added harq round for Downlink, before harqround was only tracked when retransmissions happened, this lead to confussions in the database after post processing

added if condition to U mac.demuxed-- since there were packages with 3 bytes, that didn't contain any data just header which lead to mismatches

integration_2023_w37

* !1909 Rework for aarch64 support after SIMDe integration
* !1995 Fix packing/unpacking for nFAPI P5 messages
* !2228 TDLA, TDLB and TDLC channel models in DLSIM
* !2327 Bugfixes and some rewrite for nr rlc am control PDU reception
* !2331 NR UE HARQ improvements to handle more than 4 per PUCCH
* !2286 Service Models integration - KPM v03.00, GTP, MAC, RLC and PDCP
* !2336 fix(ci): HTML reporting for Ubuntu build fix
* !2338 nr-softmodem: fixed invalid length calculation, leading to segfault
* !2326 Tutorials: Update NR SA Tutorials to be compliant with OAI CN5G pre-2.0 release
* !2334 NR UE DL channel estimation fix
* !2325 Compute ULSCH TB CRC

Laurent noticed it on one of his MR and confirmed with him:
  -- Volumes have be created the 1st time they ran on avra on July 7th
  -- Volumes were created with not only the raw files but also the nr-softmodem executable

Solution: two-fold:
  1. using the `docker-compose down -v` syntax, the volume is deleted automatically
  2. I added redundancy with the custom command to remove the used volume
Signed-off-by: Raphael Defosseux <raphael.defosseux@eurecom.fr>

Merge remote-tracking branch 'origin/dreibh-trx_usrp_read-memory-access-fix' into integration_2023_w37

Merge remote-tracking branch 'origin/raphael/ci-fix-ubuntu-build-log-scheme' into integration_2023_w37

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

- when CU is used, UE ID data is filled as GNB_UE_ID_E2SM with
  CU UE ID value included

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

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

fix(ci): docker_log_split is useless with current build scheme. Just copying all txt files from ran-build image
Signed-off-by: Raphael Defosseux <raphael.defosseux@eurecom.fr>

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

- Fix 4G and 5G discrepancy in TLV padding requirements: 4G does not
  need padding, whereas 5G needs, so make separate set of function for
  padding for NR
- Fix packing/unpacking for (n)FAPI PARAM and CONFIG.request/response
  messages
- Handle message padding to next 32-bit boundary, as per SCF
  specification in packing and unpacking functions
- Account for message header size upon receiving message in PNF and VNF

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

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

  can compile the target ldpc_generators and generate_T with cross-compiler