Because it is SPSC, atomic variables are enough to synchronize the two
threads. Concretely put() is modified to ensure that read_idx is
"acquired" so that the read index has been written, including memory, by
the other thread. "Release" ensures that the write operation (including
the memory to the buffer) is written before it is visible to the other
thread in get() (which in turn "acquires" it). For more information, see
also [1].
The rest of the library has been simplified to work only with put() and
get(), reducing total code and the surface for possible bugs.
This (and the previous) version has been tested with the
threadSanitizer:
TSAN_OPTIONS=halt_on_error=1 ./common/utils/ds/tests/test_spsc_q_perf
On my machine, using Google Benchmark, I measure a considerable 5x speed
improvement:
$ /tmp/benchmark/tools/compare.py benchmarks pthread.json atomic.json
Comparing pthread.json to atomic.json
Benchmark Time CPU Time Old Time New CPU Old CPU New
--------------------------------------------------------------------------------------------------------------
BM_spsc_q/10 -0.8201 -0.0740 266779912 47989020 52387 48512
BM_spsc_q/16 -0.8301 -0.0520 249656592 42428540 51462 48784
BM_spsc_q/32 -0.8003 -0.0841 230248798 45972155 53841 49311
BM_spsc_q/64 -0.7995 -0.0506 210429791 42199674 50690 48124
BM_spsc_q/128 -0.7930 -0.1101 205212273 42483155 52745 46936
BM_spsc_q/160 -0.7880 -0.1663 216644738 45938247 53400 44518
OVERALL_GEOMEAN -0.8057 -0.0904 0 0 0 0
[1] https://en.cppreference.com/w/c/atomic/memory_order.html
Assisted-By: Claude:claude-sonnet-4-6
Signed-off-by: Robert Schmidt <robert.schmidt@openairinterface.org>
This data structure is designed to support the use cases of the L1,
i.e., provide a queue which can be quickly used to order L1 jobs in FIFO
order with simple iteration of jobs to treat. It is a Single-Consumer
Single-Producer queue library.
Signed-off-by: Robert Schmidt <robert.schmidt@openairinterface.org>
phy_procedures_nr_gNB.c calls pack_srs() functions to pack individual
SRS "subPDUs". Correspondingly, the library depends on NFAPI_LIB. In
nr_ulsim, in a later commit, we will need to unpack the SRS.indication,
so remove dummy functions and implement correctly.
Signed-off-by: Robert Schmidt <robert.schmidt@openairinterface.org>
When receiving single-entry PHR with HARQ PID -1
(nr_mac_process_mac_pdu() is called with such parameter in at least two
places), the error handling path calls continue skipping the accounting
of the current sub-pdu (pduP and pdu_len). This results in an infinite
loop. Fix by moving functionality into a separate function.
Fixes: 2e0493632d ("RLC data ind: single call for complete TB at gNB")
Signed-off-by: Robert Schmidt <robert.schmidt@openairinterface.org>
The existing ltebox startup sequence on nepes occasionally fails.
Running
sudo su -c 'screen -dm -S simulated_hss /opt/hss_sim0609/starthss' \
&& sleep 1 && sudo /opt/ltebox/tools/start_mme && sudo /opt/ltebox/tools/start_xGw \
& sleep 1 && ip a show dev tun1
Reliably shows that no tun1 interface is created, because that takes
around 1s itself. Increasing the second sleep to 2 reliably deploys it
correctly.
Do the same for sabox and ltebox on nano for consistency.
Signed-off-by: Robert Schmidt <robert.schmidt@openairinterface.org>
Functionality existed to print all output of asn1c on error, but the
form
asn1c || cat X
makes the command exit with a 0 exit code (success). Fix by explicitly
returning non-zero exit code in this case:
asn1c || ( cat X && false )
Signed-off-by: Robert Schmidt <robert.schmidt@openairinterface.org>
Both do practically the same. Harmonize them into one while guaranteeing
backwards compatibility.
Signed-off-by: Robert Schmidt <robert.schmidt@openairinterface.org>
This fixes two problem:
1. an omitted --build-lib library could yield strange grep errors. For
instance, calling build_oai --build-lib --cmake-opt OPTION passed
"--cmake-opt" to grep, that does not know this option. Handle by
explicitly separating options and search pattern through --.
2. Even with this fixed, set -e triggered an exit because grep exits
with non-zero error code. Remove set -e to have the error handling
code pass.
With set -e removed, properly check that the build passes.
Update a log line to make it clearer.
Assisted-by: Claude:claude-sonnet-4-6
Signed-off-by: Robert Schmidt <robert.schmidt@openairinterface.org>
Remove a lot of unused code, defines, and config options
Remove unused header files and defines.
