Merge remote-tracking branch 'origin/tx_symbol_processing' into integration_2026_w15 (!3489)

Thread-pool support for TX symbol processing. This MR adds thread-pool support for TX symbol processing. It allows modulation /mapping/layer-precoding to run in parallel and offers a speedup of around 3 compared to single-thread execution. This is particularly important for large bandwidths and 4 or more TX antenna ports where the precoding operation is quite computationally-intensive. Please see the description of !3489 for the performance comparison results and the nr_dlsim timing measurements.
2026-04-09 18:11:03 +02:00
parent 83ffa53ccc 03350ab21b
commit 8583689349
8 changed files with 300 additions and 137 deletions
@@ -193,6 +193,7 @@ sudo ethtool -G enp1s0f0 tx 4096 rx 4096
 - Enable Performance Mode `sudo cpupower idle-set -D 0`
 - If you get real-time problems on heavy UL traffic, reduce the maximum UL MCS using an additional command-line switch: `--MACRLCs.[0].ul_max_mcs 14`.
 - You can also reduce the number of LDPC decoder iterations, which will make the LDPC decoder take less time: `--L1s.[0].max_ldpc_iterations 4`.
+- Read more on system tuning in the [dedicated document on performance tuning](./tuning_and_security.md#performance-tuning).

 ### 6.3 Uplink issues related with noise on the DC carriers

@@ -69,6 +69,109 @@ sudo sysctl -n -e -q -w net.core.wmem_default=134217728
 sudo sysctl -n -e -q -w net.core.wmem_max=134217728
 ```

+### System tuning
+In order to get an optimal real-time behavior, a few tunings can be performed on the host system:
+- The use of isolated cores for the softmodem prevents competitions on the usage of core between the softmodem and other processes.
+  Core isolation is enabled through the kernel command line.  
+  **Warning: modifying the kernel command line can harm the OS behavior. Proceed with caution.**  
+  Refer to the [OAI 7.2 Fronthaul Interface Tutorial](./ORAN_FHI7.2_Tutorial.md) for examples.
+
+### Softmodem tuning
+The way the NR softmodem uses the computing ressource can be configured.
+It can have a significant effect on the performance and real-time behavior:
+- The L1 TX and L1 RX threads are the two main threads
+  executing the L1 RX and L1 TX pipelines.  
+  These threads are ideally assigned to two dedicated cores.
+  To be dedicated, the cores should be isolated in the
+  kernel parameters and not be assigned elsewhere.  
+  They can be assigned to specified cores with options
+  `--L1s.[0].L1_tx_thread_core` and `--L1s.[0].L1_rx_thread_core`
+  followed by a core id.
+- The thread pool is a group of processor cores over which
+  some baseband processing worker cores execute.  
+  It is configured by providing a list of core ids
+  after option `--thread-pool`.  
+  `-1` can also be passed instead of a core id
+  in order to use a floating core.  
+  By default, the thread pool is 8 floating cores.
+- PDSCH generation (i.e., layer mapping and precoding) is by default executed
+  in the L1 TX thread but can be multithreaded using the thread pool.  
+  This is enabled by option `--L1s.[0].L1_num_tx_sym_per_thread` followed
+  by the number of symbols that should be processed in each thread.  
+  This option can also be set in the gNB configuration file in field
+  `L1_num_tx_sym_per_thread` in the `L1s` section.
+
+### Workarounds
+If the real-time performance remains bad after tuning the system and softmodem,
+some workarounds allow to lower the computing demand
+at the cost of lower network performance:
+- If you get real-time problems on heavy UL traffic,
+  reduce the maximum UL MCS using an additional
+  command-line switch: `--MACRLCs.[0].ul_max_mcs 14`.  
+  This comes at the cost of a lower spectral efficiency
+  (i.e., less data for the same radio resource).
+- You can also reduce the number of LDPC decoder iterations,
+  which will make the LDPC decoder take less time:
+  `--L1s.[0].max_ldpc_iterations 4`.  
+  The default number of LDPC iterations is 8.
+  Lowering the number of iteration comes at the cost
+  of more unsuccessful transmissions.  
+  OAI offers multiple implementation of LDPC coding,
+  including offloading to an accelerator,
+  the number of LDPC iteration should be chosen accordingly.
+
+### Known hardware behaviors
+Here is a **non-exhaustive** list of known behaviors related to hardware architecture:
+- On some AMD EPYC series processors with Zen architecture
+  (at least every Zen4, Zen4c, Zen5 and Zen5c based processors
+  experience this behavior),  
+  the processor is made of multiple dies holding one or multiple
+  core complexes which are groups of cores with an L3 cache.  
+  This means that cores from different core complexes
+  do not share the same L3 cache and communication between
+  these cores implies inter L3 cache communication  
+  within a die or, even worse, between dies,
+  which has a cost in term of latency.  
