Home / ubuntu / Questions / 1051446
In Process
rich
Asked: 2018-07-03 02:44:51 +0800 CST

Hades Canyon NUC8i7HNK can install but can't boot

  • 772

I've installed 18.04 and updated the BIOS to 44.
I've tried adding nomodeset to the end of the linux line.
I've tried kernels 4.17.3 and 4.18-rc3 too. I've tried running off each individual stick of RAM and both together.

It freezes with just the text:

Loading Linux 4.15.0-23-generic
Loading initial ramdisk

Annoyingly, the Live CD just runs fine with nomodeset, but the installed version apparently doesn't, although I can chroot it from the live CD, fetch updates, etc fine.

Any suggestions?

grub2 boot intel-nuc 18.04

4 Answers

  1. I was having exactly the same problem with a NUC8i7hnk, but in my case I was trying to install archlinux. I tried everything possible to solve it with no success.

    So, I just gave up grub and tried using systemd-boot as the bootloader. Everything worked in the first try :). I think there is a bug when using GRUB with Intel NUCs

    • 1

  2. I was previously able to install Ubuntu 18.04 and get the GPU working fully by following these instructions: Graphics drivers for Intel NUC Hades Canyon NUC8i7HVK (AMD Radeon RX Vega GH)

    Did you make sure to copy over and install the GPU firmware files?

    However, that only worked for me in legacy boot mode, which was discontinued after BIOS 0033. In UEFI boot mode I can install 18.04 server/desktop and boot to a stock state (GPU not working, video is coming from LLVM-pipe), but if I install a newer kernel it will not boot and hang at Loading initial ramdisk, it will still boot to 4.15 though.

    EDIT (July 12 2018):

    I tried a few things differently this time so I'm not sure what fixed it, but here are the steps that worked for me on the HVK today:

    1. Disabled Intel GPU in BIOS
    2. Installed Ubuntu 18.04 Desktop using nomodeset as a grub option during install and regular boot
    3. Updated all packages
    4. Installed latest stable 18.1.3-1 mesa Ubuntu release from this PPA: https://launchpad.net/~ubuntu-x-swat/+archive/ubuntu/updates
    5. Installed daily Ubuntu kernel release (2018-07-09): http://kernel.ubuntu.com/~kernel-ppa/mainline/daily/2018-07-09/
    6. Rebooted (still using nomodeset)
    7. Installed AMDGPU Vega M firmware from: https://people.freedesktop.org/~agd5f/radeon_ucode/vegam/
    8. Removed nomodeset

    EDIT 2 (Jul 17 2018):

    I tried the same set of steps on a NUC8i7HNK and the machine still hangs at "loading initial ramdisk" after installing the 4.18 kernel. So HVK is fine, HNK has issues, at least with Ubuntu 18.04 + v4.18 kernel. It gets stuck prior to even installing the firmware files.

    • 0

  3. Once you set nomodeset in /etc/default/grub did you make sure to run an update?

    sudo update-grub

    I was able to boot into the new kernel and similarly run video from LLVM-pipe, but I can't get the GPU working (when I remove nomodeset). When I SSH in and run

    systemctl status gpu-manager.service

    the log yields the following:

    gpu-manager[1670]: /etc/modprobe.d is not a file
gpu-manager[1670]: /etc/modprobe.d is not a file
gpu-manager[1670]: /etc/modprobe.d is not a file
gpu-manager[1670]: /etc/modprobe.d is not a file
gpu-manager[1670]: Error: can't open /lib/modules/4.15.0-24-generic/updates/dkms
systemd[1]: Started Detect the available GPUs and deal with any system changes.

    Even switching to lightdm didn't help.

    Seems to be an issue in gdm3 with starting the xserver. Running the command,

    vdpauinfo

    shows:

    vdpauinfo: cannot connect to X server
    • 0

  4. Same problem, or similar I think. Brand new Intel Hades Canyon NUC8i7HNK. Clean install Ubuntu 18.04 Installer seemed to work perfectly. Hung during the install. 

    [FAILED] Failed to start User Manager for UID 121.
....repeating

    Went into BIOS, disabled Intel iGD

    Seems to have worked.

    • 0