No free virtual function found vGPU S7150

erfant

Recently installed a pair of Radeon S7150's (single slot) gpu's in a couple of HPE DL360 G9 servers.

The cards showed up as passthrough capable in XCP-NG Center,
I then installed the host drivers (mxgpu-2.0.0.amd),
rebooted,
MxGPU (AMD vGPU) was now enabled.

I assign a vGPU to a Windows VM and when i tried to start the VM I got the below error in XCP-NG

"Internal Error: No free virtual function found"
Xen orchestra also produced the same error...

I attempted the solution outlined in this thread, https://xcp-ng.org/forum/topic/3404/firepro-s7150x2-sr-iov-errors
and after a reboot I was never greeted to the xcp-ng server screen and the server no longer connected to the network.
I had to reboot go through the list of options in the grub boot screen until I managed to get one that booted back into XCP-NG, vGPU still does not work.

send help

tuxen

@erfant could you upload lspci -k and dmesg ? Also, have you checked/tried:

SR-IOV enabled in BIOS;
On vmlinuz entry, boot with pci=realloc;
On vmlinuz entry, boot with pci=realloc pci=assign-busses

erfant

lspci -k.txt dmesg.txt

@tuxen I've attached both outputs, they're quite long actually.
I can also confirm SR-IOV is definitely enabled in the bios (double checked).

I'm not too sure what you mean steps 2 & 3.

tuxen

@erfant after seeing your uploaded dmesg, the steps 2 & 3 boot options can be put aside for while because the error isn't the same as the other topics.

The log is showing MxGPU driver probe/initialization errors. After some digging, could be the case of a GPU firmware being incompatible with UEFI. Do you have any spare server for testing XCP-ng boot in legacy/BIOS with this GPU?

[  119.418930]        gim error:(gim_probe:123) gim_probe(08:00.0)
[  121.145663]        gim error:(wait_cmd_complete:2387)  wait_cmd_complete -- time out after 0.003044131 sec
[  121.145719]        gim error:(wait_cmd_complete:2390)   Cmd = 0x17, Status = 0x0, cmd_Complete=0
[  121.145984]        gim error:(init_register_init_state:4643) Failed to INIT PF for initial register 'init-state'

Edited for clarification.

erfant

@tuxen
That's very bizarre..
I don't have any spare servers for testing at the moment.
I do have a NVMe drive plugged into one of the PCIe slots, could that be causing an issue?

olivierlambert

Unrelated: just wanted to chime in and thanks @tuxen for your great community help!

tuxen

@erfant probably not because the nvme driver is loaded and there're no nvme errors in the logs.

@olivierlambert thank you and your team for this great project and community! It's a nice place to share knowledge and learn new stuff. I learn a lot here!