Intel Arc Pro B60 - Host Freeze with VFIO Passthrough and ReBAR Enabled

[Qmo37]

Intel Arc Pro B60 - VFIO Passthrough Issue and Small BAR Compute Question

Description

I am experiencing host system freezes when using Intel Arc Pro B60 24GB with VFIO/KVM passthrough on Proxmox VE. I've done some investigation and would like to ask if there are any known workarounds or if my assumptions about the cause are correct.

---

Environment

Proxmox Host

Component	Details
GPU	Intel Arc Pro B60 24GB (PCI ID: 8086:e211)
GPU Firmware	BMG_21.1174
GOP (vBIOS)	23.0.1061
CPU	Intel Core Ultra 7 265K
Motherboard	ASUS PRIME Z890-P WIFI-CSM (BIOS 2401)
RAM	96GB DDR5
Host OS	Proxmox VE 9.x
Host Kernel	6.17.4-1-pve

Guest VM

Component	Details
OS	Ubuntu 24.04.3 LTS
Kernel	6.14.0-1008-intel
Driver	xe

---

What I Observed

With ReBAR Enabled

At VM startup, I see this warning:

QEMU: kvm: warning: vfio_container_dma_map(0x380000000000, 0x800000000, ...) = -22 (Invalid argument)
0000:04:00.0: PCI peer-to-peer transactions on BARs are not supported

BAR configuration (from Proxmox host):

Region 2: Memory at a000000000 (64-bit, prefetchable) [size=32G]
BAR 2: current size: 32GB, supported: 256MB 512MB 1GB 2GB 4GB 8GB 16GB 32GB

IOMMU configuration (from Proxmox host):

DMAR: Host address width 42

After 6-24 hours, the host freezes. I observed three different freeze behaviors:

1. Full Host Freeze: Entire Proxmox becomes unresponsive, requires hard reset
2. Partial Host Freeze: VMs continue running, web UI partially works, but shell/SSH becomes inaccessible
3. VM Freeze: VM shows "running (internal-error)", host stays responsive

Kernel soft lockup (captured from console before host became unresponsive):

[33275.695601] watchdog: BUG: soft lockup - CPU#14 stuck for 5212s! [ps:360867]
[33255.695867]  ? x64_sys_call+0x1151/0x2330
[33255.695868]  ? do_syscall_64+0xb8/0xa30
[33255.695869]  ? x64_sys_call+0x2093/0x2330
[33255.695869]  ? do_syscall_64+0xb8/0xa30
[33255.695870]  ? count_memcg_events+0xd7/0x1a0
[33255.695872]  ? handle_mm_fault+0x254/0x370
[33255.695874]  ? do_user_addr_fault+0x2f8/0x830
[33255.695875]  ? irqentry_exit_to_user_mode+0x2e/0x290
[33255.695877]  ? irqentry_exit+0x43/0x50
[33255.695878]  ? exc_page_fault+0x90/0x1b0
[33255.695879]  entry_SYSCALL_64_after_hwframe+0x76/0x7e

The CPU was stuck for ~87 minutes (5212 seconds) before the watchdog reported it.

With ReBAR Manually Disabled

To test, I manually limited the BAR to 256MB via VFIO-PCI rebind on Proxmox host (and disabled ReBAR option in BIOS):

echo "0000:04:00.0" > /sys/bus/pci/drivers/vfio-pci/unbind
echo "0000:04:00.0" > /sys/bus/pci/drivers/vfio-pci/bind

lspci -vvv -s 04:00.0 | grep "Region 2"
# Region 2: Memory at a000000000 (64-bit, prefetchable) [size=256M]

qm start 101

dmesg | tail -20 | grep -i "vfio\|dma\|p2p"
# [611.243671] vfio-pci 0000:04:00.0: resetting
# [611.350629] vfio-pci 0000:04:00.0: reset done
# [611.375782] vfio-pci 0000:05:00.0: enabling

Result: The host no longer freezes.

However, inside the VM I see:

$ lspci -vvs 01:00.0 | grep Region
    Region 0: Memory at 80000000 (64-bit, non-prefetchable) [size=16M]
    Region 2: Memory at 380000000000 (64-bit, prefetchable) [size=256M]

$ python3 -c "import torch; print(torch.xpu.is_available(), torch.xpu.device_count())"

WARNING: Small BAR detected for device 0000:01:00.0
XPU device count is zero!
False 0

GPU compute is disabled with small BAR configuration.

