CPU: Ryzen 3900X
Motherboard: MSI B450M mortar
BIOS: tried 7B89v1B, 7B89v1A, 7B89v19, 7B89v17
GPU: RTX2060 Super and RX580
OS: Proxmox 6.0
Configuration:
/etc/default/grub
...
GRUB_CMDLINE_LINUX_DEFAULT="quiet amd_iommu=on iommu=pt video=efifb:off rootdelay=10"
...
/etc/modules
vfio
vfio_iommu_type1
vfio_pci
vfio_virqfd
/etc/modprobe.d/blacklist.conf
blacklist nouveau
blacklist nvidia
blacklist nvidiafb
blacklist snd_hda_codec_hdmi
blacklist snd_hda_intel
blacklist snd_hda_codec
blacklist snd_hda_core
blacklist radeon
blacklist amdgpu
/etc/modprobe.d/kvm.conf
options kvm ignore_msrs=Y
/etc/modprobe.d/vfio-pci.conf
options vfio-pci ids=1002:67df,1002:aaf0 disable_vga=1
-
If I put RTX2060S in slot 1 (cpu) and RX580 card in slot 3 (pch), iommu will broken during os initialization like this:
It boots when i use only one card (both slot 1 and slot 3 works) or change two cards’ position, so there may be some issues with amd’s gpu and chipset (when there’s also an nvidia card plug-in). -
If i start kvm with pcie passthrough, there will be some outputs in dmesg like this, vm will run out of all available rams and no video outputs:
Aug 2 18:30:27 BelliashPC kernel: pcieport 0000:00:03.1: AER: Uncorrected (Non-Fatal) error received: 0000:00:03.1 Aug 2 18:30:27 BelliashPC kernel: pcieport 0000:00:03.1: AER: PCIe Bus Error: severity=Uncorrected (Non-Fatal), type=Transaction Layer, (Requester ID) Aug 2 18:30:27 BelliashPC kernel: pcieport 0000:00:03.1: AER: device [1022:1483] error status/mask=00100000/04400000 Aug 2 18:30:27 BelliashPC kernel: pcieport 0000:00:03.1: AER: [20] UnsupReq (First) Aug 2 18:30:27 BelliashPC kernel: pcieport 0000:00:03.1: AER: TLP Header: 34000000 07000010 00000000 00000000 Aug 2 18:30:27 BelliashPC kernel: pcieport 0000:00:03.1: AER: Device recovery successful
I have tried the same card on intel platforms and it works, so it shouldn’t be problem with vbios.
It happens whether i put the card in slot 1 (cpu) or slot 3 (pch).
I have tried multiple BIOS with no luck, and i have seen similar problem here for problem 1, maybe it’s not fixed for a long time: VFIO / PCI Passthrough Saga, a tragedy in 3 parts
So i think i should buy X570 instead. I want to buy MSI MEG X570 Ace, because it’s cheaper in my country and I like MSI’s BIOS way more than Gigabyte’s, but there’s very few comments about it in this forum.
I found someone who shows an iommu group with this board:
IOMMU group 10: [1022:1482] 00:08.0 Host bridge: Advanced Micro Devices, Inc. [AMD] Starship/Matisse PCIe Dummy Host Bridge
[1022:1484] 00:08.1 PCI bridge: Advanced Micro Devices, Inc. [AMD] Starship/Matisse Internal PCIe GPP Bridge 0 to bus[E:B]
[1022:1484] 00:08.2 PCI bridge: Advanced Micro Devices, Inc. [AMD] Starship/Matisse Internal PCIe GPP Bridge 0 to bus[E:B]
[1022:1484] 00:08.3 PCI bridge: Advanced Micro Devices, Inc. [AMD] Starship/Matisse Internal PCIe GPP Bridge 0 to bus[E:B]
[1022:1485] 30:00.0 Non-Essential Instrumentation [1300]: Advanced Micro Devices, Inc. [AMD] Starship/Matisse Reserved SPP
[1022:149c] 30:00.3 USB controller: Advanced Micro Devices, Inc. [AMD] Matisse USB 3.0 Host Controller
[1022:1487] 30:00.4 Audio device: Advanced Micro Devices, Inc. [AMD] Starship/Matisse HD Audio Controller
[1022:7901] 31:00.0 SATA controller: Advanced Micro Devices, Inc. [AMD] FCH SATA Controller [AHCI mode] (rev 51)
[1022:7901] 32:00.0 SATA controller: Advanced Micro Devices, Inc. [AMD] FCH SATA Controller [AHCI mode] (rev 51)
https://ngabbs.com/read.php?&tid=18238681&pid=358926295&to=1
It doesn’t look that good, and i’m not sure if it can be fixed in further BIOS updates. This guy also has problem 2 like me. Someone says it appears after flashing 1.0.0.3ABB microcode, but i have this problem also in older BIOS versions.
I also see that Linus Tech Tips uses ASUS C8H as its MacOS+Windows dual boot build.
Can anyone give me some advice? Thanks