Code 43 NVIDIA driver error within Windows 10 VM, Geforce 750 Ti Passthrough

[Hipurnism]

all info I have found on the internet, is to use to romfile option for this bug.
also, Maybe try to change pci slot ?

Made no difference. GT630 can only booted once, the Quadro can be booted everytime.

can you post your vm config file ?

Sure:

balloon: 0
bootdisk: virtio0
cores: 4
cpu: host
hostpci0: 04:00,pcie=1,x-vga=on,romfile=/tmp/gt630.rom
hotplug: disk,network,usb
ide2: none,media=cdrom
machine: q35
memory: 8192
name: w10-test01
net0: virtio=0A:C9:E0:EE:94:56,bridge=vmbr0
numa: 0
ostype: win10
scsihw: virtio-scsi-pci
smbios1: uuid=b9c0680a-6068-4bba-bf9e-571d8f76cd3a
sockets: 1
usb0: host=1-1.3
usb1: host=3-4
virtio0: local-lvm:vm-200-disk-1,size=32G

I've tried everything inside the VM I've found online: turn balloning off, fixed size memory allocation... The virtio0 disk is in raw format with no-cache.
I'll rebuild my VMs as I have another W10 one from the same template and inside it the I/O is around 75-90% and the system feels snappier. It still very high for a fresh system, but better.

Update: With sata the disk usage is 0-25%, constantly 0 on an idle system. With virtio its 100%. WTF? Any idea? Does anybody else experiences the same?

I have found some post saying to pass to grub kernel options:

"video=efifb:off"

Man, this did the trick!
This kernel parameter + romfile is the magic combo!
Maybe it'd be best if you'd update the wiki with this info. It might save some souls that want to use a headless host with graphical VM.
Note: with this option the host loses the screen as soon as the kernel loads.

 

[Dorin]

I followed the thread from: https://forum.proxmox.com/threads/gpu-passthrough-tutorial-reference.34303/ and i ended with a Win7 vm which is unable to complete the boot process (https://forum.proxmox.com/threads/gpu-passthrough-tutorial-reference.34303/#post-181095).

Now i have a Win10 os guest vm with the error code 43 reported in device status. If i try to install the video card's driver (GTX 1050 Ti) i will get an error ("This NVIDIA graphics driver is not compatible with this version of Windows." ...) then the installation process is stopped.

Reading this thread i saw some existing solutions for this issue (of error "Code 43"), but i'm a little bit confused and i'm not sure with what should i start..

The current config is:

bios: ovmf
bootdisk: virtio0
cores: 4
cpu: host
efidisk0: local-zfs:vm-101-disk-3,size=128K
ide0: none,media=cdrom
machine: q35
memory: 8192
name: Win10x64
net0: e1000=E2:F8:A4:AC:1A:32,bridge=vmbr0
numa: 0
ostype: win10
scsi0: none,media=cdrom
scsihw: virtio-scsi-pci
smbios1: uuid=35da8b04-26a6-4bf4-8f85-a142eaf5eabd
sockets: 1
virtio0: local-zfs:vm-101-disk-1,cache=none,size=250G
virtio1: local-zfs:vm-101-disk-2,cache=none,size=250G
#hostpci0: 01:00,x-vga=on,pcie=1,romfile=vbios.bin
hostpci0: 01:00,x-vga=on,pcie=1

pveversion -v:

proxmox-ve: 5.0-19 (running kernel: 4.10.17-2-pve)
pve-manager: 5.0-30 (running version: 5.0-30/5ab26bc)
pve-kernel-4.10.17-2-pve: 4.10.17-19
libpve-http-server-perl: 2.0-6
lvm2: 2.02.168-pve3
corosync: 2.4.2-pve3
libqb0: 1.0.1-1
pve-cluster: 5.0-12
qemu-server: 5.0-15
pve-firmware: 2.0-2
libpve-common-perl: 5.0-16
libpve-guest-common-perl: 2.0-11
libpve-access-control: 5.0-6
libpve-storage-perl: 5.0-14
pve-libspice-server1: 0.12.8-3
vncterm: 1.5-2
pve-docs: 5.0-9
pve-qemu-kvm: 2.9.0-3
pve-container: 2.0-15
pve-firewall: 3.0-2
pve-ha-manager: 2.0-2
ksm-control-daemon: 1.2-2
glusterfs-client: 3.8.8-1
lxc-pve: 2.0.8-3
lxcfs: 2.0.7-pve4
criu: 2.11.1-1~bpo90
novnc-pve: 0.6-4
smartmontools: 6.5+svn4324-1
zfsutils-linux: 0.6.5.9-pve16~bpo90

The result from https://forum.proxmox.com/threads/c...e-750-ti-passthrough.23746/page-2#post-131446 is due to the applied patch from https://forum.proxmox.com/threads/c...e-750-ti-passthrough.23746/page-2#post-131464 ?
Does this patch is still valid for PVE 5? In some short words, how should be done?

