Thank you for the information, David.

At 10:06 -0800 11/22/2002, David Gatwood wrote:
>On Thu, 21 Nov 2002, Jeff Walther wrote:

>>  Is there information on the PCI PowerMacs interrupt controller
>>  available somewhere?   That's handled by the Grand Central chip, yes?
>
>Or later ones used Heathrow/Gatwick.  Yeah.

Oh, I'm definitely living in a PowerSurge world.  :-)

>>  I'm wondering how many interrupts there are, which are allocated
>>  where, and what unused interrupts might be available.  In particular,
>>  I'm wondering if there are unused interrupts run out to pins on the
>>  Grand Central.
>
>The answer to those questions are "it depends".  The interrupt assignment
>varies from board to board.  The top level interrupt handling code gets
>the mapping info from Open Firmware.  As for whether there are extra pins,
>I have no idea.

Hmmm.  Is there a listing of interrupts and their respective devices 
for the 9500 and the 7500/8500?  Even without the pinouts that would 
help me match numbers and see what's available where.  I've seen the 
external interrupt listing in the ANS Hardware Developer Notes where 
it explains how they differ from the 9500.  However, that note only 
lists ten interrupts.   I suspect that the video on the 7500/8500 has 
its own set rather than reusing the lower slot interrupts on the 
9500, but I'm not certain.

Is the mapping info from Open Firmware something which is coded into 
the OF version.  That is, is it hard coded into the ROM?  Or is it 
something that OF detects from the motherboard?   The ROM and 
therefore the OF for the entire x500 line and the PCC clones (except 
PowerBase) and the Umax Storm Surge clones are all the same.   Of 
course, the OF could have a branch in there depending on which 
machine it finds itself installed in.

>>  For example each slot in the six slot machines has it's own
>>  interrupt.  The interrupt for each PCI slot is connected to its own
>>  pin on Grand Central.  So three of those interrupts are not used in a
>>  three slot machine, unless they are used by the built-in video
>>  circuitry on the {7,8} {5,6} 00.
>
>That is probably, but not definitely, correct.  It does assume that only
>one grand central variant exists.  I don't know for certain whether that
>is true or not.

Hmmm.  The Apple part number on the Grand Central chips are all the 
same as far as I can tell (343S1125-03).    For a more definite test 
I could move the Grand Central chip from one model machine to another 
and see if the recipient board still works properly.   Do you have a 
suggestion for a donor and a recipient motherboard model which you 
think would be a good test?

Jeff Walther