maki/openairinterface5g
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

F1-U in gtp-U LTE case, working

Browse Source
This commit is contained in:
Laurent THOMAS
2021-09-22 10:52:38 +02:00
parent c7f9365ccf
commit 32ebb876f4
5 changed files with 27 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files
openair2/F1AP/f1ap_cu_ue_context_management.c
+2 -2
View File
@@ -522,8 +522,8 @@ int CU_send_UE_CONTEXT_SETUP_REQUEST(instance_t instance,
f1ap_ue_context_setup_req->drbs_to_be_setup[i].up_ul_tnl[j].teid=
newGtpuCreateTunnel(getCxt(CUtype, instance)->gtpInst,
f1ap_ue_context_setup_req->rnti,
f1ap_ue_context_setup_req->drbs_to_be_setup[i].drb_id-4,
f1ap_ue_context_setup_req->drbs_to_be_setup[i].drb_id-4,
f1ap_ue_context_setup_req->drbs_to_be_setup[i].drb_id,
f1ap_ue_context_setup_req->drbs_to_be_setup[i].drb_id,
0xFFFF, // We will set the right value from DU answer
addr,
f1ap_ue_context_setup_req->drbs_to_be_setup[i].up_dl_tnl[0].port,
openair2/F1AP/f1ap_du_ue_context_management.c
+9 -3
View File
@@ -55,7 +55,11 @@ boolean_t lteDURecvCb( protocol_ctxt_t *ctxt_pP,
// The buffer comes from the stack in gtp-u thread, we have a make a separate buffer to enqueue in a inter-thread message queue
mem_block_t *sdu=get_free_mem_block(sdu_buffer_sizeP, __func__);
memcpy(sdu->data, sdu_buffer_pP, sdu_buffer_sizeP);
du_rlc_data_req(ctxt_pP,srb_flagP, false, rb_idP,muiP, confirmP, sdu_buffer_sizeP, sdu);
// weird rb id management in 4G, not fully understand (looks bad design)
// overcomplex: if i understand, on the interface DRB start at 4 because there can be SRB 0..3
// but it would be much simpler to use absolute numbering
// instead of this "srb flag" associated to these +/-4
du_rlc_data_req(ctxt_pP,srb_flagP, false, rb_idP-4,muiP, confirmP, sdu_buffer_sizeP, sdu);
return true;
}
@@ -175,6 +179,7 @@ int DU_handle_UE_CONTEXT_SETUP_REQUEST(instance_t instance,
addr,
2152,
lteDURecvCb);
drb_p->up_dl_tnl_length++;
}
}
}
@@ -220,7 +225,6 @@ int DU_handle_UE_CONTEXT_SETUP_REQUEST(instance_t instance,
ctxt.enb_flag = 1;
mem_block_t *pdcp_pdu_p = get_free_mem_block(ieRRC->value.choice.RRCContainer.size, __func__);
memcpy(&pdcp_pdu_p->data[0], ieRRC->value.choice.RRCContainer.buf, ieRRC->value.choice.RRCContainer.size);
/* for rfsim */
du_rlc_data_req(&ctxt, 1, 0x00, 1, 1, 0, ieRRC->value.choice.RRCContainer.size, pdcp_pdu_p);
} else {
LOG_E(F1AP, " RRCContainer in UEContextSetupRequestIEs size id 0\n");
@@ -229,9 +233,11 @@ int DU_handle_UE_CONTEXT_SETUP_REQUEST(instance_t instance,
LOG_W(F1AP, "can't find RRCContainer in UEContextSetupRequestIEs by id %ld \n", F1AP_ProtocolIE_ID_id_RRCContainer);
}
//DU_send_UE_CONTEXT_SETUP_RESPONSE(instance, f1ap_ue_context_setup_req);
if (RC.nrrrc && RC.nrrrc[instance]->node_type == ngran_gNB_DU)
itti_send_msg_to_task(TASK_RRC_GNB, instance, msg_p);
else
// in 4G, race conditon is to fix
DU_send_UE_CONTEXT_SETUP_RESPONSE(instance, f1ap_ue_context_setup_req);
return 0;
}