I recently replaced my GT 1030 with an RTX 5060 and now GPU passthrough fails — the VM won't even start.

With the GT 1030, passthrough worked fine even without binding the GPU to VFIO (it passed through successfully regardless). However, with the RTX 5060, it simply doesn't work. I've double- and triple-checked my configuration, and when I swap back to the GT 1030, everything works normally again.

Has anyone successfully passed through an RTX 5060 in Unraid? I'm wondering if there are known compatibility issues or special steps required for this card.

I've already tried:

• Double-checking BIOS settings (Above 4G Decoding, SR-IOV, etc.)

• Ensuring IOMMU groups are properly set up

• Updating Unraid to the latest stable version

Edited July 28, 2025Jul 28 by Yang0hui

Please see diagnostics attached

 

Thank you for your help 🙏 

unraid-diagnostics-20250728-2028.zip

Today I got an RTX 4060, and passthrough is working normally. Are there any Unraid experts here? Why can't the 5060 passthrough? I'm willing to provide any necessary clues to resolve this issue. The GPU is brand new, and I don’t want it to go to waste.

Hey,

I was in the same ditch as you, but mine was RTX 5050. There are issues with it. I am depressed and returned it and bought an RTX 4060.

I guess we will have to wait until 5000 series is fully supported for unraid.

On 8/9/2025 at 2:58 AM, PUNGGOKERS said:

Hey,

I was in the same ditch as you, but mine was RTX 5050. There are issues with it. I am depressed and returned it and bought an RTX 4060.

I guess we will have to wait until 5000 series is fully supported for unraid.

I still have my RTX 5060 here. I'm planning to build a temporary Ubuntu system using a 12th or 13th gen i3 for now, and once Unraid resolves the issue, I'll reinstall the GPU and switch back to it.

Did you ever try to use this in the append line:

vfio-pci.disable_idle_d3=1

This will prevent the GPU to go to d3cold before even attempting to turn on the VM. My issue was I can't even open the VM because it automatically goes to d3cold which gives the unknow pci header type 127.

I was able to run the VM but I am currently in black screen now. Let me know if you tried that one.

I was reading your log, you are at the same ditch as me, try adding that in your append line, it may start working.

The problem:

Unknown PCI Header type 127.

Cause, according to SpaceInvaderOne, the reason this happens is when GPU is abruptly shutdown, the only way to fix this is to restart the server or use his script (which works btw).

However, the problem remains when we run the VM and gives the same message again.

Possible Solutions in Append line:

vfio-pci.disable_idle_d3=1       # Prevent GPU power state issues - This prevents the D3cold state.
----
vfio-pci.ids=xxxx:xxxx,xxxx:xxxx # Bind your GPU + audio to VFIO
video=efifb:off                  # Disable EFI framebuffer
pcie_aspm=off                    # Disable PCIe power management
pci=noats                        # Disable Address Translation Services
intel_iommu=on iommu=pt          # Enable IOMMU passthrough

After adding the vfio-pci.disable_idle_d3=1 to the append line, this what it should look like if you run this in the terminal: dmesg | grep -i vfio

root@Punggokers:~# dmesg | grep -i vfio 
[    0.000000] Command line: BOOT_IMAGE=/bzimage isolcpus=0-17 vfio-pci.ids=10de:2882,10de:22be vfio-pci.disable_idle_d3=1 video=efifb:off pcie_aspm=off intel_iommu=on iommu=pt initrd=/bzroot 
[    0.088917] Kernel command line: BOOT_IMAGE=/bzimage isolcpus=0-17 vfio-pci.ids=10de:2882,10de:22be vfio-pci.disable_idle_d3=1 video=efifb:off pcie_aspm=off intel_iommu=on iommu=pt initrd=/bzroot 
[    1.779123] VFIO - User Level meta-driver version: 0.3 
[    1.779152] vfio-pci 0000:01:00.0: vgaarb: VGA decodes changed: olddecodes=io+mem,decodes=io+mem:owns=none 
[    1.779190] vfio-pci 0000:82:00.0: vgaarb: VGA decodes changed: olddecodes=io+mem,decodes=io+mem:owns=none 
[    1.779223] vfio_pci: add [10de:2882[ffffffff:ffffffff]] class 0x000000/00000000 
[    1.779330] vfio_pci: add [10de:22be[ffffffff:ffffffff]] class 0x000000/00000000 
[   34.475709] vfio-pci 0000:01:00.0: vgaarb: VGA decodes changed: olddecodes=io+mem,decodes=io+mem:owns=none 
[   34.475711] vfio-pci 0000:82:00.0: vgaarb: VGA decodes changed: olddecodes=io+mem,decodes=io+mem:owns=none root@Punggokers:~#

This tells us that the GPU and Audio is properly binded and is now ready to be in a VM.

