Andre Puschmann
f992350d2d
rlc_um_nr_pdu_test: add testcases for 12bit SN length packing/unpacking
4 years ago
AlaiaL
1cc2cc4e7f
Fixed typo
4 years ago
AlaiaL
fd3c2b855e
- Renamed log to logger.
...
- Added a log entry in the metrics when skipping a measurement due to
time constrains
4 years ago
AlaiaL
bbaebd1274
Added a warning in the log when the cpu metrics are not registered due
...
to the number cpu cores is greater than supported.
4 years ago
AlaiaL
c9d1c77e8c
- Deleted the magic number of maximum supported cores, now it is a
...
constexpr.
- When the system has more cores that the maximum supported, the cpu
metrics will not be registered.
4 years ago
AlaiaL
6cd9bba641
Added the cpu metrics to the system metrics.
4 years ago
Xavier Arteaga
5bdf2c93b3
Initial NR PDSCH CSI-RS RE skip and other changes
4 years ago
Pedro Alvarez
5bc55ec48c
Changed discard_if to appyly_if in circular_buffer.h.
...
This was done so it would work when circular buffer holds other things
that are not unique_pointers. Queue and pop_func had to be made public
to be able to call the pop_func when an SDU is discarded.
4 years ago
Andre Puschmann
1b1cfa40c9
rlc_am_lte: add some more sanity checks around retx of PDUs
...
this patch doesn't change any functionality but adds a bit
of further error checking around the handling of status PDUs
and the retx of PDUs
4 years ago
Francisco
6d9709fe54
adt - make circular buffer work with types without default ctor
4 years ago
Francisco
ef5329bec0
adt - extend circular map unit test to test the container with move-only types
4 years ago
Xavier Arteaga
7704c09ce2
Fix typos
4 years ago
faluco
e8395c7474
- Implement a pool to store dyn arg store objects in srslog to avoid allocations when passing arguments to the backend.
...
- Use string views inside the log formatters to avoid allocations.
- Avoid a std::string when using fmt::vsprintf, instead favour fmt::vprintf.
4 years ago
Andre Puschmann
1a0891df51
all: check return value of make_byte_buffer() and handle them safely
4 years ago
Francisco
46d46e70ff
add documentation for srsran_assert and srsran_expect macros
4 years ago
Francisco
dd336c53ea
changed srsran_warning to srsran_expect. Applied new macro to rlc am
4 years ago
Francisco
b5692037a2
created macros for assertions and warnings. The assert macro prints the whole callstack on crash. The warning macro can just log a warning or call an assertion fail depending on the cmake flag STOP_ON_WARNING
4 years ago
Andre Puschmann
9d8d897d3e
mac_pcap_base: log warning when PDUs can't be added to write queue
4 years ago
Andre Puschmann
26e7964694
mac_pcap_base: increase PDU queue size to 512
4 years ago
Pedro Alvarez
d91119baf6
RLC discard ( #2515 )
...
* Added ability to discard to dyn_block_queue
* Change way of keeping track of SDUs
* Check nullptr in poping callback. Starting to check for nullptr in RLC read_pdu.
* Adding RLC discard tests
* Clearing PDCP info when RLC discard happens
* Read SDUs until they are no longer nullptr
* Changed discard_if to use template argument
4 years ago
Francisco
0d5038dd34
fix wrong lib linking in cmaker
4 years ago
Francisco
88f5add1fb
adt - fix circular map erase function and added extra asserts
4 years ago
Francisco
5d676199ab
adt bugfix - fix copy/move ctor of circular map
4 years ago
Francisco
4869b14de3
application of new circular map to mac::ue_db
4 years ago
Francisco
98d0637ded
adt addition - implemented reusable circular map
4 years ago
Codebot
4523ee6087
rename srsLTE to srsRAN
4 years ago
David Rupprecht
ad3a3af490
Adding more info into pcap test
4 years ago
Francisco
977c194cbc
gtpu,bugfix - handle the case when gtpu fails to allocate buffer for end marker. Also, added a timer that when expired, it autoremoves the GTPU handover tunnel
4 years ago
Pedro Alvarez
dde8157bf1
Make sure that S1 Setup request is restarted regardless of which error made the procedure fail.
