Out of the blue my Alder Lake (N100) iGPU stopped working in my Debian Trixie Docker VM, and I cannot figure out why.
The error I am getting in the vm (dmesg | grep -iE 'i915|drm|vga')
[ 0.000000] Command line: BOOT_IMAGE=/boot/vmlinuz-6.12.73+deb13-amd64 root=PARTUUID=774c99d7-c654-41d1-84da-c66ee52b3e49 ro console=tty0 console=ttyS0,115200 earlyprintk=ttyS0,115200 consoleblank=0 intel_iommu=on i915.enable_guc=3 module_blacklist=xe
[ 0.143303] Kernel command line: BOOT_IMAGE=/boot/vmlinuz-6.12.73+deb13-amd64 root=PARTUUID=774c99d7-c654-41d1-84da-c66ee52b3e49 ro console=tty0 console=ttyS0,115200 earlyprintk=ttyS0,115200 consoleblank=0 intel_iommu=on i915.enable_guc=3 module_blacklist=xe
[ 1.277264] pci 0000:06:10.0: vgaarb: setting as boot VGA device
[ 1.278875] pci 0000:06:10.0: vgaarb: bridge control possible
[ 1.280381] pci 0000:06:10.0: vgaarb: VGA device added: decodes=io+mem,owns=none,locks=none
[ 1.281201] vgaarb: loaded
[ 3.824465] systemd[1]: Starting modprobe@drm.service - Load Kernel Module drm...
[ 3.910892] ACPI: bus type drm_connector registered
[ 4.035935] systemd[1]: modprobe@drm.service: Deactivated successfully.
[ 4.038329] systemd[1]: Finished modprobe@drm.service - Load Kernel Module drm.
[ 5.263744] i915 0000:06:10.0: [drm] Found ALDERLAKE_P/ADL-N (device ID 46d1) display version 13.00 stepping D0
[ 7.270767] i915 0000:06:10.0: [drm] *ERROR* Device is non-operational; MMIO access returns 0xFFFFFFFF!
[ 7.273633] i915 0000:06:10.0: [drm] *ERROR* Device initialization failed (-5)
[ 7.275159] i915 0000:06:10.0: probe with driver i915 failed with error -5
Proxmox host: 6.17.4-2-pve kernel / v25.10.10 SR-IOV. Other VMs like my Windows 11 VM are working. In the logs I can see the usual VFx FLR reset entries - no errors, nothing out of the ordinary on this end.
VM side: 6.12.73+deb13-amd64 / v25.10.10 SR-IOV. Settings: q35 machine (latest, vIOMMU on "none"), passthrough: Raw device 0000:00:02.2, Primary GPU on, All functions off, ROM-Bar on, everything else off.
Yesterday it was working, today it isn't. There were some recent Proxmox updates, including a new kernel, but as you can see, I have an older one pinned still.
Any ideas?
EDIT:
Seems like an issue with the newest main Trixie kernel: 6.12.73+deb13-amd64
i915 works when reverting to the previous kernel: 6.12.69+deb13-amd64
Wonder if anyone else got the same and how they fixed it with the newest kernel. Seen some issues here with the recent SR-IOV releases but they were with regards to newer kernels, don't think it's related....
Out of the blue my Alder Lake (N100) iGPU stopped working in my Debian Trixie Docker VM, and I cannot figure out why.
The error I am getting in the vm (dmesg | grep -iE 'i915|drm|vga')
Proxmox host: 6.17.4-2-pve kernel / v25.10.10 SR-IOV. Other VMs like my Windows 11 VM are working. In the logs I can see the usual VFx FLR reset entries - no errors, nothing out of the ordinary on this end.
VM side: 6.12.73+deb13-amd64 / v25.10.10 SR-IOV. Settings: q35 machine (latest, vIOMMU on "none"), passthrough: Raw device 0000:00:02.2, Primary GPU on, All functions off, ROM-Bar on, everything else off.
Yesterday it was working, today it isn't. There were some recent Proxmox updates, including a new kernel, but as you can see, I have an older one pinned still.
Any ideas?
EDIT:
Seems like an issue with the newest main Trixie kernel: 6.12.73+deb13-amd64
i915 works when reverting to the previous kernel: 6.12.69+deb13-amd64
Wonder if anyone else got the same and how they fixed it with the newest kernel. Seen some issues here with the recent SR-IOV releases but they were with regards to newer kernels, don't think it's related....