[Bug]: Nvidia Configure Flatpak loop

[battall]

Describe the Bug

i get this error in loop when trying to run docker container, it restarts forever.

[ /etc/cont-init.d/80-configure_flatpak.sh: executing... ]
**** Configure Flatpak ****
mount: /proc: none already mounted or mount point busy.
dmesg(1) may have more information after failed mount system call.

Steps to Reproduce

1. have an gtx 1080 with 530 driver. (idk if its relevant) (edit: same error happens with rtx 3060 ti too)
2. try to run docker

Expected Behavior

it should continue installation

Screenshots

No response

Relevant Settings

• docker-ce and nvidia-container-runtime
• edited .env docker runtime nvidia instead of runc

Version

Build: [2023-07-22 02:43:14] [master] [af1c17f] [debian]

Platform

Distribution: Ubuntu - 18.04.6 LTS (Bionic Beaver)
Linux Kernel: 5.4.0-150-generic x86_64 x86_64 GNU/Linux
GPU Driver versions: NVIDIA-SMI 530.41.03 Driver Version: 530.41.03 CUDA Version: 12.1 |
Docker: Docker version 24.0.2, build cb74dfc

Relevant log output

No response

[Josh5]

I'm unable to reproduce this. Perhaps the kernel version is too old.

[deviantintegral]

I'm seeing the same thing, but I don't actually have a nvidia card. Actually, I don't have any real video cards as I'm using this as a host for caching downloads.

I'm on Ubuntu 20.04.6 LTS, but a very similar kernel of 5.4.0-156-generic.

Running bisect I can see the issue was introduced in one of these three commits. I'm guessing the first one, since I'm only ever using the Debian build.

commit 26b775f66eeebc9abb0c4ce3ddd25929476c4020 (refs/bisect/bad)
Author: Josh.5 <jsunnex@gmail.com>
Date:   Mon Jul 10 17:34:45 2023 +1200

    Rebase debian build to Debian 12 (Bookworm)

commit 555bd8991de2c842272ee4a3f5bb90dcee8f9604 (refs/bisect/skip-555bd8991de2c842272ee4a3f5bb90dcee8f9604)
Author: Josh.5 <jsunnex@gmail.com>
Date:   Mon Jul 10 17:33:05 2023 +1200

    Some more tweaks to the Arch build

commit 42ebc826b8548ba4f8838cb4ce5ceb4c8a5fd201 (HEAD, refs/bisect/skip-42ebc826b8548ba4f8838cb4ce5ceb4c8a5fd201)
Author: Josh.5 <jsunnex@gmail.com>
Date:   Sun Jul 9 20:23:42 2023 +1200

    Update Arch build with latest changes

[janhelleman]

Describe the Bug

i get this error in loop when trying to run docker container, it restarts forever.

[ /etc/cont-init.d/80-configure_flatpak.sh: executing... ] **** Configure Flatpak **** mount: /proc: none already mounted or mount point busy. dmesg(1) may have more information after failed mount system call.

Steps to Reproduce

1. have an gtx 1080 with 530 driver. (idk if its relevant) (edit: same error happens with rtx 3060 ti too)
2. try to run docker

Expected Behavior

it should continue installation

Screenshots

No response

Relevant Settings

• docker-ce and nvidia-container-runtime
• edited .env docker runtime nvidia instead of runc

Version

Build: [2023-07-22 02:43:14] [master] [af1c17f] [debian]

Platform

Distribution: Ubuntu - 18.04.6 LTS (Bionic Beaver) Linux Kernel: 5.4.0-150-generic x86_64 x86_64 GNU/Linux GPU Driver versions: NVIDIA-SMI 530.41.03 Driver Version: 530.41.03 CUDA Version: 12.1 | Docker: Docker version 24.0.2, build cb74dfc

Relevant log output

No response

Did you find a solution?

[battall]

Did you find a solution?

updating the ubuntu version (which includes kernel updates) fixed the problem for me, but if you cant update, delete that file and container works fine (at least did for me), run 'docker exec rm /etc/cont-init.d/80-configure_flatpak.sh' right after docker run

[janhelleman]

Did you find a solution?

updating the ubuntu version (which includes kernel updates) fixed the problem for me, but if you cant update, delete that file and container works fine (at least did for me), run 'docker exec rm /etc/cont-init.d/80-configure_flatpak.sh' right after docker run

Awesome thank you, I'll give it a shot when I am at home!