+  Depending on the system configuration,
+  the NUMA topology may reflect this physical topology,
+  which can induce even further latency for
+  inter core complex communication.  
+  The softmodem is sensitive to this latency and its performance
+  can be harmed if it uses cores across the border of dies or core complexes,  
+  especially when multithreading of PDSCH generation is enabled
+  (argument of `--L1s.[0].L1_num_tx_sym_per_thread` is superior to 0).  
+  **Solution**: We recommend to use only one core complex
+  or one die for allocating cores to nr-softmodem process.
+- NUMA architecture: Make sure you don't assign cores to nr-softmodem from
+  different numa nodes. Cores from different numa cores can induce latency.
+  It is preferred to use the cores from numa node
+  which is used by the Fronthaul NIC.
+
+**Example**: How to assign cores to `nr-softmodem` process for gNB/DU connected to an o-ran 7.2 O-RU on EPYC 9575F (64 Zen5)
+assuming that we have configured `L1s.[0].L1_num_tx_sym_per_thread = 1`
+
+There are 8 cores per core complex and one core complex per die. You can visulize this topology using 
+
+```bash
+# Ubuntu 25.04
+for X in $(seq 0 63); do echo -n "cpu$X -> die "; cat /sys/devices/system/cpu/cpu$X/topology/die_id; done
+```
+   
+A 100MHz 4x4 FR1 gNB/DU with FHI 7.2 can be executed with full capabilities (4DL and 2UL layers)
+on a single core complex (cpus 0-7 in this example) by folowing the instructions of the
+[OAI 7.2 Fronthaul Interface Tutorial](./ORAN_FHI7.2_Tutorial.md)
+with the following core assignment:
+
+- `L1s.[0].L1_tx_thread_core = 0`
+- `L1s.[0].L1_rx_thread_core = 1`
+- `RUs.[0].ru_thread_core = 2`
+- `fhi_72.system_core = 3`
+- `fhi_72.io_core = 4`
+- `fhi_72.worker_cores.[0] = 5` (minimum 1 core)
+- `thread-pool: 2,3,6,7`
+  (thread pool can overlap with `RUs.[0].ru_thread_core` and `fhi_72.system_core`)
+
+
 ## Capabilities

 Historically, all softmodems are executed as `root`, typically using `sudo`.
@@ -208,8 +208,7 @@ static size_t dump_L1_meas_stats(PHY_VARS_gNB *gNB, RU_t *ru, char *output, size
  output += print_meas_log(&gNB->dlsch_encoding_stats, "DLSCH encoding", NULL, NULL, output, end - output);
  output += print_meas_log(&gNB->dlsch_scrambling_stats, "DLSCH scrambling", NULL, NULL, output, end-output);
  output += print_meas_log(&gNB->dlsch_modulation_stats, "DLSCH modulation", NULL, NULL, output, end - output);
-  output += print_meas_log(&gNB->dlsch_resource_mapping_stats, "DLSCH resource mapping", NULL, NULL, output,end-output);
-  output += print_meas_log(&gNB->dlsch_precoding_stats, "DLSCH precoding", NULL, NULL, output,end-output);
+  output += print_meas_log(&gNB->dlsch_pdsch_generation_stats, "PDSCH generation", NULL, NULL, output, end - output);
  output += print_meas_log(&gNB->phy_proc_rx, "L1 Rx processing", NULL, NULL, output, end - output);
  output += print_meas_log(&gNB->ts_deinterleave, "UL segment deinterleaving", NULL, NULL, output, end - output);
  output += print_meas_log(&gNB->ts_rate_unmatch, "UL segment rate recovery", NULL, NULL, output, end - output);
@@ -280,8 +279,7 @@ void *nrL1_stats_thread(void *param) {
  reset_meas(&gNB->rx_pusch_stats);
  reset_meas(&gNB->dlsch_scrambling_stats);
  reset_meas(&gNB->dlsch_modulation_stats);
-  reset_meas(&gNB->dlsch_resource_mapping_stats);
-  reset_meas(&gNB->dlsch_precoding_stats);
+  reset_meas(&gNB->dlsch_pdsch_generation_stats);
  while (!oai_exit) {
    sleep(1);
    if (ftruncate(fileno(fd), 0) != 0 || fseek(fd, 0, SEEK_SET) != 0) {
@@ -309,8 +307,6 @@ void init_gNB_Tpool(int inst)
  PHY_VARS_gNB *gNB;
  gNB = RC.gNB[inst];
  gNB_L1_proc_t *proc = &gNB->proc;
-  // PUSCH symbols per thread need to be calculated by how many threads we have
