8349 Commits (d950433cbd613f17e892308b4cba2038a41e1c90)
 

Author SHA1 Message Date
Francisco 60d5c6df7b implementation of memory pool with individual memory blocks per UE that get accessed via their rnti 4 years ago
Andre Puschmann db496df1e6 mac_nr: add support for periodic BSR reporting
building on the previous refactor this patch now adds support
for peridoic BSR reporting (using short BSR). It furthermore does
the following changes:

* add BSR packing
* add proc_bsr_nr unit test
* move mac_nr test code into test folder under src (needs to be done with other test code too)
4 years ago
Andre Puschmann 7188603f2e mac_nr: don't push MAC CE LCIDs to RLC when unpacking DL MAC PDU
this should fix log warnings then demux was trying to feed MAC CEs
to RLC:

10:50:54.845116 [MAC ] [I] [ 1153] Handling subPDU 1/2: rnti=0x0 lcid=61, sdu_len=1
10:50:54.845117 [RLC ] [E] Radio bearer id must be in [0:11] - 61
10:50:54.845118 [RLC ] [W] LCID 61 doesn't exist. Dropping PDU.
10:50:54.845119 [MAC ] [I] [ 1153] Handling subPDU 2/2: rnti=0x0 lcid=63, sdu_len=33
10:50:54.845119 [RLC ] [E] Radio bearer id must be in [0:11] - 63
4 years ago
Andre Puschmann 6dd8c21073 mac_sch_pdu_nr: add unpacking of TA CE in DL-SCH
also add basic test to verify correct unpacking
4 years ago
Francisco 9b7060a098 check ptr value before dereferencing 4 years ago
Francisco dc763b5045 rlc am - fix bugs in buffered_pdcp_pdu_list
The count and container elements were not being correctly updated in the events
of SN overwrite, clear(), clear_pdcp_sdu(sn) of sn that doesnt exist
4 years ago
Francisco ebbb74e639 rlc am - use of circular_map container in RLC AM window
The window elements are not being correctly cleaned up when clear() is called
or when overwritten by newer SN. Furthermore, the window count member is not being
correctly updated when the insertion of a newer SN overwrites the previous one.

I used the circular_map container to avoid this sort of bugs
4 years ago
Ismael Gomez c4069b58af Call HARQ new tti only for the current PID 4 years ago
David Rupprecht 82c9636156 Refactor pdcp ctor seperate the configure function 4 years ago
Andre Puschmann 1a42e0b1d1 epc,nas: add error handling for all make_byte_buffer() calls
not handling the case where allocation was failing was causing a
segfault. It seems that when doing high rate UDP iperf and
disconnecting the UE causes the buffer pool to become empty due
to the PDU buffering. Thus, causing a crash when attaching the UE again.
4 years ago
Andre Puschmann cbe7980d7d enb,mac_controller: apply new MAC config with disabled DRB
fix regression in RRC that updated the local MAC config to
disable DRB scheduling after maxRetx, but didn't actually apply the config
4 years ago
David Rupprecht 56e947bf27 Added functionality for releasing a NR bearer 4 years ago
David Rupprecht cc6d8128ad Added gw interface if it is running 4 years ago
David Rupprecht 7a481c40ab Added mac prach nr proc error handling 4 years ago
David Rupprecht f9ea02b6bf Added functions for rrc reestablishing procedure if nr reconfig failes 4 years ago
David Rupprecht 43ec714ce1 Added functionality for printing nr metrics 4 years ago
Xavier Arteaga 58a9ef6e7f Minor NR PHY tweaks and aesthetic 4 years ago
Xavier Arteaga db1a1c059c NR PHY workers TTI synchronization between DL and UL work 4 years ago
Xavier Arteaga 991c6e7016 Refactored NR HARQ-ACK feedback enqueue 4 years ago
Xavier Arteaga bd11b66b6c Fix unitialised NR PHY arguments 4 years ago
Xavier Arteaga 22239b29c5 Disable smoothing filter for NR PDCCH DMRS 4 years ago
Xavier Arteaga 9f54c6ae6d Fix NR PUCCH info string and memory leak 4 years ago
Francisco bb9601c13b gtpu - fix gptu manager members destruction order 4 years ago
Francisco b7d4a72ea1 adt - fix circular map clear() and erase(it) functions 4 years ago
Francisco 643c73acee gtpu - fix removal of gtpu tunnels when rnti is released 4 years ago
Andre Puschmann 7459a65a8e enb,rrc_ue: remove unused variable 4 years ago
Andre Puschmann 0fdcea2924 rlc_am: count every segment retx as a full retx for this SN
previously we were only counting retx if we retx the start of a segment.
this could lead to unwanted behaviour, i.e., not counting retx
correctly and thus not triggering the maxretx attempt, if the receive
always sends NACKs with a SO_start.