4 years ago
Xavier Arteaga
2782d96170
SRSUE: compute speed from TA commands
4 years ago
Andre Puschmann
47654af717
pdcp_entity_lte: fix printf formatter for size_t
...
we've seen a heap-buffer overflow in fmt because printf wasn't using
the right formtter for size_t, which should be %zu
this patch fixes it for the PDCP LTE entity but we might have it elsewhere too
[1m[31m==7595==ERROR: AddressSanitizer: heap-buffer-overflow on address 0x629000e6f1fc at pc 0x562273a45289 bp 0x7f35567641f0 sp 0x7f35567641e0
[1m[0m[1m[34mREAD of size 4 at 0x629000e6f1fc thread T12 (STACK)[1m[0m
0 0x562273a45288 in fmt::v7::basic_format_arg<fmt::v7::basic_printf_context<std::back_insert_iterator<fmt::v7::detail::buffer<char> >, char> > fmt::v7::detail::make_arg<fmt::v7::basic_printf_context<std::back_insert_iterator<fmt::v7::detail::buffer<char> >, char>, unsigned int>(unsigned int const&) (/osmo-gsm-tester-srsue/srslte/bin/srsue+0x9dc288)
1 0x562273a3aa86 in void fmt::v7::dynamic_format_arg_store<fmt::v7::basic_printf_context<std::back_insert_iterator<fmt::v7::detail::buffer<char> >, char> >::emplace_arg<unsigned int>(unsigned int const&) (/osmo-gsm-tester-srsue/srslte/bin/srsue+0x9d1a86)
2 0x562273a308e7 in void fmt::v7::dynamic_format_arg_store<fmt::v7::basic_printf_context<std::back_insert_iterator<fmt::v7::detail::buffer<char> >, char> >::push_back<unsigned int>(unsigned int const&) /mnt/data/jenkins/workspace/srslte_ogt_trial_builder_x86-ubuntu1804-asan/srsLTE/lib/include/srslte/srslog/bundled/fmt/core.h:1548
3 0x562274361541 in void srslog::log_channel::operator()<unsigned int&, unsigned int&, unsigned long>(char const*, unsigned int&, unsigned int&, unsigned long&&) /mnt/data/jenkins/workspace/srslte_ogt_trial_builder_x86-ubuntu1804-asan/srsLTE/lib/include/srslte/srslog/log_channel.h:101
4 0x56227430d9e7 in srslte::pdcp_entity_lte::update_rx_counts_queue(unsigned int) /mnt/data/jenkins/workspace/srslte_ogt_trial_builder_x86-ubuntu1804-asan/srsLTE/lib/src/upper/pdcp_entity_lte.cc:451
4 years ago
Francisco
fb48f4484c
sched refactor - removal or cleanup of unused sched methods
4 years ago
Andre Puschmann
2cfc657fbb
rlc_am_lte: refactor logging when PDCP SDUs are enqueued
...
the patch refactor the logging when a new PDCP SDU is enqueued for
transmission at RLC.
If the SN is already present, only a warning is logged. From the RLC
perspective operation continues and the SDU will be transmitted.
The patch also changes the order of logs. When the SN cannot be inserted
inside the queue of undelivered SDUs, only one message is logged.
4 years ago
faluco
358c4de0bd
Fix a valgrind issue in a unit test in log_channel_test where the buffer was not being null terminated.
4 years ago
faluco
eb27efd867
Modify the policy of the buffered sink to flush its contents continuously instead of doing it once.
