8131 and 8151 in static.c

[ollie at lanl.gov]

> Li-Ta Lo <ollie at lanl.gov> writes:
>
>> Eric,
>>
>> Why there is no
>> 	struct southbridge_amd_amd8131_config
>> or
>> 	struct southbridge_amd_amd8151_config
>
> Because they don't need an enable_dev method.
>
>> in ter static.c although we have
>> 	chip southbridge/amd/amd8131
>> and
>> 	chip southbridge/amd/amd8151
>>
>> in the Config.lb?
>
> The chips are found by their pci_ids.
>
>> There is no
>> 	struct chip_operations southbridge_amd_amd8131_ops
>> nor
>> 	struct chip_operations southbridge_amd_amd8151_ops
>> nether.
>
> Currently struct xxxxx_config and struct chip_operations xxx_ops
> are tied together.  If you have you have both.
>
> We don't currently require a struct chip_operations.
>
> I have not thought enough about this to know if it is a good or
> a bad thing.  It is simply the way it was done and I have not changed it.
> If we always required this we could remove the config directive
> from the configuration language.
>
> Eric
>

How does the config tool tell if there is chip_operation or not ?
How does it know to generate struct xxx_config for northbridge
but not the southbridge?

Ollie