-  gNB->num_pusch_symbols_per_thread = 1;
  // ULSCH decoding threadpool
  initTpool(get_softmodem_params()->threadPoolConfig, &gNB->threadPool, cpumeas(CPUMEAS_GETSTATE));

@@ -17,8 +17,6 @@
 #include "executables/softmodem-common.h"
 #include "SCHED_NR/sched_nr.h"

-// #define DEBUG_DLSCH
-// #define DEBUG_DLSCH_MAPPING
 #include <simde/x86/avx512.h>
 #define USE128BIT

@@ -46,21 +44,9 @@ static int do_ptrs_symbol(const nfapi_nr_dl_tti_pdsch_pdu_rel15_t *rel15,
      /* check if cuurent RE is PTRS RE*/
      uint16_t beta_ptrs = 1;
      txF[k] = c16mulRealShift(mod_ptrs[ptrs_idx], beta_ptrs * amp, 15);
-#ifdef DEBUG_DLSCH_MAPPING
-      printf("ptrs_idx %d\t \t k %d \t \t txdataF: %d %d, mod_ptrs: %d %d\n",
-             ptrs_idx,
-             k,
-             txF[k].r,
-             txF[k].i,
-             mod_ptrs[ptrs_idx].r,
-             mod_ptrs[ptrs_idx].i);
-#endif
      ptrs_idx++;
    } else {
      txF[k] = c16mulRealShift(*in++, amp, 15);
-#ifdef DEBUG_DLSCH_MAPPING
-      printf("k %d \t txdataF: %d %d\n", k, txF[k].r, txF[k].i);
-#endif
    }
    k++;
  }
@@ -74,9 +60,6 @@ typedef union {

 static inline int interleave_with_0_signal_first(c16_t *output, c16_t *mod_dmrs, const int16_t amp_dmrs, int sz)
 {
-#ifdef DEBUG_DLSCH_MAPPING
-  printf("doing DMRS pattern for port 0 : d0 0 d1 0 ... dNm2 0 dNm1 0 (ul %d, rr %d)\n", upper_limit, remaining_re);
-#endif
  // add filler to process all as SIMD
  c16_t *out = output;
  int i = 0;
@@ -126,9 +109,6 @@ static inline int interleave_with_0_signal_first(c16_t *output, c16_t *mod_dmrs,

 static inline int interleave_with_0_start_with_0(c16_t *output, c16_t *mod_dmrs, const int16_t amp_dmrs, int sz)
 {
-#ifdef DEBUG_DLSCH_MAPPING
-  printf("doing DMRS pattern for port 2 : 0 d0 0 d1 ... 0 dNm2 0 dNm1\n");
-#endif
  c16_t *out = output;
  int i = 0;
  int end = sz / 2;
@@ -177,9 +157,6 @@ static inline int interleave_with_0_start_with_0(c16_t *output, c16_t *mod_dmrs,

 static inline int interleave_signals(c16_t *output, c16_t *signal1, const int amp, c16_t *signal2, const int amp2, int sz)
 {
-#ifdef DEBUG_DLSCH_MAPPING
-  printf("doing DMRS pattern for port 0 : d0 X0 d1 X1 ... dNm2 XNm2 dNm1 XNm1\n");
-#endif
    // add filler to process all as SIMD
  c16_t *out = output;
  int i = 0;
@@ -311,22 +288,22 @@ static inline void neg_dmrs(c16_t *in, c16_t *out, int sz)
    *out++ = i % 2 ? (c16_t){-in[i].r, -in[i].i} : in[i];
 }

-static inline int do_onelayer(NR_DL_FRAME_PARMS *frame_parms,
-                              int slot,
-                              const nfapi_nr_dl_tti_pdsch_pdu_rel15_t *rel15,
-                              int layer,
-                              c16_t *output,
-                              c16_t *txl_start,
-                              int start_sc,
-                              int symbol_sz,
-                              int l_symbol,
-                              uint16_t dlPtrsSymPos,
-                              int n_ptrs,
-                              int amp,
-                              int16_t amp_dmrs,
-                              int l_prime,
-                              nfapi_nr_dmrs_type_e dmrs_Type,
-                              c16_t *dmrs_start)
+static inline void do_onelayer(NR_DL_FRAME_PARMS *frame_parms,
+                               int slot,
+                               const nfapi_nr_dl_tti_pdsch_pdu_rel15_t *rel15,
+                               int layer,
+                               c16_t *output,
+                               c16_t *txl_start,
+                               int start_sc,
+                               int symbol_sz,
+                               int l_symbol,
+                               uint16_t dlPtrsSymPos,
+                               int n_ptrs,
