[Ubuntu20.04/5.13.0-1031-aws/NVIDIA T4G/515.43.04] Weston doesn't launch due to no drm device

Problem

I can’t launch weston on AWS EC2 G5g instance/Ubuntu20.04(5.13.0-1031-aws)//NVIDIA T4G/515.43.04.
The weston logs are the following. The root cause seems that weston can’t find DRM/KMS device.

$ cat /tmp/weston.log 
Date: 2022-07-11 UTC
[04:15:43.680] weston 8.0.0
               https://wayland.freedesktop.org
               Bug reports to: https://gitlab.freedesktop.org/wayland/weston/issues/
               Build: 8.0.0
[04:15:43.680] Command line: /usr/bin/weston --backend=drm-backend.so --drm-device=card0 --modules=systemd-notify.so --log=/tmp/weston.log
[04:15:43.680] OS: Linux, 5.13.0-1031-aws, #35~20.04.1-Ubuntu SMP Mon Jun 13 22:33:23 UTC 2022, aarch64
[04:15:43.680] Starting with no config file.
[04:15:43.681] Output repaint window is 7 ms maximum.
[04:15:43.682] Loading module '/usr/lib/aarch64-linux-gnu/libweston-8/drm-backend.so'
[04:15:43.691] initializing drm backend
[04:15:43.693] logind: session control granted
[04:15:43.693] ERROR: DRM device 'card0' is not a KMS device.
[04:15:43.693] no drm device found
[04:15:43.693] fatal: failed to create compositor backend

Question

I’ve confirmed the following information. Is there anything else I should confirm?
Or How to solve the problem?

nvidia-smi

$ nvidia-smi
Mon Jul 11 05:09:03 2022       
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 515.43.04    Driver Version: 515.43.04    CUDA Version: 11.7     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  NVIDIA T4G          On   | 00000000:00:1F.0 Off |                    0 |
| N/A   45C    P8    16W /  70W |     72MiB / 15360MiB |      0%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|    0   N/A  N/A       878      G   /usr/lib/xorg/Xorg                 59MiB |
|    0   N/A  N/A       918      G   /usr/bin/gnome-shell                9MiB |
+-----------------------------------------------------------------------------+

lsmod

I confirmed nvidia module has loaded as follows, right?

$ lsmod | grep -i nvidia
nvidia_drm             69632  3
nvidia_modeset       1196032  3 nvidia_drm
nvidia              40022016  164 nvidia_modeset
drm_kms_helper        286720  1 nvidia_drm
drm                   610304  7 drm_kms_helper,nvidia,nvidia_drm

nvidiadrm modeset

The document says that executing modprobe nvidia-drm modeset=1 is needed to enable DRM KMS.
So I did and confirmed that configuration is set as follows, right?

$ modprobe -c | grep nvidia_drm
options nvidia_drm modeset=1
options nvidia_drm modeset=1
alias pci:v000010DEd*sv*sd*bc03sc00i00* nvidia_drm
alias pci:v000010DEd*sv*sd*bc03sc02i00* nvidia_drm
alias pci:v000010DEd*sv*sd*bc06sc80i00* nvidia_drm