Describe the bug
On Raspberry Pi OS Trixie with kernel 6.18.34+rpt-rpi-v8, the desktop eventually freezes while connected through NoMachine.
SSH continues to work, but the X desktop becomes unusable and NoMachine can no longer reconnect.
The kernel repeatedly reports V3D GPU resets:
v3d: Resetting GPU for hang
V3D_ERR_STAT: 0x00001000
Disabling glamor acceleration appears to eliminate the problem.
To reproduce
Hardware:
- Raspberry Pi 4 (4GB)
- HDMI to VGA adapator connected
- Raspberry Pi OS Trixie 64-bit
- LightDM
- X11 (not Wayland)
Software:
- Kernel 6.18.34+rpt-rpi-v8
- NoMachine 9.8.2
Logs
Kernel:
v3d: Resetting GPU for hang.
V3D_ERR_STAT: 0x00001000
No other GPU errors are present before the failure.
Additional observations
The X server becomes unresponsive after the GPU hang.
From SSH:
DISPLAY=:0 XAUTHORITY=/var/run/lightdm/root/:0 xdpyinfo
blocks once the failure occurs.
Workaround
Creating:
/etc/X11/xorg.conf.d/20-modesetting-noaccel.conf
with:
Section "Device"
Identifier "Modesetting"
Driver "modesetting"
Option "AccelMethod" "none"
EndSection
results in:
() modeset(0): Option "AccelMethod" "none"
() modeset(0): glamor disabled
After disabling glamor:
- NoMachine remains usable.
- rtl_airband and TwoToneDetect continue running.
- No V3D errors occur.
- The system remained stable for at least 47 minutes (still testing).
This strongly suggests the issue is in the accelerated X11/V3D path rather than NoMachine itself.
Steps to reproduce the behaviour
Steps:
- Boot normally.
- Connect to the desktop using NoMachine.
- Use the desktop normally (mostly terminal windows).
- After roughly 10-20 minutes the desktop freezes.
- SSH remains functional.
- Kernel repeatedly reports V3D GPU reset messages.
Device (s)
Raspberry Pi 4 Mod. B
System
https://pastebin.com/BVBwgayG
Logs
sudo dmesg -T | grep -iE 'v3d|drm_sched|gpu_sched|job timedout|hung task' | tail -80
[Thu Jul 23 20:06:34 2026] v3d fec00000.v3d: [drm] Transparent Hugepage support is recommended for optimal performance on this platform!
[Thu Jul 23 20:06:34 2026] [drm] Initialized v3d 1.0.0 for fec00000.v3d on minor 0
[Thu Jul 23 20:18:37 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR Resetting GPU for hang.
[Thu Jul 23 20:18:37 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR V3D_ERR_STAT: 0x00001000
[Thu Jul 23 20:18:38 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR Resetting GPU for hang.
[Thu Jul 23 20:18:38 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR V3D_ERR_STAT: 0x00001000
[Thu Jul 23 20:18:39 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR Resetting GPU for hang.
[Thu Jul 23 20:18:39 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR V3D_ERR_STAT: 0x00001000
[Thu Jul 23 20:18:40 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR Resetting GPU for hang.
[Thu Jul 23 20:18:40 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR V3D_ERR_STAT: 0x00001000
[Thu Jul 23 20:18:40 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR Resetting GPU for hang.
[Thu Jul 23 20:18:40 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR V3D_ERR_STAT: 0x00001000
[Thu Jul 23 20:18:41 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR Resetting GPU for hang.
[Thu Jul 23 20:18:41 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR V3D_ERR_STAT: 0x00001000
[Thu Jul 23 20:18:41 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR Resetting GPU for hang.
[Thu Jul 23 20:18:41 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR V3D_ERR_STAT: 0x00001000
[Thu Jul 23 20:18:42 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR Resetting GPU for hang.
[Thu Jul 23 20:18:42 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR V3D_ERR_STAT: 0x00001000
[Thu Jul 23 20:18:43 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR Resetting GPU for hang.