The RLC spec is not clear on how this should be handled correctly but
IMHO using an integer number of retx is reasonable, even for segments
that might be retransmitted more often.

The alternative of using a fractional retx counter that may be increamented
proportional to the segment size that is retx is another alternative
but considered too complex to implement (and test correctly).
4 years ago
Andre Puschmann c74661ae95 rlc_am: add log message when received NACK SN is already on retx queue 4 years ago
Francisco d02cc51e33 use mutex or atomic in running flag to signal the interruption of a thread 4 years ago
Francisco 89628b691f add thread sanitizer cmake flag 4 years ago
Xavier Arteaga 49731ab5a1 Initial NR-DCI refactor
Added NR DCI 1_1 size

Several DCI NR fixes
4 years ago
Xavier Arteaga dee2ec30cd SRSUE: use new MAC interface for NR SR procedure 4 years ago
Xavier Arteaga 01990c2e1d SRSUE: Add PHY-MAC SR related interfaces 4 years ago
Xavier Arteaga 64acf03158 Fix sch_nr_test 4 years ago
Xavier Arteaga f4ae2d73aa SRSUE: report positive ACK for retx 4 years ago
Xavier Arteaga 20a4c39fa4 SRSUE: Disable PDSCH retransmissions in NR 4 years ago
Xavier Arteaga 1dcd75a6c7 Ignore CRC for RV!=0 in phy_dl_nr_test 4 years ago
Xavier Arteaga 3504bfcf5a Fix NR-SCH test 4 years ago
Xavier Arteaga 39e97cbcf4 Fix unbounded memory access 4 years ago
Xavier Arteaga 29c6adad93 Avoid unused LDPC memory in AVX2 decoder 4 years ago
Xavier Arteaga 392a8cc1b6 Fix LDPC RM number of useful iterations 4 years ago
Xavier Arteaga 3568f633c3 Use LDPC RM number of useful LLR and rename functions 4 years ago
Xavier Arteaga ecd2c1cdcc Improved LDPC decoded performance and other NR aesthetical changes 4 years ago
David Rupprecht 506adca930 Set eutra nr band info based config value 4 years ago
Francisco cdf72248f3 Reimplement batch-based background object/memory pool
Main changes:
- addition of pool utilities
- The node size/alignment and batch allocation threshold are now runtime arguments
- object pool and memory pool are not anymore based on the same class.
  The object pool cannot use intrusive free list because it would overwrite the object
  memory
4 years ago
Francisco fbeb87c53e remove unused parameters from mac::ue 4 years ago
Francisco 62a6d9a8dc fix background memory pool destruction 4 years ago
Francisco b004c2024d integration of object pool in mac::ue for allocation of softbuffers. Definition of common object pool interface class 4 years ago
Francisco 460d7a8f4f fix memory pool test - placed the pool state into a shared_ptr so that the callbacks still have a valid handle when the pool is destroyed 4 years ago
Francisco 9bec13731a adt unit test - fix memory pool test 4 years ago