+                               int amp,
+                               int16_t amp_dmrs,
+                               int l_prime,
+                               nfapi_nr_dmrs_type_e dmrs_Type,
+                               c16_t *dmrs_start)
 {
  c16_t *txl = txl_start;
  const uint sz = rel15->rbSize * NR_NB_SC_PER_RB;
@@ -408,7 +385,7 @@ static inline int do_onelayer(NR_DL_FRAME_PARMS *frame_parms,
  } else { // no PTRS or DMRS in this symbol
    txl += no_ptrs_dmrs_case(output + start_sc, txl, amp, sz);
  } // no DMRS/PTRS in symbol
-  return txl - txl_start;
+  return;
 }

 static inline void do_txdataF(c16_t **txdataF,
@@ -472,28 +449,123 @@ static inline void do_txdataF(c16_t **txdataF,
  } // RB loop: while(rb < rel15->rbSize)
 }

+typedef struct pdschSymbolProc_s {
+  PHY_VARS_gNB *gNB;
+  NR_DL_FRAME_PARMS *frame_parms;
+  const nfapi_nr_dl_tti_pdsch_pdu_rel15_t *rel15;
+  unsigned int slot;
+  unsigned int startSymbol;
+  unsigned int numSymbols;
+  task_ans_t *ans;
+  unsigned int layerSz2;
+  unsigned int dlPtrsSymPos;
+  unsigned int n_ptrs;
+  unsigned int beam_nb;
+  unsigned int re_beginning_of_symbol[14];
+  c16_t *tx_layers[4];
+  time_stats_t dlsch_resource_mapping_stats;
+  time_stats_t dlsch_precoding_stats;
+} pdschSymbolProc_t;
+
+static void nr_pdsch_symbol_processing(void *arg)
+{
+  pdschSymbolProc_t *rdata = (pdschSymbolProc_t *)arg;
+
+  PHY_VARS_gNB *gNB = rdata->gNB;
+  NR_DL_FRAME_PARMS *frame_parms = rdata->frame_parms;
+  const nfapi_nr_dl_tti_pdsch_pdu_rel15_t *rel15 = rdata->rel15;
+  int slot = rdata->slot;
+  c16_t *tx_layers[rel15->nrOfLayers];
+  for (int l = 0; l < rel15->nrOfLayers; l++)
+    tx_layers[l] = rdata->tx_layers[l];
+  const int nb_re_dmrs = rel15->numDmrsCdmGrpsNoData * (rel15->dmrsConfigType == NFAPI_NR_DMRS_TYPE1 ? 6 : 4);
+  const int n_dmrs = (rel15->BWPStart + rel15->rbStart + rel15->rbSize) * nb_re_dmrs;
+  // Loop Over OFDM symbols:
+  c16_t mod_dmrs[(n_dmrs + 63) & ~63] __attribute__((aligned(64)));
+  const int symbol_sz = frame_parms->ofdm_symbol_size;
+
+  c16_t **txdataF = gNB->common_vars.txdataF[rdata->beam_nb];
+  uint16_t start_sc = (rel15->rbStart + rel15->BWPStart) * NR_NB_SC_PER_RB;
+
+  for (int l_symbol = rdata->startSymbol; l_symbol < rdata->startSymbol + rdata->numSymbols; l_symbol++) {
+    start_meas(&rdata->dlsch_resource_mapping_stats);
+    int l_prime = 0; // single symbol layer 0
+    int l_overline = get_l0(rel15->dlDmrsSymbPos);
+
+    /// DMRS QPSK modulation
+    if ((rel15->dlDmrsSymbPos & (1 << l_symbol))) { // DMRS time occasion
+      // The reference point for is subcarrier -1 of the lowest-numbered resource block in CORESET 0 if the corresponding
+      // PDCCH is associated with CORESET -1 and Type0-PDCCH common search space and is addressed to SI-RNTI
+      // 2GPP TS 38.211 V15.8.0 Section 7.4.1.1.2 Mapping to physical resources
+      if (l_symbol == (l_overline + 1)) // take into account the double DMRS symbols
+        l_prime = 1;
+      else if (l_symbol > (l_overline + 1)) { // new DMRS pair
+        l_overline = l_symbol;
+        l_prime = 0;
+      }
+      const uint32_t *gold = nr_gold_pdsch(frame_parms->N_RB_DL,
+                                           frame_parms->symbols_per_slot,
+                                           rel15->dlDmrsScramblingId,
+                                           rel15->SCID,
+                                           slot,
+                                           l_symbol);
+      // Qm = 1 as DMRS is QPSK modulated
+      nr_modulation(gold, n_dmrs * DMRS_MOD_ORDER, DMRS_MOD_ORDER, (int16_t *)mod_dmrs);
+
+    }
+    uint32_t dmrs_idx = rel15->rbStart;
+    if (rel15->refPoint == 0)
+      dmrs_idx += rel15->BWPStart;
