libvirt.libvirtError: internal error: libxenlight failed to attach pci device 0000:04:00.0

np, i don’t really have much experience with qubes. I’ll try this when i wake up in the morning… it’s 6 am for me, having a restless night with cptsd.

1 Like

Alright. So where do I reset them? Also I think part of the issue may be I didn’t pci block the audio device (was I supposed to???) – that may be part of it, after sleeping on it.

Your GPU can consist of several PCI devices and you need to passthrough all of them:

I do know all of the PCI devices that my GPU has, I just didn’t pass all of them through – just the GPU itself. My bad. Let me go edit my grub and see if it makes it work. Ty.

added the devices to the hvm, and now it’s giving the error: “qrexec-daemon startup failed: 2023-12-29 15:12:41.637 qrexec-daem,on[8642]: qrexec-daemon.c:144:sigchld_parent_handler: Connection to the VM failed” whenever I try to boot up the Windows HVM

And it says this message -immediately- after booting up, then closes down. I did try the button you were talking about, the strict reset button – and I tried it on both pci devices, and still nothing.

though I will say – when it does show for the 1 second it’s up, it says “guest not initialized” – looking into this now

going to try to use the qvm-create-windows-qube again – this time manually installing the windows tools

Nope. Still doesn’t work, even with the qvm-create-windows-qube setup.

Set the qube memory to 2 GB and try to boot for a test.

the qube boots fine without the pci devices. just, when I add them, it says that connection message

Try to boot qube with PCI devices and memory set to 2 GB.

I’ll try that tomorrow - ty

If it runs with let’s say 1848MB RAM, after that you want to patch xen.xml and related:

Having the ram on either 1848MB or 2048MB didn’t change anything, unfortunately.

Sorry to hear. What I do in these situations is to run in dom0 terminal sudo journalctl -f, then trying to start qube, following dom0 terminal. When it finishes, ctrl+c in terminal, and copy paste the part of the journal here in order to be helped. It is very rare not to be able to start win qube under neither circumstances.

1 Like

It’s np. I kind of break every bit of technology around me by nature, so… lol. But anyways, I ran the command, booted up the vm, the issue happened, then it closed – but the output was… a bit long. so I threw it in an unlisted Pastebin just bc I didn’t want to spam this thread – anyways, here you go: [novum@dom0 Desktop]$ journalctl -fJan 01 19:01:01 dom0 run-parts[135901]: (/e - Pastebin.com

I am seriously hoping this works, because I’m contemplating dual-booting with windows if this fails… but I know that compromises my security 10 fold, so I’m hoping we can figure this out. Which I just wanted to shout out a thank you for everybody that’s tried to help so far, as well as anyone who may. Thank you guys.

Also, instinct tells me somebody may ask if I tried to custom install – I did not. I used the average 4.2.0 iso, flashed it to a thumbdrive, then installed it. Nothing custom about it – even though the whole “hostname=?” is the way it is and all those.

My bad everyone – I was a complete idiot and swapped the two gpus in my mind… so I was trying to hide, passthrough the active gpu. My sincere apologies. That said, there still is an issue that I’m in need of help for – I’m making a new thread now, since it’s not related to this fully anymore.