Nightly Builds

    Quote from palle_c

    Hey.

    thanks for checking :)

    I will test it when a new nvidia driver is implemented.


    Regards

    Hey HiassofT

    I tested the new LibreELEC-Generic-legacy.x86_64-13.0-nightly-20250428-17d2e4e.img just now on "Live" usb stick.

    I still do not work with the new nvidia driver :( itsstill "blank screen/no signal"

    Edit: forgot logs.

    https://paste.libreelec.tv/upward-sheep.log

    https://paste.libreelec.tv/eternal-labrador.log

    Regards

    Edited once, last by palle_c: logs (April 30, 2025 at 12:47 PM).

    Quote from chewitt
    Code
    kernel: nvidia-modeset: ERROR: GPU:0: Timed out waiting for core channel idle.

    No idea what that is .. Google doesn't show anything useful.

    ChatGPT ^^:

    1. Check Kernel Compatibility

    The 6.12 kernel may be ahead of NVIDIA’s QA validation curve. Try running with a more conservative, supported kernel version like 6.6.x LTS or 6.7.x to verify stability.

    2. Patch the NVIDIA Kernel Module for Compatibility

    If you must use 6.12, verify that the DKMS or manual installation succeeded without warnings. Run:

    Code
    dmesg | grep -Ei 'nvidia|gpu'
dkms status

    Check /var/lib/dkms/nvidia/570.144/build/make.log for errors. If there are symbol mismatches or KMS hook issues, NVIDIA may not yet support 6.12, and you may need to:

    • Use a community patch set (e.g., from NV-Kernel-Patcher if using open kernel modules)
    • Or manually adapt the build if the symbol changed (common with drm API updates).

    3. Test Disabling DRM KMS for Isolation

    Set in modprobe:

    Code
    echo "options nvidia-drm modeset=0" | sudo tee /etc/modprobe.d/nvidia.conf
sudo update-initramfs -u

    Reboot and see if the timeout disappears. This will narrow the issue to modeset interaction with the new kernel's DRM layer.

    4. Enable Early KMS Logging

    For deeper kernel debugging, append this to your GRUB:

    Code
    GRUB_CMDLINE_LINUX_DEFAULT="nvidia_drm.modeset=1 nvidia.NVreg_RegistryDwords=RmLogonRC=0xff"

    Then run:

    Code
    sudo update-grub && reboot

    Watch dmesg output for RmLogonRC flags — this can help pinpoint which channel stalls.

    5. Run with Open Kernel Modules (Optional)

    Since R515, NVIDIA supports open kernel modules:

    Code
    modinfo nvidia | grep filename

    ============================================================================================================
    ============================================================================================================
    ============================================================================================================

    and here some suggestions tailored to LE nighties (I'm wondering how much it all makes sense) :):

    1. Add NVIDIA DRM Modeset Boot Option

    LibreELEC lets you append kernel params via cmdline.txt on the /flash partition (accessible from SSH or directly from the SD card/USB):

    Code
    nvidia_drm.modeset=0

    💡 Purpose: Disables DRM modesetting which causes timeout in some nightly builds.

    Quote

    ✅ Highly recommended as first fix to isolate the nvidia-modeset stall.

    2. Force Legacy Rendering Path (X11 instead of GBM) — if available

    Some builds may allow fallback to X11 rendering using the x11 variant of the NVIDIA driver.

    From SSH, check if GBM is active:

    Code
    cat /var/log/Xorg.0.log | grep -i nvidia
cat /run/libreelec/debug/udev.log | grep -i gbm

    If GBM + KMS is the cause, switching to legacy X11 (if supported in this build) can reduce stalls — but not all LibreELEC nightlies retain this support.

    3. Check for Nouveau Conflicts

    Even though LibreELEC is minimal, you can verify nouveau isn’t mistakenly loaded:

    Code
    lsmod | grep nouveau

    If present, blacklist it permanently via /flash/syslinux.cfg or /flash/extlinux/extlinux.conf (depends on build):

    Code
    append boot=UUID=... quiet modprobe.blacklist=nouveau

    4. Set NVIDIA Power Management Policy

    Add to your cmdline.txt (or bootloader config):

    Code
    nvidia.NVreg_EnableGpuFirmware=0

    In some cases, firmware-based runtime power management causes timing problems on newer kernels.

    5. Enable Persistenced (if supported)

    LibreELEC often omits nvidia-persistenced, but if present:

    Code
    systemctl start nvidia-persistenced

    If not present, ignore — this is only effective in systems that use persistent GPU state management.

    Edited 2 times, last by adam.h.: Merged a post created by adam.h. into this post. (April 30, 2025 at 11:37 PM).