newbus: Update docs on child_pnpinfo_str and child_location_str

[imp]

Summary

Only one driver in the tree with a non-trivial child_pnpinfo_str (acpi_iicbus)
returns EOVERFLOW. None do for location_str. Since it's not really an an
observed interface in the other 34 instances; and since nothing does anything
with this return value. Remove it from the API contract.

[imp]

Sadly, there appears to be no man page for either of these functions.

[jhb]

In D29791#668246, @imp wrote:

Sadly, there appears to be no man page for either of these functions.

I might know which person added the functions in the first place. :)

My only thought about this change is that it no longer describes the return value at all, though perhaps you just want to remove the EOVERFLOW check. Should we also remove the check from the one driver as part of this? (And what do other drivers do, do they truncate? You were going to change it to write to an sbuf I thought at one point?)

[imp]

In D29791#670956, @jhb wrote:

In D29791#668246, @imp wrote:

Sadly, there appears to be no man page for either of these functions.

I might know which person added the functions in the first place. :)

Yes. back in my wild, carefree youth I may have had occasion to add kernel interfaces w/o man pages, much to my present-day shame :(.

My only thought about this change is that it no longer describes the return value at all, though perhaps you just want to remove the EOVERFLOW check. Should we also remove the check from the one driver as part of this? (And what do other drivers do, do they truncate? You were going to change it to write to an sbuf I thought at one point?)

Yes. I'd planned on documenting then, and cleaning up that one case because it won't be relevant.

[imp]

I'm going to punt on this and instead do the full conversion.