more on the arima/hdama pci bus problem

[ron minnich]

the 8111 is being put on pci bus 1, but Linux never sees it.

I'm not totally clear yet on why this is happening, but I've got to work 
on a different problem with this board, so if anyone has any suggestions 
I'll try them out later.

thanks

ron

ron minnich  writes:
> the 8111 is being put on pci bus 1, but Linux never sees it.
>
> I'm not totally clear yet on why this is happening, but I've got to work 
> on a different problem with this board, so if anyone has any suggestions 
> I'll try them out later.

Hmm.  I will have to sync up with you.  Are you certain the pirq table
refers to a device on bus 1?

I have a couple odd ball issues but my internal version is starting to 
look pretty solid.  The internal guys won't give me bug reports anymore
because the bugs are few enough.

I have some serious broadcom NIC tracing to do and then I should be able
to sync with the public tree again.

Eric

[ron minnich]

On 9 Sep 2003, Eric W. Biederman wrote:
> Hmm.  I will have to sync up with you.  Are you certain the pirq table
> refers to a device on bus 1?

It does. The issue is that Linux does not enumerate any devices on Bus 1. 
LinuxBIOS, however, does, and the 1022:746b device shows up on bus 1 in 
linuxbios.

The PIRQ table in the public tree is wrong, need to fix it.

Linux itself enumerates bus 0, then jumps to bus 2.
> I have some serious broadcom NIC tracing to do and then I should be able
> to sync with the public tree again.

we're working on the same thing then. I'm not going to worry about this 
for present while I look at the tg3 problems.

ron

ron minnich  writes:
> On 9 Sep 2003, Eric W. Biederman wrote:
> > Hmm.  I will have to sync up with you.  Are you certain the pirq table
> > refers to a device on bus 1?
>
> It does. The issue is that Linux does not enumerate any devices on Bus 1. 
> LinuxBIOS, however, does, and the 1022:746b device shows up on bus 1 in 
> linuxbios.
>
> The PIRQ table in the public tree is wrong, need to fix it.
>
> Linux itself enumerates bus 0, then jumps to bus 2.

For that I cannot think of a reasonable explanation.  I will have to look
and verify this.  I am pretty certain I booted a kernel after I made those
changes..
> > I have some serious broadcom NIC tracing to do and then I should be able
> > to sync with the public tree again.
>
> we're working on the same thing then. I'm not going to worry about this 
> for present while I look at the tg3 problems.

Ok.  We might have to sync up then.

Eric

[ron minnich]

On 9 Sep 2003, Eric W. Biederman wrote:
> For that I cannot think of a reasonable explanation.  I will have to look
> and verify this.  I am pretty certain I booted a kernel after I made those
> changes..

are you on 2.4.21? I'll get the serial output (node is not on ethernet 
yet)

thanks

ron