4 years ago
Francisco
0d91802495
sched optimization - swap c-arrays for bounded_vector in sched_interface to reduce time performing bzero/memcpy in the scheduler
4 years ago
Xavier Arteaga
5f954ab379
Fix UHD compilation
4 years ago
Xavier Arteaga
a8b5b8efa1
UHD: stability improvements
4 years ago
Xavier Arteaga
f9689ec956
Added PUSCH CFO estimation
4 years ago
Francisco
47f1175502
[sched, feature enhancement] - allow variable nof prbs allocated per SIB / Paging message
4 years ago
Andre Puschmann
3e07767f28
rlc_stress_test: set maxRetx for AM to 32 by default
...
since the tests are random we sometimes hit the maxRetx threshold
with 8 retx allowed that made the unit test fail.
increase threshold to lower the likelihood for this to happen.
4 years ago
Xavier Arteaga
f3c0360842
Applied aesthetical changes
4 years ago
Xavier Arteaga
d9780e1f13
Fix duplicated NR-DCI in different search spaces and CORESET and other aesthetic modification
4 years ago
Xavier Arteaga
6fa33c890d
Simultaneous HARQ-ACK, SR and CSI in NR-PUCCH
4 years ago
Xavier Arteaga
1463b11bad
Fix compilation
4 years ago
Xavier Arteaga
be8e8cbe3e
Fix radio decimation Rx stream stall
4 years ago
Xavier Arteaga
1153555ae1
SRSUE: Added UE specific cell space
4 years ago
Xavier Arteaga
10af89fcdd
NR-PDCCH aesthetic changes
4 years ago
Xavier Arteaga
fb1a0b6bce
SRSUE: added NR-PDCCH information
4 years ago
Xavier Arteaga
d3cfb00211
SRSUE: Added TDD DL/UL slot detection
4 years ago
Xavier Arteaga
f3bf0c1c68
Wrap slot index in modulus macro
4 years ago
Xavier Arteaga
4bab0b53b7
SRSUE: fix slot index overflow
4 years ago
Xavier Arteaga
6f6c20e194
SRSUE: Added procedure for multiplexing UCI in NR-PUSCH
4 years ago
faluco
e43f555fc0
- Provide a fast path for logging pre-formatted strings that are small and fit in 64 bytes without any allocations.
...
- Use this new feature in the tracer.
4 years ago
Francisco
2e18827259
avoid accessing C-array via &[0] for initialized buffer
4 years ago
Francisco
80104c7e65
adt, bugfix - fix assert when accessing bounded_vector::data() member
4 years ago
David Rupprecht
be46806495
PCAP: Enable carrier ID TAG
4 years ago
Francisco
ca7fe1349e
remove unnecessary sfinae from bounded_vector. Use bounded_vector for UE cc buffers. Set ue cc used buffers remove_pdu method to private
4 years ago
Francisco
d27e0be609
created special class to manage mac::ue currently allocated rx UL buffers. This class avoids mallocs
4 years ago
Ismael Gomez
7b33c48fe7
Use static queue in pdu_queue
4 years ago
Pedro Alvarez
f02694dd3e
PDCP status report: Fix issue in keeping track of Rx'ed COUNTs after eNB reestablishment. Fixed issue in logic that limits the size of the rx_counts info vector.
4 years ago
Pedro Alvarez
d2ef541957
Changed the structure to store rx_counts info to std::vector, to reduce
...
memory allocations.
Update the rx_count_info queue if the queue size is too large.
4 years ago
Pedro Alvarez
3f79cd6281
Changed generation of status report to use rx_counts instead of
...
undeliverd_sdus_queue. Added a queue to store information about
rx_counts received.
Added unit test for when the SNs wrap-around in status report genaration
4 years ago
Francisco
6159cb3817
add assert to ensure a valid ptr is passed to memory pool deallocator
4 years ago
Francisco
20928651c7
created batch allocator that leverages background worker pool
4 years ago
Francisco
3e9f93eb8a
refactor - remove old log_filter and logmap libraries from the codebase
4 years ago
Francisco
3b491ab06b
optimize .count() method of bounded_bitset to leverage popcount special instructions. Confirmed to work for gcc in -msse4 flag is passed.
