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

Remove code enclosed in NVIPC_CUDA_ENABLE

Browse Source 
When compiling aerial, we compile without CUDA support (see e.g.,
docker/Dockerfile.gNB.aerial.ubuntu22). Hence, do not check for it in
aerial for the moment, we won't use it for the foreseeable future.
This commit is contained in:
Robert Schmidt
2024-10-10 16:02:29 +02:00
parent c648dfd078
commit da5601b8bf
1 changed files with 0 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files
nfapi/oai_integration/aerial/fapi_nvIPC.c
-9
View File
@@ -88,11 +88,6 @@ static int ipc_handle_rx_msg(nv_ipc_t *ipc, nv_ipc_msg_t *msg)
gpu = 0; gpu = 0;
} }
#ifdef NVIPC_CUDA_ENABLE
// Test CUDA: call CUDA functions to change all string to lower case
test_cuda_to_lower_case(test_cuda_device_id, msg->data_buf, TEST_DATA_BUF_LEN, gpu);
#endif
if (gpu) { if (gpu) {
p_cpu_data = cpu_buf_recv; p_cpu_data = cpu_buf_recv;
memset(cpu_buf_recv, 0, RECV_BUF_LEN); memset(cpu_buf_recv, 0, RECV_BUF_LEN);
@@ -647,11 +642,7 @@ int load_hard_code_config(nv_ipc_config_t *config, int module_type, nv_ipc_trans
return -1; return -1;
} }
#ifdef NVIPC_CUDA_ENABLE
int test_cuda_device_id = get_cuda_device_id();
#else
int test_cuda_device_id = -1; int test_cuda_device_id = -1;
#endif
LOG_D(NFAPI_VNF,"CUDA device ID configured : %d \n", test_cuda_device_id); LOG_D(NFAPI_VNF,"CUDA device ID configured : %d \n", test_cuda_device_id);
config->transport_config.shm.cuda_device_id = test_cuda_device_id; config->transport_config.shm.cuda_device_id = test_cuda_device_id;
if (test_cuda_device_id >= 0) { if (test_cuda_device_id >= 0) {