW, it's: 3.2.9, built 
March, 2009.  Is the minor-number difference enough to prevent the former 
from cooperating with the DW drivers?

> I have a 6309 Coco system and Ident for the modules in question gives
> rbdw3 CRC $3e2550  Mine: $076c4e
> dw3   CRC $96409D  Mine: $e139ce

Those are straight from the (working) boot image.

Steve

--