Attention is currently required from: fixeria.
pespin has posted comments on this change. ( https://gerrit.osmocom.org/c/libosmo-gprs/+/33770 )
Change subject: rlcmac: Introduce L1CTL-CCCH_READY.ind primitive ......................................................................
Patch Set 1:
(3 comments)
Patchset:
PS1:
Don't you need to modify the unit tests to send the new prim?
I'd say we don't need that. Those code paths are triggered by cleanup_test() which make sure trigger the PDCH_REL_REQ to go in the "clean tbfs" direction, since afterwards we are calling osmo_gprs_rlcmac_init() again to reset the whole library state. So adding extra code in cleanup_test() to trigger CCCH_READY.ind would be going in the direction of creating more context, which is not what we want there. If a specific test wants to test the returning from CCCH_READY.ind, it can do so explicitly.
File include/osmocom/gprs/rlcmac/rlcmac_prim.h:
https://gerrit.osmocom.org/c/libosmo-gprs/+/33770/comment/126ef377_7a3775fd PS1, Line 198: ccch_ready
Having empty structs is a bit weird, but ok.
I still like it because it quickly confirms that no extra fields are available/used.
File src/rlcmac/gre.c:
https://gerrit.osmocom.org/c/libosmo-gprs/+/33770/comment/64718e16_c1048695 PS1, Line 42: struct gprs_rlcmac_entity *gre = data;
Why not removing it? Does it make sense to keep it commented out?
I can remove it, I simply thought a reader can quickly find out the gre pointer is already available there.