---

What I Already Tried

Fix Attempted	Result
Update GPU firmware in a bare-metal Windows machine (Windows driver 101.8306)	Freeze still occurred
Update BIOS (2207 → 2401)	Freeze still occurred (escalated from VM freeze to full host freeze)
Disable ASPM (via setpci)	Freeze still occurred
Enable GPU reset methods ("bus")	Immediate host crash when restarting VM
Kernel parameters (pcie_aspm=off, etc.)	Freeze still occurred

---

My Assumptions (I Could Be Wrong)

Based on what I observed:

1. The DMA mapping failure (-22 EINVAL) at address 0x380000000000 with size 0x800000000 might be related to the host's 42-bit IOMMU address width?

2. The host freeze might occur when something tries to access memory regions that weren't properly mapped due to the DMA mapping failure?

3. The kernel soft lockup call stack (exc_page_fault → handle_mm_fault) suggests a page fault handler is waiting for a PCIe/DMA transaction that never completes?

4. The "Small BAR detected" warning in the VM causes the compute runtime to disable XPU functionality?

---

Questions

1. Is there a known workaround for using Arc Pro B60 with VFIO passthrough and ReBAR?

2. Are there any kernel parameters, QEMU options, or driver settings I might have missed?

3. Is there a way to enable compute functionality with small BAR configuration? I understand there might be performance tradeoffs.

4. Are my assumptions about the cause reasonable, or should I be looking at something else?

5. Should I be filing this somewhere else (xe driver GitLab, QEMU, etc.)?

---

Environment Summary

Host:
  Proxmox VE 9.x, Kernel 6.17.4-1-pve
  ASUS PRIME Z890-P WIFI-CSM (BIOS 2401)
  Intel Core Ultra 7 265K, 96GB DDR5
  IOMMU: Host address width 42

GPU:
  Intel Arc Pro B60 24GB (8086:e211)
  Firmware: BMG_21.1174
  GOP: 23.0.1061

Guest VM:
  Ubuntu 24.04.3 LTS, Kernel 6.14.0-1008-intel
  Driver: xe
  PyTorch: 2.7.0+xpu

---

Report date: 2026-01-07
This report was prepared with assistance from Claude Code.

[jamesstringer90]

I have a similar spec host, and experience the same VM freezing, I have experienced the VM freezing (internal error reported by Poxmox) when passing through both a Arc B580 (PCI-E passthrough), and a B50 (in SR-IOV mode split 2 ways).

In my case it's a Windows 11 VM. ReBar is enabled and reported active using GPU-Z in my VM.

Both GPUs are running the latest driver on guest, and both are running up to date GPU firmware.

I observe also that it takes a number of hours for this freeze to happen. I have not experienced my Proxmox host freezing or locking up however.

CPU(s) 20 x Intel(R) Core(TM) Ultra 7 265K (1 Socket)
Kernel Version Linux 6.17.2-1-pve (2025-10-21T11:55Z)
Boot Mode EFI (Secure Boot)
Manager Version pve-manager/9.1.1
Motherboard Z890 AYW GAMING WIFI W
VM client OS: Windows 11 Pro.

[Qmo37]

@jamesstringer90 I'm glad to hear it’s not just me encountering this issue, thanks for the info!

I didn't have Proxmox freezes at first either, but after I tried a reset tweak (seems to be enabling GPU reset), the Proxmox host started to freeze and experience soft lockups with the Ubuntu VM.

I reported this issue to Sparkle, the manufacturer of my B60 card, and they replied yesterday saying that more information regarding the B60 will be released on January 12th.

[kgibala]

Hi @Qmo37,

To help us analyze this issue more effectively, we need some additional information from you:

• In the Guest OS, please run cat on the relevant file and share the output (cat /proc/iomem).
• In the Guest OS, provide the full dmesg log after running the GPU-related workload.
• In the Host OS, please share the full dmesg log and the QEMU command line used to start the VM.

Please set dmesg to the highest verbosity level for logging.

Additionally, it would be very helpful if you could provide detailed reproduction steps so we can attempt to reproduce the issue on our side.

Thank you for your assistance.

[Qmo37]

Thanks for your response @kgibala , i really appreciate it.