+    dmrs_idx *= rel15->dmrsConfigType == NFAPI_NR_DMRS_TYPE1 ? 6 : 4;
+    c16_t txdataF_precoding[rel15->nrOfLayers][symbol_sz] __attribute__((aligned(64)));
+    for (int layer = 0; layer < rel15->nrOfLayers; layer++) {
+      do_onelayer(frame_parms,
+                  slot,
+                  rel15,
+                  layer,
+                  txdataF_precoding[layer],
+                  tx_layers[layer] + rdata->re_beginning_of_symbol[l_symbol],
+                  start_sc,
+                  symbol_sz,
+                  l_symbol,
+                  rdata->dlPtrsSymPos,
+                  rdata->n_ptrs,
+                  gNB->TX_AMP,
+                  min((double)gNB->TX_AMP * sqrt(rel15->numDmrsCdmGrpsNoData), INT16_MAX),
+                  l_prime,
+                  rel15->dmrsConfigType,
+                  mod_dmrs + dmrs_idx);
+    } // layer loop
+    stop_meas(&rdata->dlsch_resource_mapping_stats);
+
+    start_meas(&rdata->dlsch_precoding_stats);
+    for (int ant = 0; ant < frame_parms->nb_antennas_tx; ant++) {
+      const size_t txdataF_offset_per_symbol = l_symbol * symbol_sz;
+      do_txdataF(txdataF, symbol_sz, txdataF_precoding, gNB, rel15, ant, start_sc, txdataF_offset_per_symbol);
+    }
+    stop_meas(&rdata->dlsch_precoding_stats);
+  }
+  // Task running in // completed
+  completed_task_ans(rdata->ans);
+}
+
 static int do_one_dlsch(unsigned char *input_ptr, PHY_VARS_gNB *gNB, NR_gNB_DLSCH_t *dlsch, int slot)
 {
-  const int16_t amp = gNB->TX_AMP;
  NR_DL_FRAME_PARMS *frame_parms = &gNB->frame_parms;

  time_stats_t *dlsch_scrambling_stats = &gNB->dlsch_scrambling_stats;
  time_stats_t *dlsch_modulation_stats = &gNB->dlsch_modulation_stats;
  const nfapi_nr_dl_tti_pdsch_pdu_rel15_t *rel15 = &dlsch->pdsch_pdu->pdsch_pdu_rel15;
  const int layerSz = frame_parms->N_RB_DL * frame_parms->symbols_per_slot * NR_NB_SC_PER_RB;
-  const int symbol_sz=frame_parms->ofdm_symbol_size;
-  const int dmrs_Type = rel15->dmrsConfigType;
  const int nb_re_dmrs = rel15->numDmrsCdmGrpsNoData * (rel15->dmrsConfigType == NFAPI_NR_DMRS_TYPE1 ? 6 : 4);
-  const int16_t amp_dmrs = min((double)amp * sqrt(rel15->numDmrsCdmGrpsNoData), INT16_MAX); // 3GPP TS 38.214 Section 4.1: Table 4.1-1
  LOG_D(PHY,
        "pdsch: BWPStart %d, BWPSize %d, rbStart %d, rbsize %d\n",
        rel15->BWPStart,
        rel15->BWPSize,
        rel15->rbStart,
        rel15->rbSize);
-  const int n_dmrs = (rel15->BWPStart + rel15->rbStart + rel15->rbSize) * nb_re_dmrs;
+  const int n_dmrs = rel15->rbSize * nb_re_dmrs;

-  const int dmrs_symbol_map = rel15->dlDmrsSymbPos; // single DMRS: 010000100 Double DMRS 110001100
  const int xOverhead = 0;
  const int nb_re =
      (12 * rel15->NrOfSymbols - nb_re_dmrs * get_num_dmrs(rel15->dlDmrsSymbPos) - xOverhead) * rel15->rbSize * rel15->nrOfLayers;
@@ -567,21 +639,9 @@ static int do_one_dlsch(unsigned char *input_ptr, PHY_VARS_gNB *gNB, NR_gNB_DLSC
  start_meas(&gNB->dlsch_pdsch_generation_stats);
  /// Resource mapping
  // Non interleaved VRB to PRB mapping
-  uint16_t start_sc = (rel15->rbStart + rel15->BWPStart) * NR_NB_SC_PER_RB;
-
-#ifdef DEBUG_DLSCH_MAPPING
-  printf("PDSCH resource mapping started (start SC %d\tstart symbol %d\tN_PRB %d\tnb_re %d,nb_layers %d)\n",
-         start_sc,
-         rel15->StartSymbolIndex,
-         rel15->rbSize,
-         nb_re,
-         rel15->nrOfLayers);
-#endif

  AssertFatal(n_dmrs, "n_dmrs can't be 0\n");
  // make a large enough tail to process all re with SIMD regardless a garbadge filler
-  c16_t mod_dmrs[(n_dmrs+63)&~63] __attribute__((aligned(64)));
-  unsigned int re_beginning_of_symbol = 0;

  start_meas(&gNB->dlsch_layer_mapping_stats);
  int layerSz2 = (layerSz + 63) & ~63;