The next problem you'll encounter is if you put your GPU + Audio as your primary GPU and Audio, and run the VM you'll probably get a BLACK SCREEN. type this in the terminal dmesg | grep -i vfio after running the VM and you should probably get the following below :

root@Punggokers:~# dmesg | grep -i vfio
[    0.000000] Command line: BOOT_IMAGE=/bzimage isolcpus=0-17 vfio-pci.ids=10de:2882,10de:22be vfio-pci.disable_idle_d3=1 video=efifb:off pcie_aspm=off intel_iommu=on iommu=pt initrd=/bzroot
[    0.089377] Kernel command line: BOOT_IMAGE=/bzimage isolcpus=0-17 vfio-pci.ids=10de:2882,10de:22be vfio-pci.disable_idle_d3=1 video=efifb:off pcie_aspm=off intel_iommu=on iommu=pt initrd=/bzroot
[    1.769221] VFIO - User Level meta-driver version: 0.3
[    1.769254] vfio-pci 0000:01:00.0: vgaarb: VGA decodes changed: olddecodes=io+mem,decodes=io+mem:owns=none
[    1.769293] vfio-pci 0000:82:00.0: vgaarb: VGA decodes changed: olddecodes=io+mem,decodes=io+mem:owns=none
[    1.769323] vfio_pci: add [10de:2882[ffffffff:ffffffff]] class 0x000000/00000000
[    1.769451] vfio_pci: add [10de:22be[ffffffff:ffffffff]] class 0x000000/00000000
[   34.399695] vfio-pci 0000:01:00.0: vgaarb: VGA decodes changed: olddecodes=io+mem,decodes=io+mem:owns=none
[   34.399697] vfio-pci 0000:82:00.0: vgaarb: VGA decodes changed: olddecodes=io+mem,decodes=io+mem:owns=none
[  404.394745] vfio-pci 0000:01:00.1: enabling device (0000 -> 0002)
[  405.546733] vfio-pci 0000:01:00.1: vfio_bar_restore: reset recovery - restoring BARs
[  405.546845] vfio-pci 0000:01:00.0: vfio_bar_restore: reset recovery - restoring BARs
[  406.266298] vfio-pci 0000:01:00.0: timed out waiting for pending transaction; performing function level reset anyway
[  407.426295] vfio-pci 0000:01:00.0: not ready 1023ms after FLR; waiting
[  408.514290] vfio-pci 0000:01:00.0: not ready 2047ms after FLR; waiting
[  410.626281] vfio-pci 0000:01:00.0: not ready 4095ms after FLR; waiting
[  414.850269] vfio-pci 0000:01:00.0: not ready 8191ms after FLR; waiting
[  423.554237] vfio-pci 0000:01:00.0: not ready 16383ms after FLR; waiting
[  440.450175] vfio-pci 0000:01:00.0: not ready 32767ms after FLR; waiting
[  476.290040] vfio-pci 0000:01:00.0: not ready 65535ms after FLR; giving up
[  476.572567] vfio-pci 0000:01:00.0: vfio_bar_restore: reset recovery - restoring BARs
[  476.572721] vfio-pci 0000:01:00.1: vfio_bar_restore: reset recovery - restoring BARs
[  476.584156] vfio-pci 0000:01:00.0: vfio_bar_restore: reset recovery - restoring BARs
[  476.584312] vfio-pci 0000:01:00.1: vfio_bar_restore: reset recovery - restoring BARs
[  476.593936] vfio-pci 0000:01:00.0: vfio_bar_restore: reset recovery - restoring BARs
[  476.594092] vfio-pci 0000:01:00.1: vfio_bar_restore: reset recovery - restoring BARs
[  476.599060] vfio-pci 0000:01:00.0: vfio_bar_restore: reset recovery - restoring BARs
[  476.599552] vfio-pci 0000:01:00.0: vfio_bar_restore: reset recovery - restoring BARs
[  476.605684] vfio-pci 0000:01:00.0: vfio_bar_restore: reset recovery - restoring BARs
[  476.605820] vfio-pci 0000:01:00.0: vfio_bar_restore: reset recovery - restoring BARs
[  476.605841] vfio-pci 0000:01:00.0: vfio_bar_restore: reset recovery - restoring BARs
[  476.605867] vfio-pci 0000:01:00.1: vfio_bar_restore: reset recovery - restoring BARs
[  476.605965] vfio-pci 0000:01:00.1: vfio_bar_restore: reset recovery - restoring BARs
[  476.605986] vfio-pci 0000:01:00.1: vfio_bar_restore: reset recovery - restoring BARs
[  476.606064] vfio-pci 0000:01:00.0: vfio_bar_restore: reset recovery - restoring BARs
[  476.606084] vfio-pci 0000:01:00.1: vfio_bar_restore: reset recovery - restoring BARs
[  476.615625] vfio-pci 0000:01:00.0: vfio_bar_restore: reset recovery - restoring BARs
[  476.615761] vfio-pci 0000:01:00.1: vfio_bar_restore: reset recovery - restoring BARs
[  476.620394] vfio-pci 0000:01:00.0: vfio_bar_restore: reset recovery - restoring BARs
[  476.620493] vfio-pci 0000:01:00.0: vfio_bar_restore: reset recovery - restoring BARs
[  476.620514] vfio-pci 0000:01:00.1: vfio_bar_restore: reset recovery - restoring BARs
[  476.620612] vfio-pci 0000:01:00.1: vfio_bar_restore: reset recovery - restoring BARs
[  476.640481] vfio-pci 0000:01:00.0: vfio_bar_restore: reset recovery - restoring BARs
[  476.640665] vfio-pci 0000:01:00.1: vfio_bar_restore: reset recovery - restoring BARs
[  476.708380] vfio-pci 0000:01:00.0: vfio_bar_restore: reset recovery - restoring BARs
[  476.708556] vfio-pci 0000:01:00.1: vfio_bar_restore: reset recovery - restoring BARs
root@Punggokers:~# 

