-
Notifications
You must be signed in to change notification settings - Fork 171
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Update baseimg from nvidia/cudagl to nvidia/cuda #315
Conversation
I wonder what do we do with |
WRT CI failure, looks like we hit docker/docker-py#3113 I'm currently trying to figure out how we workaround that. |
Thanks @patrick-5546 |
Could you please rebase onto latest |
@patrick-5546 I seem to be getting issues with it not finding a suitable Vulkan surface renderer when it tries to bring up the editor interface. It compiles the shaders fine but after that it fails. Have you run 5.0+ with the CUDA base image? Edit: I found we have to do the following apt install https://gitlab.com/nvidia/container-images/vulkan/-/blob/master/docker/Dockerfile.ubuntu#L28 to get it to report the vulkan devices properly. |
This is my fix. I have the 5.2 editor running in Docker. |
I believe I also had to install |
Okay good to know. Maybe I will make a PR for this. Now called |
The Docker image
nvidia/cudagl
has been deprecated and replaced withnvidia/cuda
. The new image has the updated cuda repo signing key so thatsudo apt update
works, but deprecates some tags: see the Announcement section of https://hub.docker.com/r/nvidia/cuda for more details, and #276 for some context.What I did is replace every instance of
cudagl
withcuda
in the codebase.References