Attention is currently required from: fixeria, neels.

pespin has posted comments on this change by pespin. ( https://gerrit.osmocom.org/c/osmo-upf/+/41482?usp=email )

Change subject: Validate session remote CP F-SEID and use it to transmit session requests ......................................................................

Patch Set 1:

(3 comments)

File src/osmo-upf/up_endpoint.c:

https://gerrit.osmocom.org/c/osmo-upf/+/41482/comment/2afd55c2_4ce685ab?usp=... : PS1, Line 176: OSMO_ASSERT(f_seid)

missing semicolon?

Done

https://gerrit.osmocom.org/c/osmo-upf/+/41482/comment/074f1a03_24ab73ee?usp=... : PS1, Line 217: nack_response

I would go for a separate function (e.g. `up_ep_tx_nack()`) to avoid code duplication.

I think it's fine having it this, way it's only 3 lines of code and it's common practice to handle nack this way imho.

It would make sense if a lot of code was needed, but it's not really the case.

File src/osmo-upf/up_session.c:

https://gerrit.osmocom.org/c/osmo-upf/+/41482/comment/c24cdc2b_eeb44a8f?usp=... : PS1, Line 115: return NULL;

This code is unreachable. […]

I also thought about it, and tbh I only added this return to avoid warnings on older compilers/static analysis tools. I prefer keeping the assert in each branch, specially the first one, since it becomes clearer that the break is not something which may not been done on purpose.

I'll add a comment here /* unreachable */, since adding more code like ASSERT could also trigger a warning by static analysis tools