[Thu Jul 23 20:18:43 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR V3D_ERR_STAT: 0x00001000
[Thu Jul 23 20:18:44 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR Resetting GPU for hang.
[Thu Jul 23 20:18:44 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR V3D_ERR_STAT: 0x00001000
Additional context
No response
Describe the bug
On Raspberry Pi OS Trixie with kernel 6.18.34+rpt-rpi-v8, the desktop eventually freezes while connected through NoMachine.
SSH continues to work, but the X desktop becomes unusable and NoMachine can no longer reconnect.
The kernel repeatedly reports V3D GPU resets:
v3d: Resetting GPU for hang
V3D_ERR_STAT: 0x00001000
Disabling glamor acceleration appears to eliminate the problem.
To reproduce
Hardware:
Software:
Logs
Kernel:
v3d: Resetting GPU for hang.
V3D_ERR_STAT: 0x00001000
No other GPU errors are present before the failure.
Additional observations
The X server becomes unresponsive after the GPU hang.
From SSH:
DISPLAY=:0 XAUTHORITY=/var/run/lightdm/root/:0 xdpyinfo
blocks once the failure occurs.
Workaround
Creating:
/etc/X11/xorg.conf.d/20-modesetting-noaccel.conf
with:
Section "Device"
Identifier "Modesetting"
Driver "modesetting"
Option "AccelMethod" "none"
EndSection
results in:
() modeset(0): Option "AccelMethod" "none"
() modeset(0): glamor disabled
After disabling glamor:
This strongly suggests the issue is in the accelerated X11/V3D path rather than NoMachine itself.
Steps to reproduce the behaviour
Steps:
Device (s)
Raspberry Pi 4 Mod. B
System
https://pastebin.com/BVBwgayG
Logs
sudo dmesg -T | grep -iE 'v3d|drm_sched|gpu_sched|job timedout|hung task' | tail -80
[Thu Jul 23 20:06:34 2026] v3d fec00000.v3d: [drm] Transparent Hugepage support is recommended for optimal performance on this platform!
[Thu Jul 23 20:06:34 2026] [drm] Initialized v3d 1.0.0 for fec00000.v3d on minor 0
[Thu Jul 23 20:18:37 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR Resetting GPU for hang.
[Thu Jul 23 20:18:37 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR V3D_ERR_STAT: 0x00001000
[Thu Jul 23 20:18:38 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR Resetting GPU for hang.
[Thu Jul 23 20:18:38 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR V3D_ERR_STAT: 0x00001000
[Thu Jul 23 20:18:39 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR Resetting GPU for hang.
[Thu Jul 23 20:18:39 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR V3D_ERR_STAT: 0x00001000
[Thu Jul 23 20:18:40 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR Resetting GPU for hang.
[Thu Jul 23 20:18:40 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR V3D_ERR_STAT: 0x00001000
[Thu Jul 23 20:18:40 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR Resetting GPU for hang.
[Thu Jul 23 20:18:40 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR V3D_ERR_STAT: 0x00001000
[Thu Jul 23 20:18:41 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR Resetting GPU for hang.
[Thu Jul 23 20:18:41 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR V3D_ERR_STAT: 0x00001000
[Thu Jul 23 20:18:41 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR Resetting GPU for hang.
[Thu Jul 23 20:18:41 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR V3D_ERR_STAT: 0x00001000
[Thu Jul 23 20:18:42 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR Resetting GPU for hang.
[Thu Jul 23 20:18:42 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR V3D_ERR_STAT: 0x00001000
[Thu Jul 23 20:18:43 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR Resetting GPU for hang.
[Thu Jul 23 20:18:43 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR V3D_ERR_STAT: 0x00001000
[Thu Jul 23 20:18:44 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR Resetting GPU for hang.
[Thu Jul 23 20:18:44 2026] v3d fec00000.v3d: [drm:v3d_reset [v3d]] ERROR V3D_ERR_STAT: 0x00001000
Additional context
No response