ESXI 6.7 passthrough GeForce GT 1030 to Ubuntu

ewok2

New Member
Apr 7, 2021
27
1
1
Hello
am trying to configure a vm (on ESXI) with Ubuntu which will have a GPU NVIDIA Corporation GP108 [GeForce GT 1030].
It works with 2 GPU

Code :
  1. lspci | grep VGA
  2. 00:0f.0 VGA compatible controller: VMware SVGA II Adapter
  3. 03:00.0 VGA compatible controller: NVIDIA Corporation GP108 [GeForce GT 1030] (rev a1)


But I have the feeling that the VMware SVGA is the one used...

if I deactivate it on ESXI with "svga.present = FALSE "
as describe here: https://elatov.github.io/2018/04/es [...] o-plex-vm/
the vm does not boot anymore... or does not answer to ping

any idea to deactivate the SVGA from VMWare?

Thanks

 

Rand__

Well-Known Member
Mar 6, 2014
5,779
1,292
113
Cant you simply remove the virtual video card from the vm config?
 

ewok2

New Member
Apr 7, 2021
27
1
1
depending on which answer I follow it give different answer...

glxinfo|egrep "OpenGL vendor|OpenGL renderer"
OpenGL vendor string: Mesa/X.org
OpenGL renderer string: llvmpipe (LLVM 11.0.0, 256 bits)
seems to show that it is not the NVIDIA

lspci -vnnn | perl -lne 'print if /^\d+\:.+(\[\S+\:\S+\])/' | grep VGA
00:0f.0 VGA compatible controller [0300]: VMware SVGA II Adapter [15ad:0405] (prog-if 00 [VGA controller])
03:00.0 VGA compatible controller [0300]: NVIDIA Corporation GP108 [GeForce GT 1030] [10de:1d01] (rev a1) (prog-if 00 [VGA controller])
and this show both...

So I still don't know for sure :)
 

ewok2

New Member
Apr 7, 2021
27
1
1
I did not try the low rank advise...
and it does not work...
nvidia-smi -L
NVIDIA-SMI has failed because it couldn't communicate with the NVIDIA driver. Make sure that the latest NVIDIA driver is installed and running.
despite the fact I installed the "ubuntu-drivers devices"
ubuntu-drivers devices
== /sys/devices/pci0000:00/0000:00:0f.0 ==
modalias : pci:v000015ADd00000405sv000015ADsd00000405bc03sc00i00
vendor : VMware
model : SVGA II Adapter
driver : open-vm-tools-desktop - distro free

== /sys/devices/pci0000:00/0000:00:15.0/0000:03:00.0 ==
modalias : pci:v000010DEd00001D01sv00001462sd00008C98bc03sc00i00
vendor : NVIDIA Corporation
model : GP108 [GeForce GT 1030]
driver : nvidia-driver-450-server - distro non-free
driver : nvidia-driver-390 - distro non-free
driver : nvidia-driver-450 - third-party non-free
driver : nvidia-driver-418-server - distro non-free
driver : nvidia-driver-460-server - distro non-free
driver : nvidia-driver-460 - third-party non-free recommended
driver : xserver-xorg-video-nouveau - distro free builtin
 

ewok2

New Member
Apr 7, 2021
27
1
1
superempie of course I am on a vm test...
But I did not see the open-vm-tools place to disable the driver...
And I did not see it on the link you send ?
 

ewok2

New Member
Apr 7, 2021
27
1
1
Sory I forget to answer... no monitor attached. I will try to use this vm with "ssh -x vm" :)

for information I now get this answer :
sudo inxi -Gx
X11 connection rejected because of wrong authentication.
X11 connection rejected because of wrong authentication.
Graphics: Device-1: VMware SVGA II Adapter driver: vmwgfx v: 2.15.0.0 bus ID: 00:0f.0
Device-2: NVIDIA GP108 [GeForce GT 1030] vendor: Micro-Star MSI driver: nvidia v: 460.67 bus ID: 03:00.0
Display: server: X.org 1.20.9 driver: vmware FAILED: nvidia unloaded: fbdev,modesetting,nouveau,vesa tty: 145x43
Message: Advanced graphics data unavailable for root.
and

nvidia-smi
Unable to determine the device handle for GPU 0000:03:00.0: Unknown Error
 

superempie

Member
Sep 25, 2015
70
9
8
The Netherlands
Hmm, ok. Don't have experience with that scenario. Interesting though. :)
My system runs ESXi 6.7U3 latest build with 2 GPU's, one for Mint and one for Windows 10 for gaming. I have passed though Nvidia GPU's just for compute, but not for X.
 

superempie

Member
Sep 25, 2015
70
9
8
The Netherlands
Sounds cool to check out if you can get it working. I just connect monitors to my GPU's, unless they are compute VM with GPU's passed through, like a BOINC VM. I use X, but I am by no means an expert on that matter.
 

ewok2

New Member
Apr 7, 2021
27
1
1
Stange thing... I have seen in another post that "Unable to determine the device handle for GPU 0000:03:00.0: Unknown Error" was link with hypervisor.cpuid.v0 = "FALSE"

and indeed this parameter has disapear form my vm.vmx file...
I have added it with the graphical web interface and know I have the folowing answer :

nvidia-smi -L
GPU 0: GeForce GT 1030 (UUID: GPU-d6d4c15e-82c7-5403-9c17-b10ecd99adb5)
Which seem's better. but I still think that the vmware one is used...
The NVIDIA card is capable of OpenGL 4.5 version :

glxinfo | grep -i version
server glx version string: 1.4
client glx version string: 1.4
GLX version: 1.4
Version: 20.2.6
Max core profile version: 4.5
Max compat profile version: 3.1
Max GLES1 profile version: 1.1
Max GLES[23] profile version: 3.2
OpenGL core profile version string: 4.5 (Core Profile) Mesa 20.2.6
OpenGL core profile shading language version string: 4.50
OpenGL version string: 3.1 Mesa 20.2.6
OpenGL shading language version string: 1.40
OpenGL ES profile version string: OpenGL ES 3.2 Mesa 20.2.6
OpenGL ES profile shading language version string: OpenGL ES GLSL ES 3.20
GL_EXT_shader_implicit_conversions, GL_EXT_shader_integer_mix,

But when I try to launch a app that require OpenGL 4.3 it says that at least OpenGL 4.3 is required...

I will try to plug a screen to the card to see something...