🧩 PCIe Fix

When a VM fails with “QEMU exited with code 1” during PCIe passthrough, it is often tied to RMRR/DMAR restrictions, legacy BIOS boot mode, or missing kernel flags. This guide provides a structured path to resolve passthrough failures on Proxmox VE using supported kernel settings and (when necessary) the relaxable‑RMRR kernel workflow.

info

Modern Proxmox kernels already include the relaxable‑RMRR patch (6.2.16-13-pve and newer). You still must enable the kernel option to activate it.

Prerequisites

• ✅ Proxmox VE 7.4+ (Debian 12 / 11)
• ✅ BIOS/UEFI access (prefer UEFI over Legacy)
• ✅ VT-d (Intel) or AMD‑Vi enabled
• ✅ PCIe device that supports passthrough
• ✅ Root shell access on the node

warning

Legacy BIOS and poorly isolated IOMMU groups can block PCIe passthrough. Always plan downtime and validate changes after each step.

Configuration Steps

Option A: Terminal (Recommended)

1) Confirm firmware mode

If possible, switch the host to UEFI. Legacy BIOS often complicates passthrough behavior.

2) Verify IOMMU is active

dmesg | grep -e DMAR -e IOMMU

Expected examples:

• DMAR: IOMMU enabled
• AMD-Vi: Interrupt remapping enabled

3) Enable relaxable‑RMRR (Intel only)

Edit GRUB:

nano /etc/default/grub

Update the kernel line:

GRUB_CMDLINE_LINUX_DEFAULT="quiet intel_iommu=on,relax_rmrr iommu=pt intremap=no_x2apic_optout"

Apply and reboot:

update-grub
reboot

Verify the option is active:

dmesg | grep 'Intel-IOMMU'

Expected output similar to:

DMAR: Intel-IOMMU: assuming all RMRRs are relaxable

4) (Proxmox 7.4 / Kernel 5.15 only) Use patched kernel

If you are on kernel 5.15 and need relaxable‑RMRR, install a patched kernel release and reboot into it.

General steps:

1. Download the appropriate kernel .deb packages from the releases page.
2. Install them with:

   dpkg -i *.deb

3. Reboot and confirm your kernel ends with pve-relaxablermrr:

   uname -r

4. Enable relax_rmrr in GRUB as shown above.

5) Confirm PCI devices and IOMMU groups

pvesh get /nodes/<your-node>/hardware/pci --pci-class-blacklist ""

6) Retry VM passthrough

After the above changes, attach the PCIe device to the VM and boot again.

Option B: Proxmox Web UI

1. Enable VT‑d/AMD‑Vi and UEFI in BIOS/UEFI.
2. Reboot the host.
3. Use Node → System → PCI Devices to confirm the device appears.
4. Add PCI device to VM hardware and boot.
5. If it still fails, return to Option A and enable relax_rmrr for Intel.

Understanding the Configuration

Flow Diagram

[UEFI + VT-d/AMD-Vi]
      ↓
[IOMMU Enabled in Kernel]
      ↓
[relax_rmrr (Intel only)]
      ↓
[Clean IOMMU Groups]
      ↓
[VM PCIe Passthrough]

Components Table

Component	Purpose	Notes
UEFI Boot	Stable firmware path	Preferred over Legacy BIOS
IOMMU	Enables device isolation	Required for passthrough
relax_rmrr	Allows restricted PCI devices	Intel platforms only
Patched kernel	Needed for 5.15 series	Not required on 6.2.16+
IOMMU groups	Isolation boundary	Better split = safer

Verification

1. Kernel flags active

   cat /proc/cmdline

2. IOMMU enabled

   dmesg | grep -e DMAR -e IOMMU

3. Relaxable RMRR active (Intel)

   dmesg | grep 'Intel-IOMMU'

4. VM boots with PCI device attached

Troubleshooting

Issue	Cause	Solution
QEMU exited with code 1	RMRR restriction	Enable relax_rmrr or use patched kernel on 5.15
Device invisible in VM	IOMMU off	Enable VT‑d/AMD‑Vi and GRUB args
Passthrough fails on Legacy BIOS	Firmware mismatch	Switch host to UEFI
Device in shared IOMMU group	Poor isolation	Enable ACS or change hardware
No Intel-IOMMU message	Flag missing	Recheck GRUB + update-grub
Kernel doesn’t include patch	Old kernel	Upgrade Proxmox kernel or use patched 5.15
VM boots but device not functional	Driver issues	Install guest drivers for the device
Host uses device driver	Not unbound	Blacklist driver and rebuild initramfs
ACS override ignored	Kernel args not applied	Reboot after update-grub
Passthrough unstable	Platform limitation	Use supported hardware/NICs
VM won’t start after changes	GRUB typo	Fix /etc/default/grub and re-run update-grub
IOMMU groups unchanged	BIOS limits	Toggle ACS in BIOS or adjust hardware

Advanced Options

1. Enable ACS override only if you understand the risk
2. Pin vCPUs to reduce latency jitter
3. Use iommu=pt for performance
4. Isolate host devices from passthrough NICs
5. Use separate management NICs
6. Keep kernel and firmware updated
7. Test with a small VM before production
8. Document IOMMU groups after every change
9. Use vfio-pci.ids to bind by device ID
10. Maintain rollback kernel in GRUB

Related Guides

• PCI Passthrough
• Post-Install Configuration
• Install MikroTik CHR
• OpenWRT LXC Auto Creation

References

• Proxmox forum discussion: https://forum.proxmox.com/threads/qemu-exited-with-code-1-pcie-passthrough-not-working.146297/
• Proxmox VE Kernel Builder: https://github.com/brunokc/pve-kernel-builder

Completion

✅ PCIe passthrough failure resolved!

Next steps:

• Attach PCI devices and validate in the guest OS
• Monitor stability after changes
• Document kernel versions and boot args