Environment

• GPU: Intel Arc Pro B60 24GB (8086:e211)
• Host: Proxmox VE 9.1.4, Kernel 6.17.4-2-pve, QEMU 10.1.2
• Guest: Ubuntu, Kernel 6.14.0-1008-intel
• Current BAR 2 size: 256MB (manually resized from 32GB)

---

1. Guest OS: /proc/iomem

00000000-00000fff : Reserved
00001000-0002ffff : System RAM
00030000-0004ffff : Reserved
00050000-0009efff : System RAM
0009f000-0009ffff : Reserved
000a0000-000bffff : PCI Bus 0000:00
000f0000-000fffff : System ROM
00100000-7c64d017 : System RAM
7c64d018-7c674457 : System RAM
7c674458-7c675017 : System RAM
7c675018-7c69c457 : System RAM
7c69c458-7c69d017 : System RAM
7c69d018-7c6ed057 : System RAM
7c6ed058-7c711fff : System RAM
7c712000-7c735fff : Reserved
7c736000-7e764fff : System RAM
7e765000-7eb32fff : Reserved
7eb33000-7eb58fff : ACPI Tables
7eb59000-7ebfdfff : ACPI Non-volatile Storage
7ebfe000-7ef79fff : System RAM
7ef7a000-7ef7bfff : ACPI Non-volatile Storage
7ef7c000-7effffff : System RAM
7f000000-7fffffff : Reserved
80000000-dfffffff : PCI Bus 0000:00
  80000000-810fffff : PCI Bus 0000:01
    80000000-80ffffff : 0000:01:00.0
      80101018-8010101b : xe.nvm.256
      80102040-801020bf : xe.nvm.256
      8035b000-8035b00f : intel_vsec.telemetry.0
      8035b010-8035b01f : intel_vsec.telemetry.0
      8035b060-8035b077 : intel_vsec.crashlog.1
      8035b078-8035b08f : intel_vsec.crashlog.1
      8035b090-8035c08f : telem1
      8035c090-8035d08f : telem2
      8035dab0-8035e2af : crashlog1
      8035e2b0-8035f2af : crashlog2
      80374000-80374ffb : xe.mei-gscfi.256
    81000000-81003fff : 0000:01:00.1
      81000000-81003fff : ICH HD audio
  81200000-81afffff : PCI Bus 0000:05
    81200000-813fffff : PCI Bus 0000:09
      81200000-81200fff : 0000:09:01.0
    81400000-815fffff : PCI Bus 0000:08
    81600000-817fffff : PCI Bus 0000:07
      81600000-81603fff : 0000:07:1b.0
        81600000-81603fff : xhci-hcd
    81800000-819fffff : PCI Bus 0000:06
      81800000-81800fff : 0000:06:13.0
      81801000-81801fff : 0000:06:12.0
      81802000-81802fff : 0000:06:08.0
      81840000-8187ffff : 0000:06:12.0
      81880000-818bffff : 0000:06:13.0
    81a00000-81a000ff : 0000:05:01.0
    81a01000-81a010ff : 0000:05:02.0
    81a02000-81a020ff : 0000:05:03.0
    81a03000-81a030ff : 0000:05:04.0
  81c00000-81dfffff : PCI Bus 0000:04
  81e00000-81ffffff : PCI Bus 0000:03
  82000000-821fffff : PCI Bus 0000:02
  82400000-82403fff : 0000:00:1b.0
    82400000-82403fff : ICH HD audio
  82404000-82404fff : 0000:00:1f.2
    82404000-82404fff : ahci
  82405000-82405fff : 0000:00:1d.7
    82405000-82405fff : ehci_hcd
  82406000-82406fff : 0000:00:1c.3
  82407000-82407fff : 0000:00:1c.2
  82408000-82408fff : 0000:00:1c.1
  82409000-82409fff : 0000:00:1c.0
  8240a000-8240afff : 0000:00:1a.7
    8240a000-8240afff : ehci_hcd
e0000000-efffffff : PCI ECAM 0000 [bus 00-ff]
  e0000000-efffffff : Reserved
    e0000000-efffffff : pnp 00:03
f0000000-febfffff : PCI Bus 0000:00
fec00000-fec003ff : IOAPIC 0
fed1f410-fed1f414 : iTCO_wdt.1.auto
feffc000-feffffff : Reserved
100000000-b7fffffff : System RAM
  7fce00000-7fe33d1ea : Kernel code
  7fe400000-7ff368fff : Kernel rodata
  7ff400000-7ff8751bf : Kernel data
  7ffd84000-8001fffff : Kernel bss
380000000000-383fffffffff : PCI Bus 0000:00
  380000000000-3807ffffffff : PCI Bus 0000:01
    380000000000-38000fffffff : 0000:01:00.0
  380800000000-380fffffffff : PCI Bus 0000:02
  381000000000-3817ffffffff : PCI Bus 0000:03
  381800000000-381fffffffff : PCI Bus 0000:04
  382000000000-383fffffffff : PCI Bus 0000:05
    382000000000-3827ffffffff : PCI Bus 0000:09
      382000000000-382000003fff : 0000:09:01.0
        382000000000-382000003fff : virtio-pci-modern
    382800000000-382fffffffff : PCI Bus 0000:08
    383000000000-3837ffffffff : PCI Bus 0000:07
    383800000000-383fffffffff : PCI Bus 0000:06
      383800000000-383800003fff : 0000:06:13.0
        383800000000-383800003fff : virtio-pci-modern
      383800004000-383800007fff : 0000:06:12.0
        383800004000-383800007fff : virtio-pci-modern
      383800008000-38380000bfff : 0000:06:08.0
        383800008000-38380000bfff : virtio-pci-modern

