On Monday 19 February 2001  8:10 pm, Dave Daniels wrote:
> I tried starting X on my A7000+ but all it did was cause NetBSD to
> die horribly. When I typed 'xinit' the screen became corrupt.
> Pressing the return key gave me the 'db>' prompt. There was no
> kernel panic message. However, typing the command 'show pool'
> produced the following output:
>
> POOL[u]vm_fault (0xf35065e0, e3720000, 1, 0) -> 1
> Unhandled trap (frame = 0xf35f9c0c)
> Data abort: 'Translation fault (section)'
>     status=005 address=e372037e PC=f017947c
> Stopped in Xarm32VIDC at 0xf017947c: ldrb r3,[r2,#0x0000]
>
> This failure happens every time I try to run X.
>
> I am quite happy to believe that this happened as a result of
> something I did. I ran xf86config beforehand to create XF86Config
> and changed the symbolic link 'X' to point at Xarm32VIDC. Is this
> correct?
>
> This was with NetBSD 1.5.

Hmm, are you allocating enough DRAM for video?  I've noticed that my risc pc 
doesn't actually have a /etc/XF86Config, I think it uses the mode currently 
in use.  you might also need a /etc/monitor.conf file, it should just be your 
moniter definitions file from RISC-OS.

As richard suggests finding the right panic would be a good start.

Cheers,
Chris