Slightly clean up gnb_config.c by removing all _IDX variables. Remove
these unused parameters:
- MACRLCs.[0].num_cc
- MACRLCs.[0].local_n_portc
- MACRLCs.[0].remove_n_portc
- MACRLCs.[0].remote_s_portc
- MACRLCs.[0].remote_s_portd
- L1s.[0].num_cc
- L1s.[0].local_n_portc
Remove the gNB_MAC_INST.eth_params_s, since it's only purpose is to
store what can be on the stack.
For the Aerial log message, don't print an empty prefix but what is
actually used for Aerial.
The indices are brittle and might be wrong. Instead, use the option name
(via gpd()) to look up config options. This will also simplify the
removal of unused parameters, as we don't have to update all indices.
Remove leading CONFIG_STRING_, as it is just repetitive and adds noise.
The constants are both short(er) and sufficiently self-explanatory
without it.
It can happen that a message is delayed, but then no UE is found for
such message, e.g., something is delayed but the UE context is deleted
in the meantime:
[02:23:16.320730] [NR_RRC] I [--] (UE ID 3 RNTI d2b1) Remove UE context
Assertion (ue_context_p) failed!
In rrc_delay_transaction() /oai-ran/openair2/RRC/NR/rrc_gNB.c:191
In that case, don't delay the message further -- all handlers gated by
this function will check for the UE context and abort the transaction if
not present.
It can happen that no PDU session can be setup (e.g., PDU session
already setup). Check that there is at least one before triggering the
message.
Also initialize bearer_req.gNB_cu_cp_ue_id early to avoid uninitialized
IDs (e.g., when no PDU session to be setup).
Split out to make a standalone NTN configuratio page that would be
easier to find. I only modify the headings in the new file and write an
introductory sentence of the tutorial, the rest is unchanged.
- all RAN code, CI code, configuration files, dockerfiles, in CSSL v1.0
- all deployment code (openshift, charts, ancillary files like shell
scripts), in MIT
- documentation in CC-BY-4.0
- exceptions might apply and are listed in NOTICE
- there is a new LICENSES folder with all licenses
- CONTRIBUTIONS.md has been updated accordingly
For automated changes based on OAI PL v1.1:
perl -i~ -0pe 's/\/\*.*Licensed to the OpenAirInterface.*openairinterface.org\n#?/\/*\n * SPDX-License-Identifier: LicenseRef-CSSL-1.0\n/s' **/*.{c,h,cpp}
perl -i~ -0pe 's/\/\*.*Licensed to the OpenAirInterface.*openairinterface.org\n#?/\/*\n * SPDX-License-Identifier: LicenseRef-CSSL-1.0\n/s' **/*.ts
perl -i~ -0pe 's/<!--.*Licensed to the OpenAirInterface.*openairinterface.org\n.*-->/<!-- SPDX-License-Identifier: LicenseRef-CSSL-1.0 -->/s' **/*.xml
The rest (cmake, files with missing license, cmake) manually.
Remove the \file directive, as it is always superfluous because the
current file is implicit [1]:
> If the file name is omitted (i.e. the line after \file is left blank)
> then the documentation block that contains the \file command will belong
> to the file it is located in.
Author names and e-mails are not relevant for us: it can always be
inferred from git blame, and is often outdated.
Eurecom code has been contributed and was under OAI PL v1.0/v1.1.
For the cpack package contact: put generic email address that is
independent of an individual and that will remain reachable.
[1] https://www.doxygen.nl/manual/commands.html#cmdfile
Do not limit DL power, because it is not done in the other pipelines
either. Set the UL target power high, as this increases the achievable
throughput high (PUCCH can be normal). Finally, increase BLER thresholds
to improve the MCS a bit.
Introduce new data structures to track the average SNR and RSSI, and
that allow to dynamically modify the target SNR (per UE) as well as
continuously get TPC updates based on the average SNR. The target SNR is
per power control loop (which a future commit will extend to use for
PUCCH).
Concretely, it maintains an average of measured SNR and RSSI. It also
updates "tpc_in_flight", which tracks TPC changes that don't show up in
the average yet. For instance, imagine that the target SNR of 15 changes
to 20. Three successive TPC commands need to be sent (+3, +1, +1), but
it will take time to show up in the average SNR. To account for this,
tpc_in_flight is updated by the TPCs sent, and an average will make it
go down back to zero at the same pace as the average SNR approaches the
target SNR. The sum of average SNR and tpc_in_flight sums up to the
actual, current SNR, which approximates the target SNR.
The SNR is kept within -1<=targetSNR<=+2dB to avoid too many TPC
changes.
The current average SNR (sum of average SNR and tpc_in_flight) is used
to continuously get TPC updates, taking into account recent TPCs sent,
the difference to target SNR, and RSSI.
