Hi I'm running PVE 9.0.5 on a GMKTek NucBox. I think I am close to getting the iGPU working on a Jellyfin LXC (unprivileged) container. See my LXC details below. I assume that since neofetch reports the iGPU from the Proxmox host, that the passthrough is occurring?

I have attempted to follow the instructions on the Jellyfin intel documentation, However transcoding does not work in Jellyfin and there are a number of other issues I'm experiencing in the container, including:
1). When I attempt to add the following on the host (from the jellyfish documentation) to /etc/pve/lxc/302.conf, the LXC will not start up until I remove it.
and,
2). The following occurs when I try to run intel_gpu_top:
I could use some help on where things are going wrong. Is the iGPU passed through to the OS or not?
Many thanks.

I have attempted to follow the instructions on the Jellyfin intel documentation, However transcoding does not work in Jellyfin and there are a number of other issues I'm experiencing in the container, including:
1). When I attempt to add the following on the host (from the jellyfish documentation) to /etc/pve/lxc/302.conf, the LXC will not start up until I remove it.
Code:
lxc.cgroup2.devices.allow: c 226:0 rwm
lxc.cgroup2.devices.allow: c 226:128 rwm
lxc.mount.entry: /dev/dri/renderD128 dev/dri/renderD128 none bind,optional,create=file
and,
2). The following occurs when I try to run intel_gpu_top:
Code:
root@jellyfin:~# intel_gpu_top
intel_gpu_top: ../tools/intel_gpu_top.c:557: get_num_gts: Assertion `!errno || errno == ENOENT' failed.
Aborted
I could use some help on where things are going wrong. Is the iGPU passed through to the OS or not?
Many thanks.
Last edited: