GPU passthrough to linux client.

ComfySofa

Member
Feb 7, 2021
59
2
13
55
Hi there. As above this could be simple for some to answer (i hope)

First and foremost im running a small system (an optiplex 3050, i5, 32gb, 2.5tb's worth of space.

Ive got the gpu passthrough working (intel 630 graphics) and that is tied to a windows vm (all good)

I appreciate its only a small card and understand that if i want to use it for another vm the windows one has to be powered off. I built a fedora Vm of which i wanted the gpu to passthrough to. Ive passed the gpu through to it....the vm is running in uefi mode (i note that i didnt have to do anywhere near as much config to get it working) the windows one i had to add some config to a file first from recollection.

Ive got VNC installed so right now i can remote in and get a black screen (gpu passed over) - i have read that i need to get one of those dummy dongles?vand also read that i could possible do it with software?

Im not opposed to running another distro if its easier....i only want it for nothing more than exposure and learning linux a bit more by using it as a day to day desktop machine (hence the gpu passthrough running) - if anyone has a link to a walkthrough...

Many thanks to anyone that takes a swing at it.!
 
Found the answer to my question. You have to create a dummy adaptor....im on Fedora so heres what worked for me....

sudo yum install xorg-x11-drv-dummy

This line didnt work....but here it is anyway... sudo cp /etc/X11/xorg.conf /etc/X11/xorg.conf.bak

I had to chmod the X11 folder but then used this... sudo cp /etc/X11/vncserver-virtual-dummy.conf /etc/X11/xorg.conf

Complete novice with linux so while i know roughly what the above does...it did however work... and then i was remoting in, using VNC