This will give us an FLR issue: The VFIO logs show that the GPU is having serious reset issues - it's failing to complete a Function Level Reset (FLR) and timing out after 65+ seconds.

The solution, when you create a VM, only put the Primary Graphics Card, NO SOUND CARD, and it should run without a black screen and should boot properly. This may be a quirk of the consumer cards, but at least it should work.

---

Warning: You need to have an iGPU on top of your primary GPU. I tried running unraid with one GPU, it will still give you a black screen. Make sure Unraid uses your iGPU or something else. Unraid must absolutely not use your VM GPU.

Edited August 14, 2025Aug 14 by PUNGGOKERS

17 minutes ago, PUNGGOKERS said:

Warning: You need to have an iGPU on top of your primary GPU. I tried running unraid with one GPU, it will still give you a black screen. Make sure Unraid uses your iGPU or something else. Unraid must absolutely not use your VM GPU.

According to this guy's suggestion, I successfully passed through the GPU and got the VM up and running. Not sure if Unraid will officially provide a perfect solution to this issue

Hello, I am having the same no display issue after updating from 4060 ti to 5060 ti.

I did the append line things and not setting the sound card in the VM. Still unable to passthrough this card. (I am coming from a working 4060 ti vm setup)

Can someone help me? I don't even know what logs to pull... Tell me what to do.

On 10/11/2025 at 12:30 AM, chilman408 said:

Hello, I am having the same no display issue after updating from 4060 ti to 5060 ti.

I did the append line things and not setting the sound card in the VM. Still unable to passthrough this card. (I am coming from a working 4060 ti vm setup)

Can someone help me? I don't even know what logs to pull... Tell me what to do.

Can you type this in the terminal BEFORE and AFTER running the VM? Is it just black screen?

dmesg | grep -i vfio 

[ 1.574629] VFIO - User Level meta-driver version: 0.3

[ 41.513044] vfio-pci 0000:01:00.0: vgaarb: deactivate vga console

[ 41.513078] vfio-pci 0000:01:00.0: vgaarb: VGA decodes changed: olddecodes=io+mem,decodes=io+mem:owns=none

[ 41.536595] vfio-pci 0000:01:00.0: Enabling HDA controller

[ 316.100314] vfio-pci 0000:01:00.0: Enabling HDA controller

[ 316.337277] vfio-pci 0000:01:00.1: enabling device (0000 -> 0002)

On 10/16/2025 at 9:28 PM, chilman408 said:

[ 1.574629] VFIO - User Level meta-driver version: 0.3

[ 41.513044] vfio-pci 0000:01:00.0: vgaarb: deactivate vga console

[ 41.513078] vfio-pci 0000:01:00.0: vgaarb: VGA decodes changed: olddecodes=io+mem,decodes=io+mem:owns=none

[ 41.536595] vfio-pci 0000:01:00.0: Enabling HDA controller

[ 316.100314] vfio-pci 0000:01:00.0: Enabling HDA controller

[ 316.337277] vfio-pci 0000:01:00.1: enabling device (0000 -> 0002)

Type it before running and print screen it, then type it again after running the VM then prints screen it.

This does not tell me anything.