Vga Compatible Controller: Cirrus Logic Gd 5446 And Opengl
I'm running the following: Ryzen 7 1700, Prime X370 Pro, 64GB RAM, Vega Frontier, Proxmox (most recent beta). I've passed the gpu through to an Ubuntu 16.04 guest OS.
I have a remote computer on AWS with Nvidia GRID card K520 and want to remotely render 3D graphics. I use Ubuntu 14.04 - 64 bit. Remote and local.For remote connection and OpenGL rendering we use VirtualGL and TurboVNC (*amd64.deb).
Ubuntu is running and I can SSH into it, but it seems to have no graphics adapter available for display! Edit: I think this was the case because I needed to restart the entire Proxmox node for various reasons before the card could be picked up again after updating the.conf; See 'Final note' in reply below. Lspci grep -i vga 00:02.0 VGA compatible controller: Cirrus Logic GD 5446 00:05.0 VGA compatible controller: NVIDIA Corporation Device 1b06 (rev a1) 00:06.0 VGA compatible controller: NVIDIA Corporation Device 1b06 (rev a1) dmesg grep NVRM.
I properly installed VirtualGL and TurboVNC and I have Nvidia drivers.
As I understand VirtualGL uses display :0 and TurboVNC uses display :1 on remote machine.
When I connect to the remote machine and I want to run a 3D app (via vglrun ) it gives me error:
extension 'GLX' missing on display ':0.0'.
I don't undrestand how VirtualGL works on a remote machine with TurboVNC, how it uses the display :0 and if it's ok.
Do I need something to setup (.conf, graphic driver). I tried installing it over and over, tried serching for answer, but I didn't find a solution.
1 Answer
Found the solution
I. Recap
The remote computer on AWS (EC2 gl2.2xlarge) is running on 64-bit Ubuntu 14.04. I properly installed VirtualGL, TurboVNC and Nvidia 352.63 drivers (cuda 7.0).
I remotely connected to the aws instance.
Run the TurboVNC server:
/opt/TurboVNC/bin/vncserver
Run TurboVNC client and connect to remote computer:
/opt/TurboVNC/bin/vncviewer
When I wanted to run application on the remote pc there was the error.
/opt/VirtualGL/bin/vglrun <application>
Error: extension “GLX” missing on display “:0.0”
II. Solution
First type on remote pc:
sudo xinit &
Run application on remote pc
/opt/VirtualGL/bin/vglrun <application>
Example:/opt/VirtualGL/bin/vglrun glxgears
sudo xinit &
What does sudo xinit &
?
To get OpenGL applications to run on the remote computer I need to start a second X Server which VirtualGL uses to talk to the NVIDIA gpu. The second X Server will start with the sudo xinit &
.
Some useful checks
Vga Compatible Controller: Cirrus Logic Gd 5446 And Opengl 1
- Check VGA devices on comp:
lspci grep VGA
This shows on AWS (EC2 gl2.2xlarge) with Nvidia GRID card K520:
- Check drivers:
nvidia-smi
If it show some useful information, then the drivers are working:
More enemies fallout 4. NVIDIA-SMI 352.63
Driver Version: 352.63
Cirrus Logic Stock
Some useful info from xorg.conf (/etc/X11/xorg.conf)