 

[mcflym]

I am running a GTX 1050ti too and had issues regarding Code 43 in Windows 10 too.
Windows 8.1 worked for me only...

Just try it

 

[Dorin]

I am running a GTX 1050ti too and had issues regarding Code 43 in Windows 10 too.
Windows 8.1 worked for me only...

Just try it

Do i need to follow any other additional instructions excepting of what i have already did for Win7 or Win10 vms?
How does perform the gpu in your guest?
Thank you for feedback.

 

[Dorin]

I tried with Win 8.1 Pro x64 and i'm facing with the same issue, but this time i was able somehow to install the driver.
However, the final conclusion: error code 43.

This was the order of events:
-vm boot
-error code 43
-driver uninstalled from device manager (nvidia)
-driver installed (nvidia)
-vm restarted
-error code 43

 

[mcflym]

Performance is nearly like running a physical machine.

The only thing I can do for you is to post my config:

bios: ovmf
cores: 2
cpu: Haswell-noTSX
hostpci0: 01:00,x-vga=on
numa: 0
ostype: win8

Nothing more in the quest config that is relevant for...

 

[Dorin]

Performance is nearly like running a physical machine.

The only thing I can do for you is to post my config:

bios: ovmf
cores: 2
cpu: Haswell-noTSX
hostpci0: 01:00,x-vga=on
numa: 0
ostype: win8

Nothing more in the quest config that is relevant for...

This is what i have in /etc/default/grub:
GRUB_CMDLINE_LINUX_DEFAULT="quiet intel_iommu=on video=efifbff"

I added disable_vga=1 in /etc/modprobe.d/vfio.conf:
options vfio-pci ids=10de:1c82,10de:0fb9 disable_vga=1

VM's conf file:
hostpci0: 01:00,x-vga=on

By default, without hostpci0 parameters in vm's conf file, i noticed the Microsoft Basic Display Adapter from: PCI bus 0, device 1, function 0 (Location, from Device Manager) which seems to be related with the foolowing entry of IOMMU group 1: /sys/kernel/iommu_groups/1/devices/0000:00:01.0.
Finally, with hostpci0 parameter enabled, i ended again with error code 43 (and error code 32 before driver installation ) ), but location info shown in device manager has changed for the GPU as shown in the previous trial (different bus, different device number): PCI bus 1, device 0, function 0, to: PCI bus 6, device 16, function 0.

I have a filling that somehow it should work because in Win7 vm the GPU was assigned, but didn't boot completely and freeze.
How are these files (/etc/modprobe.d/vfio.conf, /etc/default/grub) configured in your setup?

Thanks.


Additional info:
-Host system:
Xeon 1225 v3 CPU with onboard Intel graphic card.
The BIOS does not offer the possibility to select manually the primary display adapter.
I have an additional PCI-E 2 slot, but the GPU doesn't fit in that place. There is not enough room space in that location of the case.

-GPU in slot 1 of the system:
root@proxmox:~# lspci -n -s 01:00
01:00.0 0300: 10de:1c82 (rev a1)
01:00.1 0403: 10de:0fb9 (rev a1)

-Interrupts available (DMAR: ... ecap f010da):
root@proxmox:~# dmesg | grep ecap
[ 0.026121] DMAR: dmar0: reg_base_addr fed90000 ver 1:0 cap d2008c20660462 ecap f010da
[ 5682.589511] vfio_ecap_init: 0000:01:00.0 hiding ecap 0x19@0x900

-IOMMU groups (three entries in group 1):
root@proxmox:/# find /sys/kernel/iommu_groups/ -type l
/sys/kernel/iommu_groups/7/devices/0000:00:1c.0
/sys/kernel/iommu_groups/5/devices/0000:00:1a.0
/sys/kernel/iommu_groups/3/devices/0000:00:16.3
/sys/kernel/iommu_groups/3/devices/0000:00:16.0
/sys/kernel/iommu_groups/11/devices/0000:03:00.0
/sys/kernel/iommu_groups/1/devices/0000:01:00.1
/sys/kernel/iommu_groups/1/devices/0000:00:01.0
/sys/kernel/iommu_groups/1/devices/0000:01:00.0
/sys/kernel/iommu_groups/8/devices/0000:00:1c.1
/sys/kernel/iommu_groups/6/devices/0000:00:1b.0
/sys/kernel/iommu_groups/4/devices/0000:00:19.0
/sys/kernel/iommu_groups/2/devices/0000:00:14.0
/sys/kernel/iommu_groups/10/devices/0000:00:1f.3
/sys/kernel/iommu_groups/10/devices/0000:00:1f.2
/sys/kernel/iommu_groups/10/devices/0000:00:1f.0
/sys/kernel/iommu_groups/0/devices/0000:00:00.0
/sys/kernel/iommu_groups/9/devices/0000:00:1d.0