Code cleanup (SRS, config files and more)
- Remove deprecated elements from some configuration files (also closes#997)
- Improve SRS code with a better separation between common code and gNB/UE code
- Other minor compilation improvements
Implementing the option to have more than 1 RSRP in CSI report
In collaboration with NI/Emerson.
The standard allows up to 4 RSRP measurements in CSI report (1 full and the other
differential). In current OAI, we just configure 1 RSRP report for the strongest
SSB. This MR allows to configure at gNB more than 1 RSRP report (depending on
config file, number of SSBs and UE capabilities) and makes changes at the UE to
send this report.
CI: Generalize PhySim tests, run PhySims on Gracehopper
This changeset generalizes/harmonizes the CI infrastructure for physim
testing, and adds a new pipeline for physim tests on Gracehopper
machines.
Specifically:
- Refactor the existing CI code for running and analyzing physical
simulators. The existing code assumed those tests are deployed using
OpenShift. Instead, OpenShift/Helm-specificities (e.g., running a
"release") have been removed, and instead of hardcoding the deployment
of the physical simulators using a script for Helm, the function takes
any script that can deploy using Helm/docker/from source. The overall
appearance is the same.
- Add a new pipeline RAN-PhySim-GraceHopper-5G that executes physical
simulators on gracehopper. Reporting is done in the same way as for
existing physical simulators (e.g., like in the RAN-PhySim-Cluster-5G
pipeline), but deployed in docker instead of helm.
-> some tests don't work reliably on ARM yet and are skipped
- the RAN-gNB-N300-Timing-Phytest-LDPC pipeline has been reworked to use
the same code for testing T2 offload, with ctest, and the same
reporting.
- the CUDA LDPC test XML has been removed, as these tests don't work
reliably.
- add a new macro add_timed_physim_test() that allows to define
thresholds to be checked, so that ctest will check for them. see the
documentation for more information
Prior to this change, each test would have, as label, it's description.
This leads to many labels, which are normally reserved for grouping
tests. Move the description into a custom property instead.
Since we rely on the JSON output for the generation of HTML reports, we
need to also extract the description from the generated JSON. However,
custom properties are only logged starting from cmake 3.30. For
back-wards compatibility, write an additional, separate environment
variable TEST_DESCRIPTION from which the description can be read as a
fallback.
For the add_timed_physim_test(), this has been already done in a parent
commit.
The unit tests have been imported with this command:
xmlstarlet sel -t
-m '//testCaseList/testCase[class="Run_Physim"]' \
-v "concat('add_timed_physim_test(physim.5g-offload.',physim_test,'.testX.Y \"',desc,'\" ',physim_test,' ',physim_run_args,')')" \
-n t2_offload_enc_nr_dlsim.xml t2_offload_dec_nr_ulsim.xml >> ../../openair1/SIMULATION/tests/CMakeLists.txt
The next commit will further cleanup this output. I commit it like this
for verification purposes.
Add two new macros that allow to register physim tests for ctest,
including timing thresholds that should be fulfilled.
First, add_timed_physim_test() registers a new test using a helper
script RunTimedTest.cmake. This is because after test execution, we need
to analyze logs, and the recommended way to do multiple steps in one
test is via a helper cmake script:
https://cmake.org/pipermail/cmake-developers/2016-February/027816.html
Second, check_physim_threshold() adds new thresholds. It takes a text to
be parsed, and a condition (e.g., "< 20") to check for the number
following the threshold (which is assumed to be present right after the
threshold). It uses a test property to count the total number of checks
(limiting them to 10), and sets environment variables for the script. I
initially planned to use a test property for checks, but those are only
valid in the same directory, and RunTimedTest.cmake seems to be assumed
by cmake to be "elsewhere", hence I needed to resort to environment
variables.
RunTimedTest.cmake is called through cmake with the test parameters and
checks. It re-constructs a list of checks [1], runs the test, and pipes
the log into a separate script that is passed all checks (see below).
Afterwards, it verifies that both the test and script passed.
A script analyze-timing.sh builds an awk script from the checks passed.
The script analyzes each line of the test output for the threshold, and
compares against the threshold. analyze-timing.sh returns success if all
checks passed.
[1] I did not manage to pass a "list" of checks in a single environment
variable through the check_physim_threshold(), which would be simpler.
Change signature of the macro to take the full test name, a separate
description (previously called a label?), and the executable and test
invocation. The overall signature becomes easier.
The rational for the full test name is that a later commit will
introduce the possibility to define physim tests that are checked for
specific times, in which case we need to reuse the test name. Thus,
using the full name from the beginning leads to more clarity, even
though they are a bit longer.