@@ -605,84 +665,69 @@ static int do_one_dlsch(unsigned char *input_ptr, PHY_VARS_gNB *gNB, NR_gNB_DLSC
                                      slot,
                                      frame_parms->symbols_per_slot,
                                      bitmap);
-
-  c16_t **txdataF = gNB->common_vars.txdataF[beam_nb];
  stop_meas(&gNB->dlsch_layer_mapping_stats);
-  // Loop Over OFDM symbols:
-  for (int l_symbol = rel15->StartSymbolIndex; l_symbol < rel15->StartSymbolIndex + rel15->NrOfSymbols; l_symbol++) {
-    start_meas(&gNB->dlsch_resource_mapping_stats);
-    int l_prime = 0; // single symbol layer 0
-    int l_overline = get_l0(rel15->dlDmrsSymbPos);

-#ifdef DEBUG_DLSCH_MAPPING
-    printf("PDSCH resource mapping symbol %d\n", l_symbol);
-#endif
-    /// DMRS QPSK modulation
-    if ((dmrs_symbol_map & (1 << l_symbol))) { // DMRS time occasion
-      // The reference point for is subcarrier -1 of the lowest-numbered resource block in CORESET 0 if the corresponding
-      // PDCCH is associated with CORESET -1 and Type0-PDCCH common search space and is addressed to SI-RNTI
-      // 2GPP TS 38.211 V15.8.0 Section 7.4.1.1.2 Mapping to physical resources
-      if (l_symbol == (l_overline + 1)) // take into account the double DMRS symbols
-        l_prime = 1;
-      else if (l_symbol > (l_overline + 1)) { // new DMRS pair
-        l_overline = l_symbol;
-        l_prime = 0;
+  // spawn symbol threads
+
+  int nb_tasks = 1;
+  int num_pdsch_symbols_per_task = rel15->NrOfSymbols;
+  if (gNB->num_pdsch_symbols_per_thread > 0) {
+    // symbol processing in thread pool enabled
+    num_pdsch_symbols_per_task = gNB->num_pdsch_symbols_per_thread;
+    nb_tasks = rel15->NrOfSymbols / num_pdsch_symbols_per_task;
+    if ((rel15->NrOfSymbols % num_pdsch_symbols_per_task) > 0)
+      nb_tasks++;
+  }
+  pdschSymbolProc_t arr[nb_tasks];
+  task_ans_t ans;
+  init_task_ans(&ans, nb_tasks);
+  int sz_arr = 0;
+  unsigned int re_beginning_of_symbol = 0;
+  int res = 0;
+  for (int l_symbol = rel15->StartSymbolIndex; l_symbol < rel15->StartSymbolIndex + rel15->NrOfSymbols;
+       l_symbol += num_pdsch_symbols_per_task) {
+    pdschSymbolProc_t *rdata = &arr[sz_arr];
+    rdata->ans = &ans;
+    ++sz_arr;
+
+    rdata->gNB = gNB;
+    rdata->frame_parms = frame_parms;
+    rdata->rel15 = rel15;
+    rdata->slot = slot;
+    rdata->startSymbol = l_symbol;
+    res = rel15->NrOfSymbols - (l_symbol - rel15->StartSymbolIndex);
+    if (res >= num_pdsch_symbols_per_task)
+      rdata->numSymbols = num_pdsch_symbols_per_task;
+    else
+      rdata->numSymbols = res;
+    rdata->layerSz2 = layerSz2;
+    rdata->dlPtrsSymPos = dlPtrsSymPos;
+    rdata->n_ptrs = n_ptrs;
+    rdata->beam_nb = beam_nb;
+    for (int s = l_symbol; s < l_symbol + rdata->numSymbols; s++) {
+      rdata->re_beginning_of_symbol[s] = re_beginning_of_symbol;
+      re_beginning_of_symbol += rel15->rbSize * NR_NB_SC_PER_RB;
+      if (n_ptrs > 0 && is_ptrs_symbol(s, dlPtrsSymPos)) {
+        re_beginning_of_symbol -= n_ptrs;
+      } else if (rel15->dlDmrsSymbPos & (1 << s)) {
+        re_beginning_of_symbol -= n_dmrs;
      }
-#ifdef DEBUG_DLSCH_MAPPING
-      printf("dlDmrsScramblingId %d, SCID %d slot %d l_symbol %d\n", rel15->dlDmrsScramblingId, rel15->SCID, slot, l_symbol);
-#endif
-      const uint32_t *gold = nr_gold_pdsch(frame_parms->N_RB_DL,
-                                           frame_parms->symbols_per_slot,
-                                           rel15->dlDmrsScramblingId,
-                                           rel15->SCID,
-                                           slot,
-                                           l_symbol);
-      // Qm = 1 as DMRS is QPSK modulated
-      nr_modulation(gold, n_dmrs * DMRS_MOD_ORDER, DMRS_MOD_ORDER, (int16_t *)mod_dmrs);