If there is DTX, tpc_in_flight is modified to artificially introduce TPC
commands which increase the currently used SNR of the UE. This is
temporary, as tpc_in_flight eventually will go back to zero. If the DTX
was due to sudden SNR drop (e.g., higher path loss), this will allow to
quickly adapt the UE's SNR to new conditions.
The periodical logs are extended to not only show the current SNR, but
also the difference from the target for better analysis.
Co-authored-by: Maxime Elkael <m.elkael@northeastern.edu>
The phr_txpower_calc field actually needs to be saved in the HARQ
process for later retrieval, so set sched_pusch after phr_txpower_calc
has been calculated.
xran needs to be informed about the cores to use for packet processing.
Prior to this change, we only filled the first 64bits, but it can go up
to 128 in xran. Change the helper accordingly to support all.
The README says:
> It is just generating the XML file and it is placing it in
> $(OPENAIR_TARGETS)/SIMU/EXAMPLE/OSD/WEBXML/ or http://localhost/xmlfile/
> so that the oaisim could pars and run the simulation/emulation
oaisim does not exist anymore, and corresponding simulation cannot be
done. Also, it requires XAMPP and similar technologies, that I think
nobody uses in the context of OAI.
Notably, remove asn_codecs_prim_xer.h from the list of files, triggering
problems when re-running ninja, because the "canonical OAI asn1c"
(https://github.com/mouse07410/asn1c, commit ID
940dd5fa9f3917913fd487b13dfddfacd0ded06e) does not generate it:
ninja explain: output openair2/XNAP/MESSAGES/asn_codecs_prim_xer.h doesn't exist
In this version, a couple of files are not generated, so remove them
from the list.
Also, leave the files in the order as shown by ls.
Integration `2026.w11`
* !3864 remove xran_fh_tx_send_slot_BySymbol() and xran_fh_rx_read_slot_BySymbol()
* !3967 Harmonize the use of definition for number of symbols per slot
* !3971 Limit dl_DataToUL_ACK values to 15 as per standard
* !3414 Added NTN-FR2 FDD bands defined in release 18 in Ka-band range
* !3919 Reduce number of mutex locking in MAC->RLC direction
* !3923 XNAP: Add encode/decode and unit tests for Xn Setup Request/Response/Failure
* !3965 Enable BladeRF support in gNB Docker image
* !3904 add multi-ue support to vrtsim
* !3980 SDR reordering: correct warning, and USRP: return samples written
* !3834 Use flat buffer for txdataF in gNB L1
* !3981 Bugfix: make longer telnet cmdfunc names
* Reduce timing reference value for feptx_total in 40 MHz phytest
See merge request oai/openairinterface5g!3976
The maximum number of cmdfunc names is 20, which IMO is too low. Worse,
if it's longer, nothing will warn us because C will just fill the array
to it's end (without \0 at the end). To remedy this:
- Allow longer names till 64 bytes (I hit the previous limit of 20)
- Check that the name is within the length limit: if somebody wants a
very long name, the strnlen() will return the maximum length, hitting
that assertion.
- When reading a command during runtime, allow the corresponding maximum
length.
On that occasion, also increase total number of permitted cmdfuncs.
The original error was simply "failed to write to RF", which is not very
precise. Update with sample numbers for context, and add it in another
similar case for consistency.
Co-authored-by: Sagar Arora <sagar.arora@openairinterface.org>
At least writerProcessWaitingQueue() checks for the exact number of
samples written. In many other places, we check that the return is
non-zero.
The usrp-tx-write-thread functionality returned 0, which triggers
various problems. Return the number of samples, as in the "direct write"
case (if-block before else-block).
Co-authored-by: Sagar Arora <sagar.arora@openairinterface.org>
Adds the API, which will always prepend a 3 byte header before writing a
PDU. This can be used to call less often into RLC, and avoid potential
locking.
The API is used at the gNB DLSCH scheduler. Also, this removes the
"overhead" calculation: since we fill the entire TB at once, RLC will
leave space for headers as necessary.
This extends the functionality provided by record in that it allows
collection of data into a Clickhouse database, allowing aggregation of
data from gNB and UE for functions like channel estimation for PUSCH
transmissions that did not decode successfully at the gNB.
It is intended to be used in conjunction with Aerial Data Lake but is
not limited to it. Data access is simpler than in !3462, but could be
improved using the metadata approach from !3462.
The integration uses logfmt for fast logging, and a clickhouse DB
client, both integrated using CPM.
This change set also adds a docker-compose file for a UE on 100MHz, to
be used with Aerial configurations, and builds the tracer for Aerial and
nrUE docker files.
Ubuntu 20 is EOL [1] unless people pay for support. Therefore, remove it
from build_helper. People can still use it, but might need to install a
more recent cmake version (which the next commit will upgrade). The rest
should be the same.
[1] https://ubuntu.com/about/release-cycle
Robert Schmidt