CUDA driver version is insufficient (Solved)

Autonomous Vehicles DRIVE PX 2 DriveWorks
1

I used the sdkmanager to install driveworks and only choose host machines choice.
but when i used to run samples, like sample_drivenet, it showed CUDA driver version is insufficient.
And I change to CUDA v 9.0 and CUDA v10.0. It doesn’t help.

nvidia-smi:

±----------------------------------------------------------------------------+
| NVIDIA-SMI 390.87 Driver Version: 390.87 |
|-------------------------------±---------------------±---------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
|===============================+======================+======================|
| 0 Quadro P6000 Off | 00000000:17:00.0 On | Off |
| 26% 58C P8 20W / 250W | 490MiB / 24447MiB | 19% Default |
±------------------------------±---------------------±---------------------+
| 1 TITAN V Off | 00000000:65:00.0 Off | N/A |
| 37% 53C P8 30W / 250W | 0MiB / 12066MiB | 0% Default |
±------------------------------±---------------------±---------------------+
| 2 TITAN V Off | 00000000:B6:00.0 Off | N/A |
| 28% 42C P8 26W / 250W | 0MiB / 12066MiB | 0% Default |
±------------------------------±---------------------±---------------------+

±----------------------------------------------------------------------------+
| Processes: GPU Memory |
| GPU PID Type Process name Usage |
|=============================================================================|
| 0 1141 G /usr/lib/xorg/Xorg 196MiB |
| 0 1654 G compiz 151MiB |
| 0 2303 G …quest-channel-token=8929122395094330860 140MiB |
±----------------------------------------------------------------------------+

nvcc -V
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2018 NVIDIA Corporation
Built on Wed_Apr_11_23:16:29_CDT_2018
Cuda compilation tools, release 9.2, V9.2.88

Go to the driverworks
./sample_drivenet

[5-10-2018 17:25:32] Initialize DriveWorks SDK v1.2.400
[5-10-2018 17:25:32] Release build with GNU 4.8.5 from v1.2.0-rc11-0-ga7f5475
[5-10-2018 17:25:32] Platform: Detected Generic x86 Platform
[5-10-2018 17:25:32] TimeSource: monotonic epoch time offset is 1538761212374359
[5-10-2018 17:25:32] Driveworks exception thrown: DW_CUDA_ERROR: Platform: cannot acquire cuda context : CUDA driver version is insufficient for CUDA runtime version

terminate called after throwing an instance of ‘std::runtime_error’
what(): [2018-10-05 17:25:32] DW Error DW_CUDA_ERROR executing DW function:
dwInitialize(&context, DW_VERSION, &sdkParams)
at /builds/driveav/dw/sdk/samples/drivenet/drivenet/main.cpp:594
Aborted (core dumped)

2

Dear ruiruil,

I think you need to re-build the samples on you host machine.
So could you please refer to “Building for the Linux Host System” in DriveWorks doc folder? Thanks.

3

Dear ruiruil,
any update on this? can you check running CUDA 9.2(/usr/local/cuda-9.2) samples on host?

4

cd /usr/local/cuda-9.2/samples
sudo make

it fails.

o -c videoDecodeGL.cpp
videoDecodeGL.cpp: In function ‘void parseCommandLineArguments(int, char**)’:
videoDecodeGL.cpp:668:64: warning: format ‘%d’ expects argument of type ‘int’, but argument 2 has type ‘long int’ [-Wformat=]
printf(“YUV output @ nStartFrame = %d\n”, g_nFrameStart);
^
videoDecodeGL.cpp:673:60: warning: format ‘%d’ expects argument of type ‘int’, but argument 2 has type ‘long int’ [-Wformat=]
printf(“YUV output @ nStartEnd = %d\n”, g_nFrameEnd);
^
/usr/local/cuda-9.2/bin/nvcc -ccbin g++ -m64 -Xcompiler -no-pie -gencode arch=compute_30,code=compute_30 -o cudaDecodeGL FrameQueue.o ImageGL.o VideoDecoder.o VideoParser.o VideoSource.o cudaModuleMgr.o cudaProcessFrame.o dynlink_cuda.o dynlink_nvcuvid.o videoDecodeGL.o -L/usr/lib/nvidia-387 -lGL -lGLU -lglut -L/usr/local/cuda-9.2/lib64/stubs -lcuda -L…/…/common/lib/linux/x86_64 -lcudart -lnvcuvid -lGLEW
/usr/bin/ld: cannot find -lnvcuvid
collect2: error: ld returned 1 exit status
Makefile:421: recipe for target ‘cudaDecodeGL’ failed
make[1]: *** [cudaDecodeGL] Error 1
make[1]: Leaving directory ‘/usr/local/cuda-9.2/samples/3_Imaging/cudaDecodeGL’
Makefile:52: recipe for target ‘3_Imaging/cudaDecodeGL/Makefile.ph_build’ failed
make: *** [3_Imaging/cudaDecodeGL/Makefile.ph_build] Error 2

