I recently had a motherboard replaced in an as500. It came with a very
new version of the firmware:  V6.4-5 Feb 21 1997 16:41:14

Now when I boot netbsd, it gives very verbose output for each pci
device:

vendor 0x0080 product 0x8080 (class 0x21, subclass 0x04, interface 0xd8, revision 0xf1) at pci0 dev 30 function 0 not configured
vendor 0x0080 product 0x8080 (class 0x80, subclass 0x80, revision 0x80) at pci0 dev 30 function 2 not configured
vendor 0x0080 product 0x8080 (class 0x80, subclass 0x80, revision 0x80) at pci0 dev 30 function 3 not configured
vendor 0xc570 product 0x0152 (class 0x80, subclass 0x80, revision 0x80) at pci0 dev 30 function 4 not configured
vendor 0x0080 product 0x8080 (class 0x80, subclass 0x80, revision 0x80) at pci0 dev 30 function 5 not configured
vendor 0x0080 product 0x8080 (class 0x80, subclass 0x80, revision 0x80) at pci0 dev 30 function 6 not configured
vendor 0x0080 product 0x8080 (class 0x80, subclass 0x80, revision 0x80) at pci0 dev 30 function 7 not configured

This patch seems to shut up the messages, but I'm not sure if its the
"right thing" to do.

<3:05pm>marilyn/gallatin:sys>diff -c arch/alpha/pci/cia_pci.c{_orig,}
*** arch/alpha/pci/cia_pci.c_orig       Tue Apr  8 07:17:20 1997
--- arch/alpha/pci/cia_pci.c    Sat Apr 26 14:57:50 1997
***************
*** 189,194 ****
--- 189,199 ----
        printf("cia_conf_read: tag 0x%lx, reg 0x%lx -> %x @ %p%s\n", tag, reg,
            data, datap, ba ? " (badaddr)" : "");
  #endif
+       /* fix a bug in newer Alcor firmware where empty pci slots 
+          are probed & return 0x80800080 */
+ 
+       if(offset == PCI_ID_REG && data == 0x80800080)
+           data = 0;
  
        return data;
  }

Drew
------------------------------------------------------------------------------
Duke University				Email:		gallatin@cs.duke.edu
Department of Computer Science		Phone:		(919) 660-6590