[Bug] DIFR issue with 610 on Asus laptop

Here are the Bug reports

~$ sudo nvidia-bug-report.sh

nvidia-bug-report.sh will now collect information about your
system and create the file ‘nvidia-bug-report.log.gz’ in the current
directory. It may take several seconds to run. In some
cases, it may hang trying to capture data generated dynamically
by the Linux kernel and/or the NVIDIA kernel module. While
the bug report log file will be incomplete if this happens, it
may still contain enough data to diagnose your problem.

If nvidia-bug-report.sh hangs, consider running with the --safe-mode
and --extra-system-data command line arguments.

Please include the ‘nvidia-bug-report.log.gz’ log file when reporting
your bug via the NVIDIA Linux forum (see forums.developer.nvidia.com)
or by sending email to 'linux-bugs@nvidia.com’.

By delivering ‘nvidia-bug-report.log.gz’ to NVIDIA, you acknowledge
and agree that personal information may inadvertently be included in
the output. Notwithstanding the foregoing, NVIDIA will use the
output only for the purpose of investigating your reported issue.

Running nvidia-bug-report.sh…
Detected driver type: Running RM Driver on GPUs
complete.

Summary of Skipped Sections:

Skipped Component | Details

vulkaninfo output | vulkaninfo not found

ibstat output | ibstat not found

acpidump output | acpidump not found

mst output | mst not found

nvlsm-bug-report.sh output | nvlsm-bug-report.sh not found

Summary of Errors:

Error Component | Details | Resolution

Bug report — Reproducible kernel hang in nvidia-modeset (DIFR) on RTX 5060 Max-Q (GB206M)

Configuration

  • GPU: NVIDIA GeForce RTX 5060 Max-Q / Mobile (GB206M), Blackwell architecture
  • Machine: ASUS Vivobook (chassis V3607VM), Intel/NVIDIA hybrid graphics (Optimus), PCIe 0000:01:00.0
  • OS: Ubuntu 26.04
  • Kernel: 7.0.0-31-generic
  • Driver tested: nvidia-driver-595-open then nvidia-driver-610 (open kernel module, Dual MIT/GPL license, confirmed required for Blackwell) — identical behavior on both versions
  • Firmware: linux-firmware-nvidia-graphics up to date (latest version available as of 2026-09-14)
  • PRIME mode: on-demand (the intel-only mode disables the external HDMI output, which is wired only to the discrete GPU with no apparent hardware MUX)
  • Session: GNOME on Wayland (no Xorg session available in this Ubuntu version’s repositories)

Symptom

Complete, unrecoverable system hang (hard freeze, requiring a forced 10-second power-off), reproducible several times per week, occurring both after a period of inactivity (automatic screen blanking) and during active use. The hang consistently results in root filesystem corruption (ext4 on NVMe) on the next boot, requiring an fsck repair.

Kernel trace (reproduced identically 3 times over one week)

INFO: task KMS thread: blocked for more than 122 seconds.
Tainted: G O 7.0.0-31-generic
task KMS thread blocked on a semaphore likely last held by task nvidia-modeset

nvkms_ioctl_from_kapi_try_pmlock+0x55/0xc0 [nvidia_modeset]
ApplyModeSetConfig+0x382/0xc70 [nvidia_modeset]
? nv_drm_atomic_apply_modeset_config+0x79d/0xb40 [nvidia_drm]
? drm_mode_to_nvkms_display_mode+0x17/0x180 [nvidia_drm]
? nv_drm_crtc_atomic_check+0x4a6/0x960 [nvidia_drm]
? nv_drm_atomic_check+0x12d/0x150 [nvidia_drm]
? drm_atomic_check_only+0x199/0x420
? drm_atomic_commit+0x6c/0xf0
? drm_mode_atomic_ioctl+0x7d1/0x910
? drm_ioctl_kernel+0xb5/0x110
? drm_ioctl+0x309/0x5f0
? nv_drm_ioctl+0x38/0x420 [nvidia_drm]

task:nvidia-modeset/ state:R running task pid:976 tgid:976 ppid:2
RIP: 0010:nvWriteGpEntry+0xa4/0x370 [nvidia_modeset]
? nvPushKickoff+0x28/0x50 [nvidia_modeset]
? PrefetchHelperSurfaceEvo+0x45c/0x650 [nvidia_modeset]
? nvDIFRPrefetchSurfaces+0x135/0x1f0 [nvidia_modeset]
? DifrPrefetchEventDeferredWork+0x16/0x30 [nvidia_modeset]
? nvkms_kthread_q_callback+0xe9/0x190 [nvidia_modeset]
? _main_loop+0x99/0x160 [nvidia_modeset]

Immediately before each occurrence, rapid, repeated cycles of the following are observed:

nvidia 0000:01:00.0: Enabling HDA controller

(several occurrences within a few minutes), suggesting repeated GPU power-state transitions (runtime PM) right before the hang.

What has been tried with no effect on the bug

  1. Switching driver versions (595-open → 610) — no change.
  2. Updating the GPU firmware (linux-firmware-nvidia-graphics) — no change.
  3. Switching PRIME mode (on-demand) — no change (the intel-only mode avoids the crash but breaks HDMI output, so it cannot be tested over an extended period in real use).
  4. NVreg_EnableS0ixPowerManagement=1 + NVreg_S0ixPowerManagementVideoMemoryThreshold=256 — no change; the hang occurs even without a full system suspend, just from screen blanking.
  5. MUTTER_DEBUG_DISABLE_HW_CURSORS=1 (disabling the hardware cursor on the compositor side) — no change.
  6. NVreg_RegistryDwords="EnableDIFR=0" (confirmed loaded via cat /proc/driver/nvidia/params, shown as RegistryDwords: "EnableDIFR=0") — the hang reoccurred with an identical stack trace (still inside nvDIFRPrefetchSurfaces/DifrPrefetchEventDeferredWork), suggesting this registry key is either no longer the correct mechanism to fully disable DIFR on this driver version, or is insufficient on its own.

Question for the NVIDIA team

  • Is EnableDIFR=0 still the correct parameter to fully disable DIFR on driver 610 (or 595-open)? Is there another parameter or method to disable it completely?
  • Is this hang in nvWriteGpEntry / nvPushKickoff, reached via the DIFR prefetch path, a known issue on Blackwell mobile GPUs (GB206M) with the open kernel modules?

The nvidia-bug-report.log.gz file (generated via sudo nvidia-bug-report.sh right after a hang) is available on request and will be attached to the report.