NvRmMemInitNvmap failed with Permission denied

Autonomous Machines Jetson & Embedded Systems Jetson AGX Orin
, ,
1 
Python 3.8.10 (default, May 26 2023, 14:05:08) 
[GCC 9.4.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import torch
>>> torch.cuda.is_available()
NvRmMemInitNvmap failed with Permission denied
549: Memory Manager Not supported



****NvRmMemInit failed**** error type: 196626


*** NvRmMemInit failed NvRmMemConstructor
/usr/local/lib/python3.8/dist-packages/torch/cuda/__init__.py:115: UserWarning: CUDA initialization: Unexpected error from cudaGetDeviceCount(). Did you run some cuda functions before calling NumCudaDevices() that might have already set an error? Error 801: operation not supported (Triggered internally at /opt/pytorch/pytorch/c10/cuda/CUDAFunctions.cpp:108.)
  return torch._C._cuda_getDeviceCount() > 0
False

docker: nvcr.io/nvidia/l4t-tensorrt:r8.5.2.2-devel
PyTorch installation method: Installing PyTorch for Jetson Platform - NVIDIA Docs

start docker command:

docker run -it --rm --runtime=nvidia --gpus all nvcr.io/nvidia/l4t-tensorrt:r8.5.2.2-devel

If it is the root user, it can get the correct result.

I created a user named “walker” and it cannot be used under that user, resulting in the above error.

The command to create a user is as follows:

USER=walker
useradd --create-home --user-group --groups sudo --shell /bin/bash "$USER"

# Add sudo permission for $USER
mkdir -p /etc/sudoers.d
printf '%s ALL=(ALL:ALL) NOPASSWD: ALL\n' "$USER" | \
  tee /etc/sudoers.d/nopasswd

chmod 440 /etc/sudoers.d/nopasswd
mkdir -p /var/lib/sudo/lectured/
touch /var/lib/sudo/lectured/"$USER"
chmod 600 /var/lib/sudo/lectured/"$USER"
1 Like
2

May I know which platform and SW you’re using?

3 
env | grep JET
JETSON_CUDA_ARCH_BIN=8.7
JETSON_MODULE=NVIDIA Jetson AGX Orin
JETSON_L4T=35.2.1
JETSON_CODENAME=Concord
JETSON_MODEL=Jetson AGX Orin
JETSON_SERIAL_NUMBER=1421322011417
JETSON_SOC=tegra23x
JETSON_P_NUMBER=p3701-0000
JETSON_JETPACK=5.1
5

Hi,

Please check if you are facing the same issue of the below topic:

Thanks.

6

I have tried, but it didn’t work.

7

Hi,

Do you use the below command to launch the container?

docker run -it --rm --runtime=nvidia --gpus all nvcr.io/nvidia/l4t-tensorrt:r8.5.2.2-devel

The --gpu all is a dGPU flag. Could you remove it and try it again?

Thanks.

8

I tried.
Can be used by the root user, but not after creating a new user

9

Hi,

Thanks.
We will give it a check and get back to you later.

10

Okay, thanks.

12

Hi,

Thanks for your patience.

Please add below link to enable the authority:

usermod -aG sudo,video,i2c "$USER"

Thanks.

1 Like
13

Okay, thanks.

I’ll try and get back to you later.

14

Thanks, solved!

16

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.