Re: Are PCI bridges automatically configured?

[Peter Dufault <dufault at hda dot com>]

On Feb 25, 2005, at 3:44 PM, Kate Feng wrote:

> Peter Dufault wrote:
>
> > On Feb 25, 2005, at 3:04 PM, Joel Sherrill <joel@OARcorp.com> wrote:
> >
> > > On a board with a nice Rom monitor like PPCBug, it is
> > > quite possible that it was initialized by the Bug.
> >
> > If it was enabled by MOTLoad (the monitor on the MVME5500), then it 
> > has
> > a bug similar to RTEMS mvme5500 (arbitrarily limiting number of 
> > busses)
> > since the bus and device don't show up with pciShow (and other pciFoo)
> > commands in MOTLoad, but after increasing the number of busses in the
> > RTEMS source it is showing up in RTEMS using BSP_pciFindDevice.
>
> Good to know I am not the only one who limited it.
>
> > I'm still not sure who enabled the bridge.
>
> You did it.
> Any PMC device should have firmware to follow the PCI
> standard to respond to a PCI  bridge search.  You just enabled
> the search by increasing the limit of two.
>
> Regards,
> Kate

Here's another note about the mvme5500 "BSP_pciFindDevice" in case you 
start looking at it.  I'll look at it after I finish the drivers I'm 
writing if you don't get to it.

I've got two Acromag PMC730 PMC cards in the MVME5500 board's two PMC 
slots, and during startup both are found and configured by RTEMS.  The 
debug output during RTEMS startup is (PCI_DEBUG is enabled):
PCI0 Bus0 Slot6 DeviceID 0x445715D5
PCI0_BASE_ADDRESS_0 0x80200000
...
PCI1 Bus0 Slot6 DeviceID 0x445715D5
PCI1_BASE_ADDRESS_1 0xE1100000
...

However, when I use BSP_pciFindDevice to try to access that second 
instance it doesn't find it, so there is some inconsistency between the 
startup code and BSP_pciFindDevice.

So the good news is the card I thought I'd have trouble accessing (the 
one on the other side of the previously unsupported PCI-PCI bridge for 
the PMCSPAN) is now accessible, but the bad news is that the card on 
the MVEM5500 that MOTLoad sees and that is found during startup isn't 
accessible.  I didn't see a quick fix, so I'm back to developing the 
drivers then I'll look at this.

Thanks,

Peter Dufault
HD Associates, Inc.