-
-#ifdef DEBUG_DLSCH_MAPPING
-      printf("DMRS modulation (symbol %d, %d symbols, type %d):\n", l_symbol, n_dmrs, dmrs_Type);
-      for (int i = 0; i < n_dmrs / 2; i += 8) {
-        for (int j = 0; j < 8; j++) {
-          printf("%d %d\t", mod_dmrs[i + j].r, mod_dmrs[i + j].i);
-        }
-        printf("\n");
-      }
-#endif
    }
-    uint32_t dmrs_idx = rel15->rbStart;
-    if (rel15->refPoint == 0)
-      dmrs_idx += rel15->BWPStart;
-    dmrs_idx *= dmrs_Type == NFAPI_NR_DMRS_TYPE1 ? 6 : 4;
-    c16_t txdataF_precoding[rel15->nrOfLayers][symbol_sz] __attribute__((aligned(64)));
-    int layer_sz = 0;
-    for (int layer = 0; layer < rel15->nrOfLayers; layer++) {
-      layer_sz = do_onelayer(frame_parms,
-                             slot,
-                             rel15,
-                             layer,
-                             txdataF_precoding[layer],
-                             tx_layers[layer] + re_beginning_of_symbol,
-                             start_sc,
-                             symbol_sz,
-                             l_symbol,
-                             dlPtrsSymPos,
-                             n_ptrs,
-                             amp,
-                             amp_dmrs,
-                             l_prime,
-                             dmrs_Type,
-                             mod_dmrs + dmrs_idx);
-    } // layer loop
-    re_beginning_of_symbol += layer_sz;
-    stop_meas(&gNB->dlsch_resource_mapping_stats);
-
-    start_meas(&gNB->dlsch_precoding_stats);
-    for (int ant = 0; ant < frame_parms->nb_antennas_tx; ant++) {
-      const size_t txdataF_offset_per_symbol = l_symbol * symbol_sz;
-      do_txdataF(txdataF, symbol_sz, txdataF_precoding, gNB, rel15, ant, start_sc, txdataF_offset_per_symbol);
+    reset_meas(&rdata->dlsch_resource_mapping_stats);
+    reset_meas(&rdata->dlsch_precoding_stats);
+    for (int l = 0; l < rel15->nrOfLayers; l++)
+      rdata->tx_layers[l] = tx_layers[l];
+    if (l_symbol < rel15->StartSymbolIndex + rel15->NrOfSymbols - num_pdsch_symbols_per_task) {
+      task_t t = {.func = &nr_pdsch_symbol_processing, .args = rdata};
+      pushTpool(&gNB->threadPool, t);
+    } else {
+      nr_pdsch_symbol_processing(rdata);
    }
-    stop_meas(&gNB->dlsch_precoding_stats);
+  }
+  join_task_ans(&ans);
+  for (int i = 0; i < nb_tasks; i++) {
+    merge_meas(&gNB->dlsch_resource_mapping_stats, &arr[i].dlsch_resource_mapping_stats);
+    merge_meas(&gNB->dlsch_precoding_stats, &arr[i].dlsch_precoding_stats);
  }
  stop_meas(&gNB->dlsch_pdsch_generation_stats);
  /* output and its parts for each dlsch should be aligned on 64 bytes (or 8 * 64 bits)
@@ -485,6 +485,7 @@ typedef struct PHY_VARS_gNB_s {
  notifiedFIFO_t L1_rx_out;
  tpool_t threadPool;
  int num_pusch_symbols_per_thread;
+  int num_pdsch_symbols_per_thread;
  int dmrs_num_antennas_per_thread;
  pthread_t L1_rx_thread;
  int L1_rx_thread_core;
@@ -382,6 +382,8 @@ int main(int argc, char **argv)
  uint8_t  dlsch_threads = 0;
  int chest_type[2] = {0};
  uint8_t  max_ldpc_iterations = 5;
+  // number of PDSCH symbols per thread = 0 means do not use thread pool
+  int num_pdsch_symbols_per_thread = 0;
  if ((uniqCfg = load_configmodule(argc, argv, CONFIG_ENABLECMDLINEONLY)) == 0) {
    exit_fun("[NR_DLSIM] Error, configuration module init failed\n");
  }
@@ -583,6 +585,10 @@ int main(int argc, char **argv)
      gNBthreads[sizeof(gNBthreads)-1]=0;
      break;

+    case 'Y':
+      num_pdsch_symbols_per_thread = atoi(optarg);
+      break;
+
    case 'Z' :
      filename_csv = strdup(optarg);
      AssertFatal(filename_csv != NULL, "strdup() error: errno %d\n", errno);
@@ -657,6 +663,7 @@ int main(int argc, char **argv)
