Pfsense vm fails on intitial boot

[budney76]

Hello,

I have been trying to get Pfsense CE 2.5 to boot in a vm for a couple of days now. I've gone over the wiki, and many other blogs and forum posts without success.

As soon as I start the vm I get "Error: start failed: QEMU exited with code1". The output is:

kvm: -device vfio-pci,host=0000:0b:00.0,id=hostpci3.0,bus=pci.0,addr=0x1c.0,multifunction=on: vfio 0000:0b:00.0: device is already attached
TASK ERROR: start failed: QEMU exited with code 1

The only output that doesn't match up is: dmesg | grep -e DMAR -e IOMMU -e AMD-Vi (no DMAR listed)

[ 1.111855] pci 0000:00:00.2: AMD-Vi: IOMMU performance counters supported
[ 1.118752] pci 0000:00:00.2: AMD-Vi: Found IOMMU cap 0x40
[ 1.118753] pci 0000:00:00.2: AMD-Vi: Extended features (0x58f77ef22294a5a):
[ 1.118754] AMD-Vi: Interrupt remapping enabled
[ 1.118818] AMD-Vi: Lazy IO/TLB flushing enabled
[ 1.119663] perf/amd_iommu: Detected AMD IOMMU #0 (2 banks, 4 counters/bank).

I hope this issue can be resolved, any help will be greatly appreciated. Thanks!

My hardware is:
Ryzen 3700x
Asus Prime x570-p
Nvidia 8400 GS (for troubleshooting/temp)
Dell H310 Sata controller (for future trueNAS vm)
4 port nic Intel 82576
Realtek nic (for Pi-hole vm)
2 TB HDDs in mirror

 

[dcsapak]

can you post the vm config ? (qm config ID)

 

[budney76]

can you post the vm config ? (qm config ID)

# qm config 100
agent: 1
balloon: 0
boot: order=scsi0;ide2
cores: 2
cpu: qemu64
hostpci2: 0b:00
hostpci3: 0b:00
ide2: local:iso/pfSense-CE-2.5.1-RELEASE-amd64.iso,media=cdrom
memory: 2048
name: pfsense
numa: 0
onboot: 1
ostype: l26
scsi0: vm-storage:vm-100-disk-0,size=50G
scsihw: virtio-scsi-pci
smbios1: uuid=4a007142-12c8-4f08-a5e7-3f97c0f4b342
sockets: 1
vmgenid: c4e9321f-3e67-406d-a9a3-586a7ac736c1

 

[dcsapak]

you have passed through the same device twice, this cannot work:

hostpci2: 0b:00
hostpci3: 0b:00

 

[budney76]

you have passed through the same device twice, this cannot work:

Thanks for the help dcsapak!

Interesting, looking at that it does look like I am passing through the same device/port. Going back and deleting the two nic ports, and adding all 4 ports from the same Intel card



The output of qm config 100 is now:

hostpci0: 09:00
hostpci1: 09:00
hostpci2: 0b:00
hostpci3: 0b:00

It looks like the .0 or the .1 at the end of the ID has no barring in the vm settings???

So how do I separate the IDs enough to pass through each port??? Or is their another work around?

 

[dcsapak]

you have to deselect 'all functions' or simply use each device only once (the .X are the pci 'functions')

 

[budney76]

you have to deselect 'all functions' or simply use each device only once (the .X are the pci 'functions')

Thank you, removing the pci-e devices, then adding them back in with out "all functions" selected worked.

 

[tricochat]

Hello fellows,
I'm stuck on the same case, but with other config I guess..
I get the same error than you had : start failed: QEMU exited with code 1

Thanks for any help

----SERVER CONFIG
CPU(s) 8 x Intel(R) Xeon(R) CPU E3-1280 v6 @ 3.90GHz (1 Socket)
Kernel Version Linux 6.8.4-2-pve (2024-04-10T17:36Z)
Boot Mode EFI

CPU usage 0.82% of 8 CPU(s)
RAM 15.45 GiB
SWAP usage 9.63% (789.18 MiB of 8.00 GiB)


-----VM CONFIG
qm config 100
boot: order=ide2;virtio0;net0
cores: 4
cpu: host
ide2: local:iso/pfSense.iso,media=cdrom,size=1402120K
memory: 2048
meta: creation-qemu=8.1.5,ctime=1726708382
name: pfSense
net0: rtl8139=BC:24:11:E5:FC:1B,bridge=vmbr1,firewall=1
numa: 0
ostype: other
scsihw: virtio-scsi-single
smbios1: uuid=01f2840a-d9d5-4e8d-b0d0-634fdbbaf45d
sockets: 1
vga: qxl
virtio0: local1:vm-100-disk-0,cache=writeback,iothread=1,size=10G
vmgenid: 08d80e9e-9fbc-42f8-89a4-e6c1b0e21165