---

2. Guest OS: Full dmesg after GPU workload

Note: PyTorch XPU and sycl-ls are not installed. The xe driver loads but reports "Small BAR device":

[    2.309818] xe 0000:01:00.0: vgaarb: deactivate vga console
[    2.310512] xe 0000:01:00.0: [drm] Running in none mode
[    2.310583] xe 0000:01:00.0: [drm] Found battlemage (device ID e211) discrete display version 14.01 stepping B0
[    2.315092] xe 0000:01:00.0: [drm] Using GuC firmware from xe/bmg_guc_70.bin version 70.44.1
[    2.326646] xe 0000:01:00.0: [drm] Using GuC firmware from xe/bmg_guc_70.bin version 70.44.1
[    2.330014] xe 0000:01:00.0: [drm] Using HuC firmware from xe/bmg_huc.bin version 8.2.10
[    2.354859] xe 0000:01:00.0: vgaarb: VGA decodes changed: olddecodes=io+mem,decodes=none:owns=none
[    2.357008] xe 0000:01:00.0: [drm] VISIBLE VRAM: 0x0000380000000000, 0x0000000010000000
[    2.357349] xe 0000:01:00.0: [drm] Small BAR device
[    2.357350] xe 0000:01:00.0: [drm] VRAM[0, 0]: Actual physical size 0x0000000600000000, usable size exclude stolen 0x00000005fa000000, CPU accessible size 0x0000000010000000
[    2.357351] xe 0000:01:00.0: [drm] VRAM[0, 0]: DPA range: [0x0000000000000000-600000000], io range: [0x0000380000000000-380010000000]
[    2.357352] xe 0000:01:00.0: [drm] VRAM: 0x0000000600000000 is larger than resource 0x0000000010000000
[    2.357352] xe 0000:01:00.0: [drm] Total VRAM: 0x0000380000000000, 0x0000000600000000
[    2.357353] xe 0000:01:00.0: [drm] Available VRAM: 0x0000380000000000, 0x00000005fa000000
[    2.370161] xe 0000:01:00.0: [drm] Finished loading DMC firmware i915/bmg_dmc.bin (v2.6)
[    2.549012] xe 0000:01:00.0: [drm] ccs2 fused off
[    2.549014] xe 0000:01:00.0: [drm] ccs3 fused off
[    2.568888] xe 0000:01:00.0: [drm] vcs1 fused off
[    2.568889] xe 0000:01:00.0: [drm] vcs3 fused off
[    2.568890] xe 0000:01:00.0: [drm] vcs4 fused off
[    2.568891] xe 0000:01:00.0: [drm] vcs5 fused off
[    2.568891] xe 0000:01:00.0: [drm] vcs6 fused off
[    2.568891] xe 0000:01:00.0: [drm] vcs7 fused off
[    2.568892] xe 0000:01:00.0: [drm] vecs2 fused off
[    2.568892] xe 0000:01:00.0: [drm] vecs3 fused off
[    2.613151] [drm] Initialized xe 1.1.0 for 0000:01:00.0 on minor 1
[    3.620526] fbcon: xedrmfb (fb0) is primary device
[    3.620529] xe 0000:01:00.0: [drm] fb0: xedrmfb frame buffer device
[    3.640126] xe 0000:01:00.0: [drm] Using mailbox commands for power limits
[    3.640516] xe 0000:01:00.0: [drm] PL2 is supported on channel 0
[    3.650506] Creating 4 MTD partitions on "xe.nvm.256":
[    3.650511] 0x000000000000-0x000000001000 : "xe.nvm.256.DESCRIPTOR"
[    3.651438] 0x000000001000-0x00000054e000 : "xe.nvm.256.GSC"
[    3.652456] 0x00000054e000-0x00000074e000 : "xe.nvm.256.OptionROM"
[    3.653433] 0x00000074e000-0x00000075e000 : "xe.nvm.256.DAM"
[    6.084462] snd_hda_intel 0000:01:00.1: bound 0000:01:00.0 (ops i915_audio_component_bind_ops [xe])

