pciback error - what does it mean?

pciback error - what does it mean?

[Stefan Neuwirth]

I'm still playing around with xen-3.0.2-3 and the pciback.hide
option. Currently I'm trying to give the guest a pcmcia-to-pci card
with a prism54 wlan card in the slot. When starting the guest the
following message apears in the bootlog:

...
pciback 0000:02:09.0: can't export pci devices that don't have a
normal (0) or bridge (1) header type!
...

The lspci of the card:

0000:02:09.0 CardBus bridge: Ricoh Co Ltd RL5c475 (rev 81)
        Flags: bus master, medium devsel, latency 168, IRQ 16
        Memory at e1002000 (32-bit, non-prefetchable)
        Bus: primary=02, secondary=04, subordinate=07, sec-latency=176
        Memory window 0: 30000000-31fff000 (prefetchable)
        Memory window 1: 34000000-35fff000
        I/O window 0: 00004400-000044ff
        I/O window 1: 00004800-000048ff
        16-bit legacy interface ports at 0001

What does the message mean?

   Stefan

RE: pciback error - what does it mean?

[Ian Pratt]

> I'm still playing around with xen-3.0.2-3 and the pciback.hide
> option. Currently I'm trying to give the guest a pcmcia-to-pci card
> with a prism54 wlan card in the slot. When starting the guest the
> following message apears in the bootlog:
> 
> ...
> pciback 0000:02:09.0: can't export pci devices that don't have a
> normal (0) or bridge (1) header type!
> ...

Cardbus bridges have a "2" in this field. No idea what will happen if
you disable this test in pciback, but it's worth a go...

Ian

 

Re: pciback error - what does it mean?

[Keir Fraser]

On 28/8/06 9:26 pm, "Ian Pratt" <m+Ian.Pratt@cl.cam.ac.uk> wrote:

>> I'm still playing around with xen-3.0.2-3 and the pciback.hide
>> option. Currently I'm trying to give the guest a pcmcia-to-pci card
>> with a prism54 wlan card in the slot. When starting the guest the
>> following message apears in the bootlog:
>> 
>> ...
>> pciback 0000:02:09.0: can't export pci devices that don't have a
>> normal (0) or bridge (1) header type!
>> ...
> 
> Cardbus bridges have a "2" in this field. No idea what will happen if
> you disable this test in pciback, but it's worth a go...

It'll almost certainly just work. I'm not sure a test on device type is
really necessary.

 -- Keir

Re: pciback error - what does it mean?

[Stefan Neuwirth]

Keir Fraser <Keir.Fraser@cl.cam.ac.uk> writes:

> On 28/8/06 9:26 pm, "Ian Pratt" <m+Ian.Pratt@cl.cam.ac.uk> wrote:
>
>>> I'm still playing around with xen-3.0.2-3 and the pciback.hide
>>> option. Currently I'm trying to give the guest a pcmcia-to-pci card
>>> with a prism54 wlan card in the slot. When starting the guest the
>>> following message apears in the bootlog:
>>> 
>>> ...
>>> pciback 0000:02:09.0: can't export pci devices that don't have a
>>> normal (0) or bridge (1) header type!
>>> ...
>> 
>> Cardbus bridges have a "2" in this field. No idea what will happen if
>> you disable this test in pciback, but it's worth a go...
>
> It'll almost certainly just work. I'm not sure a test on device type is
> really necessary.

Just disable the test will lead to trouble in function
pciback_config_header_add_fields
(.../xen/pciback/conf_space_header.c). There are two structures called
header_0 and header_1 that are used by pciback_config_add_fields. So
someone decided to not implement header_2 - maybe there was a reason
for the decision...

  Stefan

RE: pciback error - what does it mean?

[Ian Pratt]

> >> Cardbus bridges have a "2" in this field. No idea what will happen
if
> >> you disable this test in pciback, but it's worth a go...
> >
> > It'll almost certainly just work. I'm not sure a test on device type
is
> > really necessary.
> 
> Just disable the test will lead to trouble in function
> pciback_config_header_add_fields
> (.../xen/pciback/conf_space_header.c). There are two structures called
> header_0 and header_1 that are used by pciback_config_add_fields. So
> someone decided to not implement header_2 - maybe there was a reason
> for the decision...

Looking at include/linux/pci_regs.h there's quite a lot of similarity
between a 'bridge' header type 1 and a 'cardbus bridge' header type2.
It may well work defining a header_2 the same as header_1 and adding it
to pciback_config_header_add_fields

Ian

Re: pciback error - what does it mean?

[Stefan Neuwirth]

"Ian Pratt" <m+Ian.Pratt@cl.cam.ac.uk> writes:

>> >> Cardbus bridges have a "2" in this field. No idea what will happen
> if
>> >> you disable this test in pciback, but it's worth a go...
>> >
>> > It'll almost certainly just work. I'm not sure a test on device type
> is
>> > really necessary.
>> 
>> Just disable the test will lead to trouble in function
>> pciback_config_header_add_fields
>> (.../xen/pciback/conf_space_header.c). There are two structures called
>> header_0 and header_1 that are used by pciback_config_add_fields. So
>> someone decided to not implement header_2 - maybe there was a reason
>> for the decision...
>
> Looking at include/linux/pci_regs.h there's quite a lot of similarity
> between a 'bridge' header type 1 and a 'cardbus bridge' header type2.
> It may well work defining a header_2 the same as header_1 and adding it
> to pciback_config_header_add_fields

Ok, everything back to start. It's not necessary to give the
card-bridge device to the guest system. I can use the pcmcia card in
the guest system just by giving the pci id of this card to the guest
and let the dom0 system init and manage the bridge - I'm really
suprised that it works this way.