      printf("-T Enable PTRS, arguments list L_PTRS{0,1,2} K_PTRS{2,4}, e.g. -T 2 0 2 \n");
      printf("-U Change DMRS Config, arguments list DMRS TYPE{0=A,1=B} DMRS AddPos{0:2} DMRS ConfType{1:2}, e.g. -U 3 0 2 1 \n");
      printf("-X gNB thread pool configuration, n => no threads\n");
+      printf("-Y Number of symbols processed per PDSCH generation thread\n");
      printf("-Z Output filename (.csv format) for stats\n");
      exit (-1);
      break;
@@ -715,6 +722,7 @@ int main(int argc, char **argv)

  AssertFatal((gNB->if_inst = NR_IF_Module_init(0)) != NULL, "Cannot register interface");
  gNB->if_inst->NR_PHY_config_req = nr_phy_config_request;
+  gNB->num_pdsch_symbols_per_thread = num_pdsch_symbols_per_thread;

  NR_ServingCellConfigCommon_t *scc = calloc(1,sizeof(*scc));;
  prepare_scc(scc);
@@ -28,6 +28,10 @@
 #define L1_MAX_LDPC_ITERATIONS               "max_ldpc_iterations"
 #define L1_RX_THREAD_CORE                    "L1_rx_thread_core"
 #define L1_TX_THREAD_CORE                    "L1_tx_thread_core"
+#define L1_NUM_RX_SYM_PER_THREAD             "L1_num_rx_sym_per_thread"
+#define HLP_L1_NUM_RX_SYM_PER_THREAD         "number of symbols processed per PUSCH generation thread"
+#define L1_NUM_TX_SYM_PER_THREAD             "L1_num_tx_sym_per_thread"
+#define HLP_L1_NUM_TX_SYM_PER_THREAD         "number of symbols processed per PDSCH generation thread"
 #define HLP_TP_SIZ "thread_pool_size paramter removed, please use --thread-pool"
 #define L1_TX_AMP_BACKOFF_dB                 "tx_amp_backoff_dB"
 #define HLP_L1TX_BO "Backoff from full-scale output at the L1 entity(frequency domain), ex. 12 would corresponding to 14-bit input level (6 dB/bit). Default 36 dBFS for OAI RU entity"
@@ -58,6 +62,8 @@
  {L1_MAX_LDPC_ITERATIONS,               NULL,       0,         .uptr=NULL,           .defintval=8,               TYPE_UINT,     0},         \
  {L1_RX_THREAD_CORE,                    NULL,       0,         .iptr=NULL,           .defintval=-1,              TYPE_INT,      0},         \
  {L1_TX_THREAD_CORE,                    NULL,       0,         .iptr=NULL,           .defintval=-1,              TYPE_INT,      0},         \
+  {L1_NUM_RX_SYM_PER_THREAD,             HLP_L1_NUM_RX_SYM_PER_THREAD, 0, .iptr=NULL, .defintval=1,               TYPE_INT,      0},         \
+  {L1_NUM_TX_SYM_PER_THREAD,             HLP_L1_NUM_TX_SYM_PER_THREAD, 0, .iptr=NULL, .defintval=0,               TYPE_INT,      0},         \
  {L1_TX_AMP_BACKOFF_dB,                 HLP_L1TX_BO,0,         .uptr=NULL,           .defintval=36,              TYPE_UINT,     0},         \
  {L1_PHASE_COMP,                        HLP_L1_PHASE_COMP,PARAMFLAG_BOOL, .uptr=NULL,.defintval=1,               TYPE_UINT,     0},         \
  {L1_NUM_ANTENNAS_PER_THREAD,           HLP_NUM_ARX,0,         .uptr=NULL,           .defintval=1,               TYPE_UINT,     0},         \
@@ -841,6 +841,9 @@ void RCconfig_NR_L1(void)
      gNB->L1_rx_thread_core = *gpd(params, np, L1_RX_THREAD_CORE)->iptr;
      gNB->L1_tx_thread_core = *gpd(params, np, L1_TX_THREAD_CORE)->iptr;
      LOG_I(NR_PHY, "thread cores for L1_RX %d L1_TX %d\n", gNB->L1_rx_thread_core, gNB->L1_tx_thread_core);
+      // PUSCH symbols per thread need to be calculated by how many threads we have
+      gNB->num_pusch_symbols_per_thread = *gpd(params, np, L1_NUM_RX_SYM_PER_THREAD)->iptr;
+      gNB->num_pdsch_symbols_per_thread = *gpd(params, np, L1_NUM_TX_SYM_PER_THREAD)->iptr;
      gNB->TX_AMP = min(32767.0 / pow(10.0, .05 * (double)(*gpd(params, np, L1_TX_AMP_BACKOFF_dB)->uptr)), INT16_MAX);
      LOG_I(NR_PHY, "TX_AMP = %d (-%d dBFS)\n", gNB->TX_AMP, *gpd(params, np, L1_TX_AMP_BACKOFF_dB)->uptr);
      AssertFatal(gNB->TX_AMP > 300, "TX_AMP is too small, must be larger than 300 (is %d)\n", gNB->TX_AMP);