5

Thank you for your reply and I couldn’t re-build

/usr/local/driveworks-1.2/build$

sudo cmake …/samples/src/drivenet -DCMAKE_BUILD_TYPE=Release -DCMAKE_TOOLCHAIN_FILE=/usr/local/driveworks-1.2/samples/cmake/Toolchain-V5L.cmake -DVIBRANTE_PDK:STRING=/home/byton-ad/nvidia/nvidia_sdk/DRIVE/Linux/5.0.10.3/OS/DriveSDK/drive-t186ref-linux /usr/local/driveworks-1.2/samples/src/drivenet/
– VIBRANTE_PDK = /home/byton-ad/nvidia/nvidia_sdk/DRIVE/Linux/5.0.10.3/OS/DriveSDK/drive-t186ref-linux
– VIBRANTE_PDK_DEVICE = t186ref
– VIBRANTE_PDK_BRANCH = 5.0.10.3
CMake Warning (dev) at /usr/local/driveworks-1.2/samples/cmake/Toolchain-V5L.cmake:127 (add_definitions):
Policy CMP0005 is not set: Preprocessor definition values are now escaped
automatically. Run “cmake --help-policy CMP0005” for policy details. Use
the cmake_policy command to set the policy and suppress this warning.
Call Stack (most recent call first):
/usr/share/cmake-3.5/Modules/CMakeDetermineSystem.cmake:98 (include)
CMakeLists.txt
This warning is for project developers. Use -Wno-dev to suppress it.

– Vibrante version 5.0.10.3
– VIBRANTE_PDK = /home/byton-ad/nvidia/nvidia_sdk/DRIVE/Linux/5.0.10.3/OS/DriveSDK/drive-t186ref-linux
CMake Warning (dev) at /usr/local/driveworks-1.2/samples/cmake/Toolchain-V5L.cmake:127 (add_definitions):
Policy CMP0005 is not set: Preprocessor definition values are now escaped
automatically. Run “cmake --help-policy CMP0005” for policy details. Use
the cmake_policy command to set the policy and suppress this warning.
Call Stack (most recent call first):
/usr/local/driveworks-1.2/build/CMakeFiles/3.5.1/CMakeSystem.cmake:6 (include)
CMakeLists.txt
This warning is for project developers. Use -Wno-dev to suppress it.

– Vibrante version 5.0.10.3
– The C compiler identification is GNU 4.9.4
– The CXX compiler identification is GNU 4.9.4
– Check for working C compiler: /home/byton-ad/nvidia/nvidia_sdk/DRIVE/Linux/5.0.10.3/OS/DriveSDK/drive-t186ref-linux/…/toolchains/tegra-4.9-nv/usr/bin/aarch64-gnu-linux/aarch64-gnu-linux-gcc
– Check for working C compiler: /home/byton-ad/nvidia/nvidia_sdk/DRIVE/Linux/5.0.10.3/OS/DriveSDK/drive-t186ref-linux/…/toolchains/tegra-4.9-nv/usr/bin/aarch64-gnu-linux/aarch64-gnu-linux-gcc – works
– Detecting C compiler ABI info
– Detecting C compiler ABI info - done
– Detecting C compile features
– Detecting C compile features - done
– Check for working CXX compiler: /home/byton-ad/nvidia/nvidia_sdk/DRIVE/Linux/5.0.10.3/OS/DriveSDK/drive-t186ref-linux/…/toolchains/tegra-4.9-nv/usr/bin/aarch64-gnu-linux/aarch64-gnu-linux-g++
– Check for working CXX compiler: /home/byton-ad/nvidia/nvidia_sdk/DRIVE/Linux/5.0.10.3/OS/DriveSDK/drive-t186ref-linux/…/toolchains/tegra-4.9-nv/usr/bin/aarch64-gnu-linux/aarch64-gnu-linux-g++ – works
– Detecting CXX compiler ABI info
– Detecting CXX compiler ABI info - done
– Detecting CXX compile features
– Detecting CXX compile features - done
CMake Error at drivenet/CMakeLists.txt:23 (cuda_add_executable):
Unknown CMake command “cuda_add_executable”.

CMake Warning (dev) in CMakeLists.txt:
No cmake_minimum_required command is present. A line of code such as

cmake_minimum_required(VERSION 3.5)

should be added at the top of the file. The version specified may be lower
if you wish to support older CMake versions for this project. For more
information run “cmake --help-policy CMP0000”.
This warning is for project developers. Use -Wno-dev to suppress it.

– Configuring incomplete, errors occurred!
See also “/usr/local/driveworks-1.2/build/CMakeFiles/CMakeOutput.log”.

6

Dear ruiruil,
The issue is with driver version. CUDA 9.2 requires r396 driver. Can you please update your driver?
PLease check Ubuntu 16.04 CUDA 9.2 with driver 390.59, insufficient driver - CUDA Setup and Installation - NVIDIA Developer Forums for more details. On my ubuntu 16.04 machine, I searched for additional drivers and selected 396 driver version

7

Thank you for your advice, it works, and I think I shouldn’t install cuda 9.0 first as well.