Attention is currently required from: laforge.

fixeria has posted comments on this change by fixeria. ( https://gerrit.osmocom.org/c/libosmocore/+/34427?usp=email )

Change subject: gsm_12_21.h: fix values for NM_IPAC_F_GPRS_CODING_MCS[1-9] ......................................................................

Patch Set 1:

(1 comment)

Patchset:

PS1:

well, it is probably simply a question of endianness of the underlying CPU, right? It's a 16 bit va […]

These flags are applied to a `uint32_t` value in **native** endianness for the host. nanoBTS and osmo-bts report a value in network byte order, which needs to be converted to host's byte order before doing any `u32 & NM_IPAC_F_` checks.

* osmo-bsc: see `ipacc_parse_supp_flags()` for a usage example, * osmo-bts: grep for `NM_IPAC_F_CHANT_`.

The problem this patch is solving is not about endianness, it's about wrong offset values assigned to MCS related `NM_IPAC_F_` constants. We should get this merged to fix parsing of the feature reports sent by nanoBTS. As a side effect, this will change the bit-vector sent by osmo-bts, but it's not critical because we never request GPRS related features from osmo-bts and so far only using this for informative purposes.