4 years ago
Francisco
7dcb703d06
adt lib,bugfix - fix bounded_bitset resize to clear bits outside of mask correctly
4 years ago
Francisco
d1483dc0f8
sched,bugfix - fix bitmask formatting. fix msg3 adaptive retx pdcch allocation
4 years ago
faluco
173dcdd421
- Import latest srslog version.
...
- Remove string allocs in some log lines in radio.cc.
Add new tracing buffer capcity options into srsenb and srsue.
Add missing file.
4 years ago
faluco
c7542daf43
Remove rlc traces and add new ones into tti run functions both in the enb and ue.
4 years ago
faluco
82db6544fb
Add instrumentation points to rlc_am_lte::rlc_am_lte_tx::handle_control_pdu using srslog.
4 years ago
Ismael Gomez
8e9d28e7e0
Change wait_for interface for wait_until in circular_buffer
4 years ago
Andre Puschmann
16de8668e0
rlc_am_lte: convert mutexes to std::mutex
...
replace all pthread_mutex with std::mutex and use lock_guard and unique_lock (where needed)
4 years ago
Andre Puschmann
20cbc48f90
rlc_am_lte: refactor PDCP/RLC interface to use fix sized data structure
...
this patch replaces the std::vector type used in the interface between
PDCP and RLC to signal delivery status between both layers. The new
data type is a configurable but fixed-size vector.
The RLC AM doesn't need to dynamically allocate the vector for every SN but
uses the tx_window for storage.
4 years ago
Francisco
ee77343f26
fix compilation issue for centos7
4 years ago
Francisco
20e341d0a1
mac,bugfix - fixed formatting of MAC and scheduler user-defined types
4 years ago
Francisco
f72cd4151c
rlc am,bugfix - fixed formatting of RLC AM header and status pdu
4 years ago
Francisco
4c1067bcf6
bitset - fix compilation issue for centos7
4 years ago
Francisco
0f9d73012f
use srslte::move_callback instead of std::function in timers to avoid allocations
4 years ago
Francisco
5a1bf28fe1
optimization - minimization of number of std::string allocations for logging
4 years ago
Francisco
1ffc4cef86
rlc am,optimization - change helper log methods to avoid mallocs
4 years ago
Francisco
d77b6e1d9c
sched,optimization,refactor - use of custom formatter that avoids mallocs for getting bitmasks strings in hex and binary formats
4 years ago
Xavier Arteaga
a03c78a777
Fix NR-PUSCH defect
4 years ago
Xavier Arteaga
05f4d6af71
Fix and validated UCI multiplex in NR-PUSCH
4 years ago
Xavier Arteaga
2b9bd1173e
More NR-PUSCH UCI bit multiplexing
4 years ago
Xavier Arteaga
289fff9c22
Removed carrier set in UCI NR module
4 years ago
Xavier Arteaga
97435b085e
Added CSI part1 and 1/2 bit HARQ-ACK multiplexing on PUSCH
4 years ago
Xavier Arteaga
dc2542901a
Verified HARQ-ACK in PUSCH
4 years ago
Xavier Arteaga
9dffad87f2
Initial UCI bits multiplexing in PUSCH
4 years ago
Xavier Arteaga
f4e9d00ea8
Initial HARQ-ACK multiplex in PUSCH
4 years ago
Xavier Arteaga
7b588eafec
Minor aesthet change
4 years ago
Andre Puschmann
ace8764568
rlc_am_lte: add missing unlock when building status PDU fails
4 years ago
Francisco
0b6293c676
adt lib additions - add the ability to perform timedwait for popping from a blocking queue
4 years ago
Francisco
f0ed1e06a8
documentation,bugfix - added documentation to new blocking queues, changed class names, and removed blocking pushes from the critical path
4 years ago
Francisco
d1236fd62f
stack,optimization - replaced previous block_queue design for new bounded queue in several places in the enb
4 years ago
Francisco
28ef5833a2
fix compilation issue for centos7
4 years ago
Francisco
0ba93d274f
converted byte_buffer_queue to use new circular buffer-based queue to avoid mallocs
4 years ago
Francisco
822e26b63f
implemented circular buffer-based queue
4 years ago
Francisco
fb278a8b21
Fallback to default R8 TBS tables for format 1A DCI
...
