Attention is currently required from: fixeria.

pespin has posted comments on this change. ( https://gerrit.osmocom.org/c/osmo-bts/+/32196 )

Change subject: nm: Drop NM_EV_SETATTR_{ACK/NACK} ......................................................................

Patch Set 3:

(3 comments)

File src/osmo-bts-lc15/oml.c:

https://gerrit.osmocom.org/c/osmo-bts/+/32196/comment/763dfb56_8f946bcc PS3, Line 1851: rc = 0;

This is weird. […]

It follows the logic of "for that object/msg type simply ack it", see below.

https://gerrit.osmocom.org/c/osmo-bts/+/32196/comment/8c47833b_6c0857b0 PS3, Line 1854: rc = 0;

So it's fine if we get an unknown/unhandled `foh->msg_type` value?

yes, this function is called for several obj_classes, so we only have specific code for those where we need to do something in lower layers, the rest we simply return 0 (ACK it).

I know we could list all the messages here and the code could be better, but I cannot simply improve it all in one go, there's a lot of stuff which needs moving, etc, as you can see in this set of patches I'm submitting. I'm doing this in an iterative way since the code paths are a bit complex and used by different objects, different BTS types, etc.

File src/osmo-bts-virtual/bts_model.c:

https://gerrit.osmocom.org/c/osmo-bts/+/32196/comment/db027969_2217bccc PS3, Line 148: rc = 0;

Are you sure 0 is the right value here? Not `-ENOTSUP`?

Yes, it's fine for now, since ideally we should be calling this for more object types.