Codebot
8c3332f894
Merge branch 'next' into agpl_next
...
# Conflicts:
# srsenb/hdr/stack/upper/pdcp_nr.h
# srsenb/hdr/stack/upper/rlc_nr.h
# srsenb/src/stack/upper/pdcp_nr.cc
# srsenb/src/stack/upper/rlc_nr.cc
3 years ago
Andre Puschmann
bf2606f94e
ue,mac_nr: fix race condition by using thread-safe RNTI class
3 years ago
Ismael Gomez
c09f09b7be
Move rrc_common from srsue to lib
3 years ago
Fabian Eckermann
8830c2796f
Add support for NSA FDD on band n3, n5 and n7.
3 years ago
Andre Puschmann
bed4cfb5e5
ue,mac_nr: fix generation of padding BSRs
...
* in Msg3 don't add explicit SBSR (rely on padding BSR to add)
* fix padding BSR generation to always use padding bytes to add it
3 years ago
faluco
fa351cd285
Rename functions that contain the unsafe word to nolock.
...
The term "unsafe" does not clearly state that the function is not locking the shared date and it sounds "dangerous", so use a more explicit term.
3 years ago
Codebot
ecd0580b9e
Merge branch 'next' into agpl_next
3 years ago
Bedran Karakoc
43915550cc
NGAP GTP Bearer Establishment
3 years ago
David Rupprecht
3712983308
Added PDU Session est request
3 years ago
David Rupprecht
c670383ae9
Added NAS 5G functions incoming msg and reg proc
3 years ago
faluco
55f5f74ff3
Fixed case where calling gw::setup_if_addr after ue attachment would spawn a new thread without closing the current one.
...
This would end up being a resource leak as the number of threads grow forever until the application crashes.
3 years ago
Andre Puschmann
26cc6d7886
ue: remove stack type parameter
...
for the moment we always use the LTE stack (also for NSA mode).
For SA we likely pick a smarter approach to pick the stack type, i.e.
using the ARFCN specified in the config.
3 years ago
Ismael Gomez
ad646f436c
Ignore PDSCH during cell selection
3 years ago
Xavier Arteaga
de00b80228
Review NR NSA CFO estimation and compensation
3 years ago
Codebot
b87b760326
Merge branch 'next' into agpl_next
3 years ago
faluco
c2b705c5ce
Fix several data races in proc_phr, class needs to be fully protected since most member variables are accessed by different threads.
3 years ago
faluco
6b07e886b7
Fix data race in is_first_ul_grant.
3 years ago
faluco
1e3ba607ce
Fix data race in is_pending_tx_end.
3 years ago
faluco
7dae4a491e
Fix data race in ue_stack_lte::ul_dropped_sdus.
3 years ago
faluco
1796e12084
Fix data race in ul_harq_entity.
3 years ago
faluco
2b73beb3dd
Fix data race in ul_harq.
3 years ago
David Rupprecht
e8a464228d
Cleaned string helpers
3 years ago
Andre Puschmann
ab6210b1fc
ue,proc_sr_nr: add mutex to protect access from PHY and Stack
...
also detected by TSAN
3 years ago
Andre Puschmann
2248e63219
ue,proc_ra_nr: add mutex to protect access from PHY and Stack
...
detected with TSAN
3 years ago
Andre Puschmann
cc2d71183e
ue,phy,nr: fix race condition when setting UL grant for Msg3
...
the Msg3 UL grant requires the TTI in which the RAR has been received
to calcualte the correct timing. There was a race between PHY and Stack
thread.
This patch circumvents the issue by removing a PHY state member that only holds
the RAR Rx timing. In the new interface the RA proc passes the Rx TTI
to the PHY again when setting the UL grant so the PHY can calculate the
correct timing without any state.
3 years ago
Codebot
958542dc2b
Merge branch 'next' into agpl_next
3 years ago
Andre Puschmann
f382a7c7db
Revert "ue,proc_ra_nr: add temporary flag to skip RAR reception"
...
This reverts commit 28668aac34
.
3 years ago
Andre Puschmann
5a936d9940
ue,rrc: handle DRB removal correctly
...
when RRC receives a command to remove a DRB, it also needs to delete
the bearer at RLC and PDCP
3 years ago
Andre Puschmann
54ab64327a
ue,mac_nr: move BSR levels lookup table into lib header
...
prepare for shared use with gNB
3 years ago
David Rupprecht
a9a23214b0
Added more 5G security kdfs
...
Added new 5G SIM functions
Added const parameters for security functions
3 years ago
David Rupprecht
a645d74eaf
USIM functions for res star
...
Added gen_res_star in security.h
Added USIM interfaces for res star
Added RES star test case
3 years ago
Codebot
de59c88954
Merge branch 'next' into agpl_next
...
# Conflicts:
# srsue/hdr/stack/bearer_manager.h
# srsue/src/stack/bearer_manager.cc
3 years ago
Francisco Paisana
f4df55c245
stack lte: separate bearer manager into two classes, one for the ue and one for the enb
3 years ago
Andre Puschmann
db8b710442
bearer_manager: move to lib folder and extend for multiple users
...
prepare bearer manager to be used by eNB which requires
to support multiple RNTIs
3 years ago
Andre Puschmann
40499e6b6d
ue_stack_lte: use std::atomic for running var
3 years ago
Andre Puschmann
79cdc28015
nas_test_common: call NAS layer from stack thread context
...
since NAS layers insn't thread safe, all calls must come from
the same thread
3 years ago
Andre Puschmann
78d60bc138
nas: remove all uses of atomic from 4G and 5G classes
...
the NAS classes aren't thread safe and thread-safety shouldn't be
pretended by using atomics. Remove them and add explicit notice.
3 years ago
Codebot
522ca4b43a
Adding AGPL copyright to new files
3 years ago
Codebot
d1d38d1826
Merge branch 'next' into agpl_next
3 years ago
Xavier Arteaga
8aa7ae7d95
Remove unused attribute
3 years ago
Xavier Arteaga
4cb633e6e6
UE dicards NR CFO measurement if SNR is negative
3 years ago
Xavier Arteaga
83b9b2d1ff
Implemented UE open loop CFO compensation for NR
3 years ago
Ismael Gomez
7839ab09dc
Add RLF-Reportv9 to srsUE
3 years ago
Codebot
83f103fe4b
Merge branch 'next' into agpl_next
3 years ago
David Rupprecht
3217c00cfc
Adjusting NAS base security function to 5G
3 years ago
Codebot
635d0bc6e9
Merge branch 'next' into agpl_next
3 years ago
Xavier Arteaga
cba6df3722
Extended NR PHY robustnes against wrong RRC configuration
3 years ago
Andre Puschmann
28668aac34
ue,proc_ra_nr: add temporary flag to skip RAR reception
...
this will be removed again as soon as the eNB supports full RAR
transmission.
3 years ago
Andre Puschmann
99dc94ab38
nas: make state variables atomics
...
NAS states and substates maybe be requested from other threads so
they need to be protected.
Note that the caller still needs to hold it's own mutex if different
actions are required based on the state.
3 years ago
Xavier Arteaga
b57df4db10
Refactored worker_end mechanism for concurrent workers
3 years ago