Resolved conflict in Merge remote-tracking branch 'origin/episys/master-nsa' into episys/david/nsa_ci_debug_for_sa

MR !1233 : NR gNB MAC scheduler BWP start issue fix
This branch fix incorrect use of BWP start at gNB MAC scheduler (in DLSCH and ULSCH allocation and VRB map filling)

MR !1232 : SR/BSR support for NR UE
-added the support of scheduling request in NR UE. 
-added the support of regular, periodic and padding BSR in NR UE. 
-removed memcpy call in nr_ue_get_sdu(). 
-called mac_rlc_status_ind only for existing bearers. 
-Moved NR_LONG_BSR_TABLE and NR_SHORT_BSR_TABLE to nr_mac_common.c

MR !1239 : NR SDAP gNB - Transparent Mode

MR !1257 : Nr ldpc cuda softmodem
Move ldpc decoder accelerated in cuda GPU in a shared lib compatible with oai loader. 
libldpc_cuda.so is built as a ldpctest, nr-sotmodem, nr-uesoftmoden and ocp-gnb targets when cuda development environment is detected. 
At runtime --loader.ldcp.shlibversion _cuda option can be used to select the cuda accelerated ldcp library, except for the ldpctest exec where the -G option has been preserved. Also start implementing an opencl version of the ldpc decoder, starting "translating" the cuda version.

MR !1296 : Apply the NSA t_Reordering value also for SA
Apply the NSA t_Reordering value also for SA to improve the iperf downlink bitrate. The same iperf performance is expected at both NSA and SA

MR !1293 : hotfix SEGV and divide by 0
With latest develop, two bugs appear as long as we don't hide them with gcc -O2 optimisations

MR !1259 : gNB nrUE SA: improve security
Improves security and addition security in nrUE side

MR !1260 : NR CSET0 BWP Start fixes
Fixes a bug in selecting the start of BWP for Coreset0

Use Magma MME in CI tests
Update NSA 2x2 test

Merge NFAPI changes to new SA master branch

See merge request aburger/openairinterface5g!79

Instead of resetting queue when full,
take last item out.

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

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

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

In the previous commit, the NRUE would record
the time it received a slot and compare this
rx time to the next slot. If the next slot
came in before it had been 500us, the NRUE
would sleep a delta amount to make the duration
between the two slot indications 500 us. This
change complicated the code and did not
improve the performance so it was removed.
Furthermore, this commit removes the old
queueing mechanism of the slot indications and
uses calloc rather than an array.