PCI Passthrough Device Not Working

If a VPS with a passed-through PCI/USB device fails to start, or the device doesn't appear inside the guest, the cause is almost always that the host's IOMMU isn't actually enabled, not a Virtlix configuration problem.

Check IOMMU is on: confirm virtualization extensions are enabled in BIOS (Intel VT-d or AMD-Vi, not just VT-x/AMD-V), and that the kernel was booted with IOMMU support - check dmesg | grep -i iommu for a line confirming it initialized. On Intel systems this typically needs intel_iommu=on on the kernel command line; on AMD, amd_iommu=on. This requires a host reboot to take effect if it wasn't already set.

Check the device's IOMMU group: a device can only be passed through cleanly if it's in its own IOMMU group, or every other device sharing that group is also passed through (or unused). Hypervisor Assets' own IOMMU group scan shows you this directly - if your target device shares a group with something you need the host to keep using itself (e.g. an onboard chipset device), passthrough for just that one device won't fully isolate cleanly, which commonly shows up as the guest failing to start with a VFIO-related error in dmesg on the host.

Check the device isn't already claimed by a host driver: for a GPU specifically, the host's own driver (nouveau, amdgpu, the proprietary NVIDIA driver) needs to NOT be bound to that device - it needs to be bound to vfio-pci instead before it can be attached to a guest.

If all of the above checks out and the guest still fails to start, virsh dumpxml <vm-name> on the host and dmesg immediately after a failed start attempt almost always show the specific underlying error.