PVE Host hard crashing when booting/running Windows 11 VM [No GPU Passthrough]

F

fgohan

New Member
Apr 11, 2025
15
0
1
Hello !


I've waited a month before posting this thread because I wouldn't bother people with an already solved subject. So I made my research all over the web and the forum. I've got helped by a Sysadmin and a Software engineer (I am too). We aren't able to solve this issue and understand what is really happening...

I worked for months with my setup just making VM and CT. Everything was working properly until I tried to install a Windows 11 VM. I've started using the GPU passthrough method because I am having the hardware to work with. I encountered regularly host crashs during Win11 installation process but with some hard reboots I was able to complete the installation.
Now when the VM in turned on it takes a random timing before the host is crashing, it can be instantly,15 minutes or even once it took 5 hours.
So I’ve tried to found some logs related using journal -xe or dmesg but nothing relevant about was found.


I did install proxmox in lvm or zfs i am still having the same issue. I changed the pve-kernel a lot of times it didn't change anything.

For a while now I am just trying to install a Windows 11 VM without any tweak like GPU Passthrough.

Just for information I tried to install a windows vm on Vmware ESXI 8 with GPU Passthrough and i didn't encountered this issue. The only issue I had was when installing Vmware with a PSOD Cpu mismatch that I was able to fix by this source : https://www.virten.net/2022/11/esxi-7-and-8-installation-fails-with-fatal-cpu-mismatch-on-feature/


I did saw some weird logs already in proxmox happening with or without windows VM like :

1.
Code: 
x86/split lock detection: #AC: crashing the kernel on kernel split_locks and warning on user-space split_locks
- > was fixed by adding split_lock_detect=off


2.
Code: 
[ 3.756780] intel-ipu6 0000:00:05.0: Direct firmware load for intel/ipu/ipu6epmtl_fw.bin failed with error -2

[ 3.757506] intel-ipu6 0000:00:05.0: error -ENOENT: Requesting signed firmware intel/ipu/ipu6epmtl_fw.bin failed

[ 3.758071] intel-ipu6 0000:00:05.0: probe with driver intel-ipu6 failed with error -2

- > was fixed by copying the ipu6 of the latest linux-firmware as below
Code: 
cp linux-firmware/intel/ipu6epmtl_fw.bin /lib/firmware/intel/



