Hi Ilpo,

this is very cool. I'm slowly getting my cross-compiler setup going
again after a 2 year break, so I can work on the IP20 port.
Any ideas whether this will work on GIO32bis buses? I'll have the answer
in a few weeks, I hope, but are there many differences between GIO64 and
GIO32bis (as used on IP20, IIRC)?

Last question: I had a look on Ebay a while back if there were any GIO
cards that I could scavenge, but I've never seen one offered. Does this
sort of hardware ever come up on Ebay or is it so rare to be effectively
non-existent?

Cheers,

Gerald

Ilpo Ruotsalainen wrote:
> 
> Getting sidetracked while hacking seems to be my specialty. This time I
> started with newport hacking and ended up writing GIO bus probe.
> 
> The attached patch implements a bus probe for the GIO slots (and GFX
> slot) and I'm very curious to hear what kind of product/revision codes
> it reports for real GIO devices and the graphics boards in Indy/Indigo2
> machines. (And of course if it breaks things for you I'd like to hear
> about that before I commit this...)
> 
> Sadly the linux-mips hackers seem to have come to the conclusion that
> the newport graphics boards do not implement the mandatory GIO product
> id read properly... If someone has relevant information, please let me
> know. (How does the PROM detect what kind of graphics hardware there is?
> Any SGI insiders on this list?-)