Merge branch 'NR_SA_F1AP_5GRECORDS' of https://gitlab.eurecom.fr/oai/openairinterface5g into NR_SA_F1AP_5GRECORDS

- first PUSCH containing RRCSetupComplete not programmed for reception in PHY for some reason. DCI_0_0 received by UE.
- if RA fails, on the second attempt there is usually a crash (sometimes in L2_interface.c when getting UE context (added AssertFatal on this now).

-RRCSetupComplete successfully transferred to RLC of the UE now through SRB1, but not at the MAC yet.
-Needed to enable some additional elements for RLC config. at the gNB side (asn1_msg.c) so that the addition of the SRB does not fail at the UE.
 To verify if these are really required.

removed reading reconfig.raw with --sa. Also a segfault upon reception of RRCSetup-CellGroupConfig when this is not in the RRC context.

Conflicts:
	cmake_targets/build_oai
	openair2/LAYER2/NR_MAC_UE/nr_ue_procedures.c
	openair2/LAYER2/NR_MAC_gNB/gNB_scheduler_RA.c

Merge branch 'NR_SA_F1AP_5GRECORDS' of https://gitlab.eurecom.fr/oai/openairinterface5g into NR_SA_F1AP_5GRECORDS

Merge branch 'NR_SA_F1AP_5GRECORDS' of https://gitlab.eurecom.fr/oai/openairinterface5g into NR_SA_F1AP_5GRECORDS

Conflicts:
	openair2/LAYER2/NR_MAC_UE/nr_ue_procedures.c

This reverts commit e06c7276042d2ce3a7956cc5d2e68e9b0a0b3bed

It is sufficient to pass l_d directly to the function but considering future extension to mapping type B, this way makes more sense.

Signed-off-by: Raphael Defosseux <raphael.defosseux@eurecom.fr>

Integration Branch : 2021 week 14

See merge request oai/openairinterface5g!1121