---

3. Host OS: Full dmesg

[    0.000000] Linux version 6.17.4-2-pve #1 SMP PREEMPT_DYNAMIC PMX 6.17.4-2 (2025-12-19T07:49Z) x86_64 GNU/Linux
[    0.098732] Kernel command line: initrd=\EFI\proxmox\6.17.4-2-pve\initrd.img-6.17.4-2-pve root=ZFS=rpool/ROOT/pve-1 boot=zfs quiet intel_iommu=on iommu=pt pci=realloc=off crashkernel=256M
[    0.098766] DMAR: IOMMU enabled
[    0.660656] iommu: Default domain type: Passthrough (set via kernel command line)
[    0.696794] pci 0000:00:02.0: Adding to iommu group 0
[    0.697172] pci 0000:04:00.0: Adding to iommu group 19
[    3.612590] VFIO - User Level meta-driver version: 0.3
[    3.617432] vfio-pci 0000:04:00.0: vgaarb: VGA decodes changed: olddecodes=io+mem,decodes=none:owns=none
[    3.617504] vfio_pci: add [8086:e211[ffffffff:ffffffff]] class 0x000000/00000000
[    3.617540] vfio_pci: add [8086:e2f7[ffffffff:ffffffff]] class 0x000000/00000000
[    8.304976] vfio-pci 0000:04:00.0: vgaarb: VGA decodes changed: olddecodes=none,decodes=io+mem:owns=none
[    8.305130] pci 0000:04:00.0: BAR 2 [mem 0xa000000000-0xa7ffffffff 64bit pref]: releasing
[    8.305144] pcieport 0000:03:01.0: bridge window [mem 0xa000000000-0xa7ffffffff 64bit pref]: releasing
[    8.305145] pcieport 0000:02:00.0: bridge window [mem 0xa000000000-0xa7ffffffff 64bit pref]: releasing
[    8.305152] pcieport 0000:02:00.0: bridge window [mem 0xa000000000-0xa00fffffff 64bit pref]: assigned
[    8.305154] pcieport 0000:03:01.0: bridge window [mem 0xa000000000-0xa00fffffff 64bit pref]: assigned
[    8.305155] pci 0000:04:00.0: BAR 2 [mem 0xa000000000-0xa00fffffff 64bit pref]: assigned
[    8.305207] pci 0000:04:00.0: All device reset methods disabled by user
[    8.305223] pci 0000:05:00.0: All device reset methods disabled by user

