Raymond KnoppandRomain Beurdouche 33078f9e3d feat(cnProc_gen_BG1_128): Put back small change in 128-bit SIMD code generator for LDPC decoder (cnProc). ...

Avoid the use of simde_mm_sign_epi8 which is inefficient on aarch64.
Decoding times are significantly improved on aarch64.

Co-authored-by: Romain Beurdouche <romain.beurdouche@eurecom.fr>

2025-05-13 14:08:33 +00:00

..

doc/nrLDPC

Restore LDPC decdoer documentation

2023-01-12 18:02:45 +01:00

nrLDPC_tools

feat(cnProc_gen_BG1_128): Put back small change in 128-bit SIMD code generator for LDPC decoder (cnProc).

2025-05-13 14:08:33 +00:00

nrLDPC_bnProc.h

remove race condition on norm128 global

2023-09-13 17:20:04 +02:00

nrLDPC_cnProc_avx512.h

remove race condition on norm128 global

2023-09-13 17:20:04 +02:00

nrLDPC_cnProc.h

First-level of support for new aarch64 machines

2023-09-13 17:19:20 +02:00

nrLDPC_decoder_CL.c

Fix: modify block_length to Kprime

2025-04-11 14:40:52 +02:00

nrLDPC_decoder_kernels_CL.c

fix a warning. fix openCL regression after rebase

2022-01-04 19:06:38 +01:00

nrLDPC_decoder.c

fix(nrLDPC_decoder): Cleaning LDPC iteration accounting in NR

2025-03-10 15:11:19 +00:00

nrLDPC_init.h

fix ldpc ubsan

2023-04-17 09:30:47 +02:00

nrLDPC_lut.h

CI: fixing compilation warning in phy simulators

2019-10-05 11:24:03 +02:00

nrLDPC_mPass.h

fix ldpc ubsan

2023-04-17 09:30:47 +02:00

nrLDPC_types.h

fix(nrLDPC_coding): remove T2 legacy setCombIn from segment decoder parameters

2025-01-10 09:00:35 +00:00

nrLDPCdecoder_defs.h

Correct max number of LDPC segments for T2 processing

2024-11-19 16:15:25 +01:00