Add an additional safety check that the given simulator is actually a
target defined within the project (in other words, it actually exists).
Update the documentation correspondingly.
Most (but not all) physical simulators are defined under
openair1/SIMULATION/. Defining physical simulators below there seems to
be the more "obvious" directory anyway, as the physical simulators are
openair1/SIMULATION/tests
Phy(L1) Sim tests => PhySim (tests)
Reworking BWP configuration
This MR reworks the configuration via fail of additional BWPs. Moreover, it
allows for a single BWP to be configured for a given UE at a given time (BWP
switch possible only via reconfiguration). It also reworks PTRS configuration
for consistency with the BWP one.
gNB: adaptive DL/UL MCS based on reported SSB-SINR and measured PUSCH SNR
Currently this is used esp. if the number of rounds is limited to 1
(e.g. disable_harq).
feat(ldpc_aal): Do not try EAL init if FHI 7.2 is enabled
When FHI 7.2 is enabled, the EAL initialization is done by the FHI and
the attempt to init EAL by ldpc_aal will fail. In that case, it is not
necessary neither to provide a DPDK core list for BBDEV.
This MR offers to skip EAL init in ldpc_aal when FHI 7.2 is enabled in
order to avoid error messages in the log and to allow dropping the
useless DPDK cores argument in that case.
Also add some fixes to the LDPC offload documentation.
NR UE: add support for PDSCH rate matching with NZP CSI-RS
Currently only ZP CSI-RS is considered in function
configure_ratematching_csi, add support for NZP CSI-RS.
Closes#975
If the DCI in PDCCH is not correctly received, no PDSCH would be decoded but counted as no error.
Also if the DCI was not correctly received, a TBS of 0 would be used in some computations and memory allocations.
Fix these not counting if there is no PDSCH decoding, and by using the known TBS instead of the one received from the DCI.
When FHI 7.2 is enabled, the EAL initialization is done by the FHI and
the attempt to init EAL by ldpc_aal will fail.
In that case, it is not necessary neither to provide a DPDK core list
for BBDEV.
This commit offers to skip EAL initialization if it is already initialized
in order to allow dropping the useless DPDK cores argument in that case.
The change relies on `rte_dev_probe` to detect if EAL was already
initialized.
Ldpc encoder zc384 optimization
This is a modification for the ldpc encoder which improves performance for
high-throughput cases. It contains a newly-generated file for innermost part of
the parity bit generation which, for AVX512, uses the mm512_permutex2var_epi8
instruction and for 128-bit (aarch64) the mm_alignr_epi8 (vextq_s8) in the
computation corresponding to each non-zero bit of the H matrix. This allows for
removing a very large memcpy and reduces the memory demand for the parity matrix
computation. For the moment this is limited to largest case in BG1, namely
Zc=384. This is the only format used when the number of segments after codeblock
segmentation is large (>8). This improves the timing required for ldpc parity
bits generation by more than a factor 2 with AVX512. This significantly reduces
the total time to generate the DLSCH/PDSCH.
See the merge request description for numbers.
Speedup complex rotate for aarch64
This is a simple optimization for aarch64 for the "rotate_cpx_vector" used for
the phase_rotation in 5G NR. It provides a speedup of around 1.5 on high-end
aarch64 targets compared to the SIMDe translated version. It is implemented with
native NEON intrinsics and was largely inspired by suggestions from ChatGPT 5 :-)
newly-generated files for innermost part of the parity bit
generation which, for AVX512, uses the mm512_permutex2var_epi8
instruction and for 128-bit (aarch64) the mm_alignr_epi8 (vextq_s8) in
the computation corresponding to each non-zero bit of the H matrix.
It is generated with ldpctest -n0 -g1. the changes for this are in a
future MR. I did these changes to compare with the GPU version
(ldpc_encoder_cuda), so that the CPU version is as fast as possible. The
generator will come when we merge the GPU encoder/decoder which also has
a generator for the CUDA version. it was difficult to extract the
changes in ldpctest/ldpc_encoder.c so I just took the generated file.
This allows for removing a very large memcpy and reduces the memory
demand for the parity matrix computation. For the moment this is limited
to largest case in BG1, namely Zc=384. This is the only format used when
the number of segments after codeblock segmentation is large (>8). This
improves the timing required for ldpc parity bits generation by more
than a factor 2 with AVX512. This significantly reduces the total time
to generate the DLSCH/PDSCH
Jaroslava Fiedlerova