I forgot to say what was interesting in the logs:

I like to do side by side diff, but to each their own.

no_pnp is the log without my patch.
1. Notice that it creates an enabled dynamic device for each function, even the disabled ones.
2. Notice the resource allocation bases. Because the PNP devices get included in resource allocation, the IO base and limit are broken.
3. There are lots of incorrect resources, because no values are taken from the dts.

Thanks,
Myles