On  10 Sep, this message from Ryuichi Oikawa echoed through cyberspace:

I've tried Adaptec PowerDomain 3094UWD, dual channel UW SCSI card
with PCI-PCI bridge and two SCSI controllers under the bridge,
and I noticed driver related and OF related problems.

First I tried vger 2.2.12 kernel that caused machine check exception
at first IO write transaction. I thought that this was normal because
OF didn't properly set up P2P bridge's IO forwarding window.
The real problem was this part(aic7xxx.c):

#ifdef MMAPIO
          if ( !(temp_p->flags & AHC_MULTI_CHANNEL) ||
               ((temp_p->chip != (AHC_AIC7870 | AHC_PCI)) &&
                (temp_p->chip != (AHC_AIC7880 | AHC_PCI))) )
          {
	   ...... select MMIO and does ioremap ....
           ...... if it fails select normal IO ....
	  }
#endif

that rejected my SCSI card from MMIO forcing normal IO resulting in machin
check. So I had to ignore this statement by #if !defined(__powerpc__) ...
#endif statement.

Does anyone know what is this restriction for?

I could imagine that certain Adaptec SCSI chips use a regular PCI
memory region for their registers, while others use a PCI I/O region.
There are chances that the PCI I/O region is not enabled on your board,
and therefore accessing it results in a machine check.

Can you send me the output of lspci -vv, preferably once without the
aic7xxx driver in the kernel, and once with your fixes?

After that aic7xxx.c properly detected two SCSI chips but entered in a
timeout/resetting loop:-(  Apparently interrupts weren't fixed-up by
pcibios_fixup.

IRQs do get fixed, even on devices behind P2P bridges... iff OF did
assing IRQs, that is, as you found below.

So I scanned OF searching for correnct IRQ number and
found it at AAPL,interrupts field of P2P bridge. SCSI controllers didn't
have any AAPL,interrupts of course. It seems OF inserts AAPL,interrupts
only at the top node of the slot when it detects a card in the physical
slot. Therefore I had to change arch/ppc/kernel/prom.c as 

[snip'ed code]

Your fix looks OK to me; on PowerMacs, all PCI devices in any one slot
share the same interrupt, as the four PCI interrupt pins are OR'ed
together per slot.

I don't know if it is OF(PowerMac8500, OF 1.0.5) version specific,
or SCSI card specific, or PowerMac OF nature. Any ideas?

Either OF-specific in general, or one of the many bugs in OF 1.0.5.
Anyway, as I said, your patch wouldn't break anything, even if OF did
assign IRQs already.

Michel

-------------------------------------------------------------------------
Michel Lanners                 |  " Read Philosophy.  Study Art.
23, Rue Paul Henkes            |    Ask Questions.  Make Mistakes.
L-1710 Luxembourg              |
email   mlan@cpu.lu            |
http://www.cpu.lu/~mlan        |                     Learn Always. "


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/