Co-authored-by: xavierarteaga <xavier.arteaga@softwareradiosystems.com>
4 years ago
Andre Puschmann
83f24fbf77
rlc_am_test: fix max_retx test
...
the test should only check the correct signaling of the maxretx
to the upper layer.
4 years ago
Andre Puschmann
ceacf8508d
rlc_am_lte: fix handling of maxretx
...
this patch fixes the actions/handling after RLC detected
maxRetx reached for a given SN.
According to the TS, RLC should only inform upper layers and
not try to recover from the event itself.
As a consequence, we won't manipulate the Tx or Rx window.
As a result of this, we might retransmit a SN more than
the specified amount of times.
It's the task of RRC to reestablish the bearer to recover
from that.
4 years ago
Francisco
fb04fdbe69
sched,bugfix - fix mcs/tbs computation for 256-QAM.
4 years ago
Andre Puschmann
9f627911df
rlc_am_{control}_test: extend status PDU test
4 years ago
Andre Puschmann
6a4bf65fd8
rlc_am_lte: extend check for valid status PDU
...
check for all NACKed SNs are smaller than ACK_SN
4 years ago
Andre Puschmann
fa23be4d6d
rlc_stress_test: fix SDU generation
...
the PDCP SN was used to fill the SDU with data. But since
the max value isn't devisable by 256 it didn't always increment
correctly. Use a normal byte now that wraps correctly.
4 years ago
Andre Puschmann
62b2327178
rlc_am_lte: fix header reconstruction
...
* optimize processing of status PDU (SN is removed from window immediately)
* fix maxRetx signaling for segments
* make tx_window_t a template class, rename and use for rx_window as well
4 years ago
Andre Puschmann
5e345df439
rlc_stress_test: make pedantic operation the default
...
among cleaning up the the stress test, it does:
* remove optional pedantic flag and make it the default to catch resegmentation issues
* use random SDU and MAC grant size the default
* use factory functions to create RLC configs (make maxretx configurable)
* remove reestablishment and print error when maxRetx is triggered
* fix NR PCAP usage and removes PCAP ifdef
4 years ago
Andre Puschmann
1a59f2d5ba
rlc_am_test: extend header reconstruction test
...
this patch adds various new test cases generated from
rlc_stress_test to test header reconstruction.
4 years ago
faluco
dc6a2e5160
Implement a log spy to check if a certain message has been logged.
4 years ago
Francisco
9eeddcec5f
disabled prach vs pucch collision detection temporarily until enb cfg files are updated in the rfci
4 years ago
Francisco
102cb196e1
Addition of enb cfg option to choose between PUCCH with multiplexing and without. Change default prach offset to 4.
...
- The new prach offset of 4 accounts for the size of the PUCCH-ACK region with 100 prbs
4 years ago
Francisco
39aea96c69
fix sched tests UL HARQs. Set common pucch cfg in the sched tests
4 years ago
Andre Puschmann
99df71a914
rf_zmq: make error message when port bind failed more verbose
4 years ago
Francisco
9748c0be84
simplified tbs/mcs computation in scheduler
4 years ago
Pedro Alvarez
9c67a029b5
Trying to make calculation of offset in status report more readable
4 years ago
Pedro Alvarez
7b23a83306
Fixup tests to consider wrap-around in status report
4 years ago
Pedro Alvarez
c4e9516561
Fix issue when LMS wraps-around and FMS doesn't in PDCP status report. Fix incorrect update of FMS on clear_sdu. Adding unit test to LMS wrap-around.