3.
Code: 
[    5.008899] CPU: 6 UID: 0 PID: 870 Comm: (udev-worker) Tainted: P           O       6.14.0-2-pve #1
[    5.008902] Tainted: [P]=PROPRIETARY_MODULE, [O]=OOT_MODULE
[    5.008904] Hardware name: AZW GTi14/GTi14, BIOS GTi14T108 01/13/2025
[    5.008905] RIP: 0010:intel_bios_init+0xeb1/0x19d0 [i915]
[    5.009002] Code: 24 48 8b 78 08 4c 8b 67 50 4d 85 e4 75 03 4c 8b 27 e8 03 f8 1a e0 89 d9 4c 89 e2 48 c7 c7 e0 74 d3 c1 48 89 c6 e8 9f 02 74 df <0f> 0b e9 4b fd ff ff 4c 8b b3 d8 12 00 00 4d 39 f7 0f 84 ca 01 00
[    5.009004] RSP: 0018:ffffa714c0ad7520 EFLAGS: 00010246
[    5.009006] RAX: 0000000000000000 RBX: 0000000000000046 RCX: 0000000000000000
[    5.009007] RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000000
[    5.009008] RBP: ffffa714c0ad7598 R08: 0000000000000000 R09: 0000000000000000
[    5.009010] R10: 0000000000000000 R11: 0000000000000000 R12: ffff907083294270
[    5.009011] R13: 0000000000000005 R14: 0000000000000001 R15: ffff9070818c9420
[    5.009012] FS:  0000702a1dab88c0(0000) GS:ffff90879fb00000(0000) knlGS:0000000000000000
[    5.009013] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[    5.009015] CR2: 0000702a1cdd3d39 CR3: 0000000101fb6005 CR4: 0000000000f70ef0
[    5.009016] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[    5.009017] DR3: 0000000000000000 DR6: 00000000ffff07f0 DR7: 0000000000000400
[    5.009018] PKRU: 55555554
[    5.009019] Call Trace:
[    5.009021]  <TASK>
[    5.009022]  ? show_regs+0x6c/0x80
[    5.009027]  ? __warn+0x8d/0x150
[    5.009030]  ? intel_bios_init+0xeb1/0x19d0 [i915]
[    5.009125]  ? report_bug+0x182/0x1b0
[    5.009130]  ? handle_bug+0x6e/0xb0
[    5.009133]  ? exc_invalid_op+0x18/0x80
[    5.009135]  ? asm_exc_invalid_op+0x1b/0x20
[    5.009139]  ? intel_bios_init+0xeb1/0x19d0 [i915]
[    5.009230]  intel_display_driver_probe_noirq+0x40/0x260 [i915]
[    5.009337]  i915_driver_probe+0x685/0xc80 [i915]
[    5.009417]  ? mutex_lock+0x12/0x50
[    5.009419]  ? drm_privacy_screen_get+0x16d/0x190
[    5.009423]  ? acpi_dev_found+0x64/0x80
[    5.009427]  i915_pci_probe+0xc1/0x150 [i915]
[    5.009502]  local_pci_probe+0x44/0xb0
[    5.009506]  pci_device_probe+0xf4/0x250
[    5.009508]  really_probe+0xee/0x3b0
[    5.009512]  __driver_probe_device+0x8c/0x180
[    5.009514]  driver_probe_device+0x24/0xd0
[    5.009515]  __driver_attach+0x10b/0x210
[    5.009517]  ? __pfx___driver_attach+0x10/0x10
[    5.009519]  bus_for_each_dev+0x8a/0xf0
[    5.009521]  driver_attach+0x1e/0x30
[    5.009522]  bus_add_driver+0x14e/0x290
[    5.009524]  driver_register+0x5e/0x130
[    5.009526]  __pci_register_driver+0x5e/0x70
[    5.009528]  i915_pci_register_driver+0x23/0x30 [i915]
[    5.009602]  i915_init+0x23/0xc0 [i915]
[    5.009682]  ? __pfx_i915_init+0x10/0x10 [i915]
[    5.009762]  do_one_initcall+0x5b/0x340
[    5.009766]  do_init_module+0x97/0x2c0
[    5.009770]  load_module+0x2321/0x24e0
[    5.009773]  init_module_from_file+0x96/0x100
[    5.009775]  ? init_module_from_file+0x96/0x100
[    5.009778]  idempotent_init_module+0x110/0x300
[    5.009781]  __x64_sys_finit_module+0x6f/0xe0
[    5.009784]  x64_sys_call+0x1ec6/0x2540
[    5.009786]  do_syscall_64+0x7e/0x170
[    5.009789]  ? mmap_region+0x66/0xe0
[    5.009792]  ? vm_mmap_pgoff+0x159/0x200
[    5.009795]  ? ksys_mmap_pgoff+0x186/0x220
[    5.009798]  ? arch_exit_to_user_mode_prepare.constprop.0+0x22/0xd0
[    5.009800]  ? syscall_exit_to_user_mode+0x38/0x1d0
[    5.009802]  ? do_syscall_64+0x8a/0x170
[    5.009804]  ? do_syscall_64+0x8a/0x170
[    5.009805]  ? arch_exit_to_user_mode_prepare.constprop.0+0x22/0xd0
[    5.009807]  ? syscall_exit_to_user_mode+0x38/0x1d0
[    5.009809]  ? do_syscall_64+0x8a/0x170
[    5.009811]  ? arch_exit_to_user_mode_prepare.constprop.0+0xc8/0xd0
[    5.009812]  ? syscall_exit_to_user_mode+0x38/0x1d0
[    5.009814]  ? do_syscall_64+0x8a/0x170
[    5.009816]  ? irqentry_exit+0x43/0x50
[    5.009817]  ? sysvec_call_function+0x57/0xc0
[    5.009819]  entry_SYSCALL_64_after_hwframe+0x76/0x7e
[    5.009821] RIP: 0033:0x702a1e1c97d9
[    5.009823] Code: 08 89 e8 5b 5d c3 66 2e 0f 1f 84 00 00 00 00 00 90 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d f7 05 0d 00 f7 d8 64 89 01 48
[    5.009826] RSP: 002b:00007fffea5bc328 EFLAGS: 00000246 ORIG_RAX: 0000000000000139
[    5.009828] RAX: ffffffffffffffda RBX: 0000594d87c25e00 RCX: 0000702a1e1c97d9
[    5.009830] RDX: 0000000000000000 RSI: 0000702a1e35cefd RDI: 0000000000000013
[    5.009831] RBP: 0000702a1e35cefd R08: 0000000000000000 R09: 0000594d87bb9d20
[    5.009832] R10: 0000000000000013 R11: 0000000000000246 R12: 0000000000020000
[    5.009834] R13: 0000000000000000 R14: 0000594d87c079b0 R15: 00007fffea5bc560
[    5.009836]  </TASK>
[    5.009837] ---[ end trace 0000000000000000 ]---
- > not fixed yet

My friends and I are totally stuck right now it is why we're asking for help. I am pretty sure that more people aware of proxmox and system engineering will be more competent than us.

Please help me.

Best regards


--------------

Config :

Beelink GTI14 ultra 185h / 96 gb ram / 2xnvme 2tb
 

Attachments

Last edited:
Already did the latest Beelink Bios update.
I tried with disabling Hyperthreading or disabling Efficient cores with no success.
Just tried few hours earlier to install Incus-ui with LXD. I could not get a windows VM working. It was crashing too but only the VM this time without being able to restart the VM after the crash, it is like the VM is not working forever. No relevant logs.

I am going to try to slowing the memory as you suggested...

It is frustrating because I can make a real install of win11 No VM on a drive or using ESXI (but i would try again vmware to see if it wasn't a luck).
Having any linux VM without any crashs.
 
You must log in or register to reply here.
Top Bottom