vglrun no respondence

[renhaofan]

I met the same problem as described in Container doesn't start with volume .X11-unix:rw. I follow your advice, manually change the DISPLAY:

docker run -it --rm \
    --device /dev/dri/card0 \
    -v /tmp/.X11-unix:/tmp/.X11-unix:rw \
    -e DISPLAY=:2 \
    -p "35000:5901" \
    -p "35001:6901" \
    -e VNC_PW="" \
    accetto/ubuntu-vnc-xfce-opengl-g3:latest --tail-vnc

After that, I run the command vglrun glmark2, no correspondence.

Then I run the command vglrun -d :2 glmark2, but error occurred like below:

=======================================================
    glmark2 2021.02
=======================================================
    OpenGL Information
    GL_VENDOR:     Mesa/X.org
    GL_RENDERER:   llvmpipe (LLVM 12.0.0, 256 bits)
    GL_VERSION:    3.1 Mesa 21.2.6
=======================================================
[build] use-vbo=false:X Error of failed request:  BadMatch (invalid parameter attributes)
  Major opcode of failed request:  131 (MIT-SHM)
  Minor opcode of failed request:  3 (X_ShmPutImage)
  Serial number of failed request:  43
  Current serial number in output stream:  44

However, vglrun -d :2 glxgears look like good, and no error reported, but it's slower:

headless@2ba116776775:~$ glxgears
11525 frames in 5.0 seconds = 2304.955 FPS
11898 frames in 5.0 seconds = 2379.447 FPS
X connection to :2.0 broken (explicit kill or server shutdown).
headless@2ba116776775:~$ vglrun -d :2 glxgears
4714 frames in 5.0 seconds = 942.387 FPS
4937 frames in 5.0 seconds = 987.321 FPS
[VGL] ERROR: in readback--
[VGL]    269: Window has been deleted by window manager
headless@2ba116776775:~$ 

[accetto]

Thanks for the feedback. I'll check it and let you know.

[accetto]

Hallo @renhaofan,

I can confirm that the command vglrun -d :2 glmark2 (and also vglrun -d :1 glmark2) does not work correctly. It opens the rendering window and it closes it immediatelly. Unfortunately I cannot explain the reason.

However, I don't find such an example on the page you've mentioned above.

There is only the command vglrun glmark2, which works correctly independently of the display number.

Regards, accetto

[renhaofan]

Hallo @renhaofan,

I can confirm that the command vglrun -d :2 glmark2 (and also vglrun -d :1 glmark2) does not work correctly. It opens the rendering window and it closes it immediatelly. Unfortunately I cannot explain the reason.

However, I don't find such an example on the page you've mentioned above.

There is only the command vglrun glmark2, which works correctly independently of the display number.

Regards, accetto

Thanks for your quick response. I have tried some docker images with virtualGL, but all failed. From your repo, I realize myebe that because ${DISPLAY} and QT_X11_NO_MITSHM . Next week, I will try to solve this problem. If I solved that, I will let you know. Anyway, thanks a lot.

[accetto]

You're welcome.

I hope you've found that you can override the VNC/noVNC parameters, including the display number, using an external file. It can simplify experimenting. Check the part At VNC/noVNC startup-time in this Readme file.

[renhaofan]

You're welcome.

I hope you've found that you can override the VNC/noVNC parameters, including the display number, using an external file. It can simplify experimenting. Check the part At VNC/noVNC startup-time in this Readme file.

Yes. Before I created the issue, I have noticed that and test. Thanks for you remind.

[accetto]

FYI: I've just updated the images on Docker Hub. There is a newer VirtualGL 3.0.1. Maybe it will help.

[renhaofan]

FYI: I've just updated the images on Docker Hub. There is a newer VirtualGL 3.0.1. Maybe it will help.

I will try that later and let you know if successful.

By the way, I use this docker image to solve the VirtualGL problem, which is based on the nomachine with VirtualGL naturely. And I noticed you comment the nomachine in your Dockerfile. I hope that docker image is helpful for you.

[accetto]

Closing because of no further actions from the requester.

I do not have the appropriate hardware to develop the subject at the moment.