4 years ago
Pedro Alvarez
0d020a7f07
Disable sending and handling status report on SRBs
4 years ago
Xavier Arteaga
81cc8de2b7
Protect NAS unbounded message read
4 years ago
David Rupprecht
aef6805d03
Added pcap net feature to eNodeB
4 years ago
David Rupprecht
9e72d89fa1
MAC PCAP move run thread in base class
4 years ago
David Rupprecht
894e4d3501
Added mac pcap network class that dumps the packets to a udp network sink
4 years ago
David Rupprecht
a58f0642cb
created pcap base class for main handling
4 years ago
David Rupprecht
da658ed2d4
Changable UE ID for enodeb
4 years ago
David Rupprecht
5a95cbdfae
Disable rat assignment in pcap class
4 years ago
David Rupprecht
24bea9a860
Refactored functions for pcap udp packing and write lte mac pcap with udp frame
4 years ago
David Rupprecht
bb3e87302d
Added new pcap enable option and mac nr
4 years ago
AlaiaL
390a4c0f76
[metrics] Changing variable types.
...
Adjusting the license.
4 years ago
AlaiaL
723c56051e
[metrics] Changed the extension from cpp to cc.
...
- Fix date of CMakeLists.txt.
- Initialization of the used variables in the cpu usage calculation.
4 years ago
AlaiaL
41cf15f5ad
[metrics] Adjusted function names and documentation
4 years ago
AlaiaL
d518a625cc
[metrics] Added the system metrics to the csv
4 years ago
AlaiaL
0ba7cd7df9
[metrics] Added system metrics to the enb's stdout
4 years ago
AlaiaL
3a86c210f9
[metrics] Added the system metrics to the ue stdout
4 years ago
Francisco
290f39d805
fix crnti mac ce handling in the rrc in normal (no handover) case
4 years ago
Xavier Arteaga
0346564241
Fix NR-PUCCH encoding
4 years ago
Xavier Arteaga
784f4be72e
SRSUE: NR CSI report fixes
4 years ago
Xavier Arteaga
4679e9f871
Initial NR CSI reports
4 years ago
Francisco
202c4d7566
remove uses of method .clear() of stack timers
4 years ago
Francisco
be7e78da53
limit allocation of PDCP SDU+Timer buffer to DRB+RLC AM cases
4 years ago
Pedro Alvarez
8dfa87c377
Disable undelivered SDU queue for SRBs.
4 years ago
Francisco
a3cb21d378
move undelivered_pdcp_sdus class methods to .cc file. Fixed the forwarding of pending PDCP SNs during S1 handover
4 years ago
Francisco
17fa79f24c
couple the pdcp discard timers and sdu buffer management into same subclass. Add asserts to ensure there is no ambiguity in the PDCP SN assignment and buffering. Extend the discard timer and pdcp sdu buffering to SN lengths above 12
4 years ago
Pedro Alvarez
4723dd0aa9
Converted undelivered sdus queue in the PDCP to use a vector.
...
This was done to improve the performance of the notify_delivery().
4 years ago
Francisco
a72a659fdf
fix timers interface to avoid accessing to a timer after its release. fix gtpu tunnel activation during s1 handover
4 years ago
Xavier Arteaga
6c9ff3ac92
Fix packed sequence genration for non multiples of 8
4 years ago
Xavier Arteaga
1e55d68b55
Removed set RNTI from API
4 years ago
Xavier Arteaga
93752fb2c4
Fix LDPC compilation for GCC 10
4 years ago
Xavier Arteaga
576a923a4c
Added i8 zero vector
4 years ago
Pedro Alvarez
4011714664
Fix errors in RLC stress tests due to PDCP SN wrap-around
4 years ago