---

4. Host OS: QEMU command line

/usr/bin/kvm
-id
101
-name
Ubuntu,debug-threads=on
-no-shutdown
-chardev
socket,id=qmp,path=/var/run/qemu-server/101.qmp,server=on,wait=off
-mon
chardev=qmp,mode=control
-chardev
socket,id=qmp-event,path=/var/run/qmeventd.sock,reconnect-ms=5000
-mon
chardev=qmp-event,mode=control
-pidfile
/var/run/qemu-server/101.pid
-daemonize
-smbios
type=1,uuid=f516cedc-24e6-467a-b650-9021dbc44131
-object
{"id":"throttle-drive-efidisk0","limits":{},"qom-type":"throttle-group"}
-blockdev
{"driver":"raw","file":{"driver":"file","filename":"/usr/share/pve-edk2-firmware//OVMF_CODE_4M.secboot.fd"},"node-name":"pflash0","read-only":true}
-blockdev
{"detect-zeroes":"on","discard":"ignore","driver":"throttle","file":{"cache":{"direct":false,"no-flush":false},"detect-zeroes":"on","discard":"ignore","driver":"raw","file":{"aio":"io_uring","cache":{"direct":false,"no-flush":false},"detect-zeroes":"on","discard":"ignore","driver":"host_device","filename":"/dev/zvol/data2/vm-101-disk-0","node-name":"e3f220a61cfb2d1312d2c4bfe0d3291","read-only":false},"node-name":"f3f220a61cfb2d1312d2c4bfe0d3291","read-only":false,"size":540672},"node-name":"drive-efidisk0","read-only":false,"throttle-group":"throttle-drive-efidisk0"}
-smp
12,sockets=1,cores=12,maxcpus=12
-nodefaults
-boot
menu=on,strict=on,reboot-timeout=1000,splash=/usr/share/qemu-server/bootsplash.jpg
-vga
none
-nographic
-cpu
host,kvm=off,+kvm_pv_eoi,+kvm_pv_unhalt
-m
45056
-object
iothread,id=iothread-virtioscsi0
-object
{"id":"throttle-drive-scsi0","limits":{},"qom-type":"throttle-group"}
-global
ICH9-LPC.disable_s3=1
-global
ICH9-LPC.disable_s4=1
-readconfig
/usr/share/qemu-server/pve-q35-4.0.cfg
-device
vmgenid,guid=7bc76f71-7aed-4950-af2e-65bdc20e96b5
-device
qemu-xhci,p2=15,p3=15,id=xhci,bus=pci.1,addr=0x1b
-device
usb-tablet,id=tablet,bus=ehci.0,port=1
-device
vfio-pci,host=0000:04:00.0,id=hostpci0.0,bus=ich9-pcie-port-1,addr=0x0.0,multifunction=on
-device
vfio-pci,host=0000:05:00.0,id=hostpci0.1,bus=ich9-pcie-port-1,addr=0x0.1
-device
usb-host,bus=xhci.0,port=1,hostbus=3,hostport=1,id=usb0
-chardev
socket,path=/var/run/qemu-server/101.qga,server=on,wait=off,id=qga0
-device
virtio-serial,id=qga0,bus=pci.0,addr=0x8
-device
virtserialport,chardev=qga0,name=org.qemu.guest_agent.0
-iscsi
initiator-name=iqn.1993-08.org.debian:01:2dfae84330ee
-device
ide-cd,bus=ide.1,unit=0,id=ide2,bootindex=101
-device
virtio-scsi-pci,id=virtioscsi0,bus=pci.3,addr=0x1,iothread=iothread-virtioscsi0
-blockdev
{"detect-zeroes":"unmap","discard":"unmap","driver":"throttle","file":{"cache":{"direct":true,"no-flush":false},"detect-zeroes":"unmap","discard":"unmap","driver":"raw","file":{"aio":"native","cache":{"direct":true,"no-flush":false},"detect-zeroes":"unmap","discard":"unmap","driver":"host_device","filename":"/dev/zvol/data2/vm-101-disk-1","node-name":"ed295b9a759d5dc76d0d27e0dab7ab2","read-only":false},"node-name":"fd295b9a759d5dc76d0d27e0dab7ab2","read-only":false},"node-name":"drive-scsi0","read-only":false,"throttle-group":"throttle-drive-scsi0"}
-device
scsi-hd,bus=virtioscsi0.0,channel=0,scsi-id=0,lun=0,drive=drive-scsi0,id=scsi0,device_id=drive-scsi0,bootindex=100,write-cache=on
-netdev
type=tap,id=net0,ifname=tap101i0,script=/usr/libexec/qemu-server/pve-bridge,downscript=/usr/libexec/qemu-server/pve-bridgedown,vhost=on
-device
virtio-net-pci,mac=BC:24:11:C6:2E:38,netdev=net0,bus=pci.0,addr=0x12,id=net0,rx_queue_size=1024,tx_queue_size=256,bootindex=102,host_mtu=1500
-netdev
type=tap,id=net1,ifname=tap101i1,script=/usr/libexec/qemu-server/pve-bridge,downscript=/usr/libexec/qemu-server/pve-bridgedown,vhost=on,queues=6
-device
virtio-net-pci,mac=BC:24:11:DB:8C:DF,netdev=net1,bus=pci.0,addr=0x13,id=net1,vectors=14,mq=on,packed=on,rx_queue_size=1024,tx_queue_size=256,host_mtu=1500
-machine
pflash0=pflash0,pflash1=drive-efidisk0,hpet=off,type=q35+pve0

---

Additional Notes

• VM start shows these warnings:

  error writing '1' to '/sys/bus/pci/devices/0000:04:00.0/reset': Inappropriate ioctl for device
  failed to reset PCI device '0000:04:00.0', but trying to continue as not all devices need a reset
  

• GPU reset methods are currently disabled
• Host freezes randomly, typically within 24 hours of operation
• The freeze occurs both with VM running and sometimes even without VM running
• With 32GB BAR (ReBAR enabled), host freezes after 6-24 hours
• With 256MB BAR (current), host still freezes randomly, even when VM is not running

[ntsugar]

I have the same problem.