On Mon, Jul 03, 2000 at 01:11:56AM -0700, R. C. Dowdeswell wrote:
> 
> On 962609946 seconds since the Beginning of the UNIX epoch
> "R. C. Dowdeswell" wrote:
> >
> >Illegal device detected on primary bus in physical slot 5
> >
> >Power down the system and remove the unsupported device from slot 5
> 
> So, I switched the ZLXp-E2 into a 32bit slot and the system booted.
> I then, of course, ran into port-alpha/9328 -- the firmware doesn't
> initialize the ZLXp-E2.  But this seems to be my repsonsibility,
> so I'll have a look at it.  :-)

Taken from release/texts/alpha/HARDWARE.TXT from FreeBSD-current:

<quote>
DMA bug:
MX5 has problems with DMA via the 2 64-bit PCI slots when this DMA
crosses a page boundary. The 32 bit slots don't have this problem because
the
PCI-PCI bridge chip does not allow the offending transfers. The SRM code
knows about the problem and refuses to start the system if there is a PCI
card in one of the 64bit slots that it does not know about. Cards that are
'known good' to the SRM are allowed to be used in the 64bit slots.

If you want to fool the SRM you can type "set pci_device_override" at
the SRM prompt. Just don't complain if your data mysteriously gets mangled.

The complete command is:

        set pci_device_override <vendor_id><device_id>
        e.g. set pci_device_override 88c15333

A more radical approach is to use:

        set pci_device_override -1

This disables PCI ID checking altogether, so that you can stick in any
random PCI card without its ID getting checked. Again: do this on your own
risk.

The kernel reports it when it sees a buggy Pyxis chip:
Sep 16 18:39:43 miata /kernel: cia0: Pyxis, pass 1
Sep 16 18:39:43 miata /kernel: cia0: extended capabilities: 1<BWEN>
Sep 16 18:39:43 miata /kernel: cia0: WARNING: Pyxis pass 1 DMA bug; no
bets...

A MiataGL probes as:
Jan  3 12:22:32 miata /kernel: cia0: Pyxis, pass 1
Jan  3 12:22:32 miata /kernel: cia0: extended capabilities: 1<BWEN>
Jan  3 12:22:32 miata /kernel: pcib0: <2117x PCI host bus adapter> on cia0

MiataGL does not have the DMA problems of the MX5. PCI cards that make
the MX5 SRM choke when installed in the 64bit slots are accepted without
problems by the MiataGL SRM.

The latest mainboard revisions of MX5 contain a hardware workaround for the
bug. The SRM does not know about the ECO and will complain about unknown
cards
as before. So does the FreeBSD kernel by the way.

</quote>

Comments are invited if people have more detail. I'm the writer/maintainer
of said file.

hope this helps,
Wilko
-- 
Wilko Bulte  	 	http://www.freebsd.org  "Do, or do not. There is no try"
wilko@freebsd.org	http://www.nlfug.nl	Yoda - The Empire Strikes Back