ASM1166 on board Topton N21 can't passthough to VM

[Tander]

My setup

• Topton N21
• Proxmox 9.1.9
• nvme for OS, 3 HDD connected to onboard ASM1166

Problem
ASM1166 working just fine in bios (I see drives) and Proxmox (it sees drives), also in dmesg:
[0.343620] pci 0000:04:00.0: [1b21:1166] type 00 class 0x010601 PCIe Endpoint
[0.343661] pci 0000:04:00.0: BAR 0 [mem 0xfce82000-0xfce83fff]
[0.343669] pci 0000:04:00.0: BAR 5 [mem 0xfce80000-0xfce81fff]
[0.343672] pci 0000:04:00.0: ROM [mem 0xfce00000-0xfce7ffff pref]
[0.343723] pci 0000:04:00.0: PME# supported from D0 D3hot D3cold

But it unable to initialize in VM when I passthough it. In dmesg I always see this:
[0.509960] pci 0000:01:00.0: [1b21:1166] type 7f class exffffff conventional PCI [
[0.510064] pci 0000:01:00.0: unknown header type 7f, ignoring device [

Attempted solutions that did NOT resolve the issue (with reboot after each try, as ASM1166 doesn't recover from that dead state)

• IOMMU and virtualization
  enabled both in bios and Proxmox
• IOMMU group
  ASM1166 in separate IOMMU group
• Kernel command line parameters in GRUB
  amd_iommu=on iommu=pt libata.force=nolpm pcie_aspm=off pcie_port_pm=off
  (followed by update-grub)
• vfio-pci module configuration
  Created /etc/modprobe.d/vfio.conf with:
  options vfio-pci ids=1b21:1166
  Updated initramfs.
• Blacklisting the ahci driver
  File /etc/modprobe.d/ahci-blacklist.conf:
  

  options vfio-pci ids=1b21:1166
  options vfio-pci disable_idle_d3=1
  softdep ahci pre: vfio-pci

• Disabling ASPM and power management
  Parameters: pcie_aspm=off, pcie_port_pm=off, libata.force=nolpm
• Firmware update of ASM1166
  240625-0000-05 (factory), 241224-0000-00, 211108-0000-00, 240625-0000-05 (sent by manufacturer as factory)
• Changing Proxmox kernels
  Installed and booted: proxmox-kernel-6.17.13-7-pve-signed, proxmox-kernel-6.14.11-4-pve-signed.
• Manual driver binding/unbinding
  echo 0000:04:00.0 > /sys/bus/pci/drivers/vfio-pci/bind
echo 0000:05:00.0 > /sys/bus/pci/devices/0000:05:00.0/driver/unbind
  to check if unbinding doesn't cause this, and adapter kept working after that
• Hardware reset of the device and parent device
  echo 1 > /sys/bus/pci/devices/0000:04:00.0/reset
echo 1 > /sys/bus/pci/devices/0000:00:02.4/reset
• PCIe power management control
  echo on > /sys/bus/pci/devices/0000:04:00.0/power/control
  Checked power_state – device remained in D3hot before starting VM and D0 after that.
• Disabling reset methogs
  echo "" > /sys/bus/pci/devices/0000:05:00.0/reset_method
• disable_idle_d3 parameter
  options vfio-pci disable_idle_d3=1
• Editing VM configuration (/etc/pve/qemu-server/101.conf)
  Various combinations of hostpci (with rombar=0 and 1, pcie=0 and 1)
  Tested all combinations of bios (seabios and ovmf) and machine(q35 and i440fx)
• Different guest OSs
  TrueNAS Scale 25 (latest), TrueNAS Scale 26 (beta) and Ubuntu 24.04 LTS as guest OS

At this point I almost lost all hope to make it work and thinking about buying proper HBA (like LSI 9207-8i) and mostly posting this to make sense from a week of fruitless debugging, but if you have any input - i'd be glad to hear it!

 

[celemine1gig]

Probably a firmware and or hardware issue with the ASM1166. Someone else also recently complained, here:

S

Post in thread 'Problems with PCIe passthrough with two identical devices'

Sunday at 19:57

This is all great and good, but I’m surprised that this patch was made only for 1164, I have similar devices, but on an asm1166 chip, and they have exactly the same problem. I think that this problem applies to the entire 116* line and requires a general patch in the kernel. Tell me, how can I report this so that I can check and make changes to the update?

01:00.0 SATA controller: ASMedia Technology Inc. ASM1166 Serial ATA Controller (rev 02) (prog-if 01 [AHCI 1.0])
        Subsystem: ZyDAS Technology Corp. Device 2116
        !!! Unknown header type 7f
        Memory at 80b82000...

• sacredx72

Maybe you are better at reading, than the other guy.