On Thu, 5 Mar 1998, Peter Berg wrote:

> Hi,
> 
> now I have fitted a EtherM card into my RiscPC. If the card works, it works 
> very well, but:
> 
> after *each* power down reset the kernel crashes at the etherM initialisation 
> stage with
> 
> em0 at podulebus0 [ netslot 0 ]
> em0: Ethernet address 00:00:a4:10:ba:2b
> em0: 16KB buffer memory, jabber, UTP
> Fault with intr_depth > 0
> Data abort: "Translation fault (page)' status = 007 address = f018b000 PC = 
> f01265b4
> Instruction @Vf01265b4 = e52d0004
> Stopped at _Debugger+0x10 ldmdb r11, { r11, r13, r15 }
> db>

Could you type trace at this point?  (that's if you're loading the symbol
table from the bootloader)  I think it would help ppl see where the
problem is.  If I remember it's requested that any kernel problems have a
trace included (or it used to be).

Another thought on this is that the card can't cope with the rapid
sucession of 2 resets, I know that one of my cd-roms goes funny if I do
reboots sometimes...

> This means I have everytime I want boot RiscBSD to do a power down reset with 
> a hold down DEL key, reconfigure the whole RiscOS and boot RiscBSD from 
> RiscOS, or I have to remove the EtherM card completely from my machine.

I thought you could hold down alt or something similiar to avoid riscBSD
doing a fastboot, check in your predesktop and the relevant basic file.

On another note can anybody tell me if you have the 16bit sound card on an
issue 1 board with Risc-OS 3.5 whether you have to configure the sound to
16bit?  or does it work with another method?  If I have no responses I
can't safely say that my 16 bit sound stuff will work with RO 3.5.

Cheers,
Chris