[Testing - Trixie] Upgrade to either testing or unstable from latest stable results in no video upon boot

[nanoshot]

Hi,

I'm attempting to install testing from a clean stable install and am running into an issue after rebooting the system. The main symptom is that there is no video output. At the end of the apt full-upgrade command, I do see several errors regarding possible missing firmware, and on follow up boot I get a message saying that the system cannot find the amdgpu/gc_11_0_0_mes_2.bin file. I have tried retrieving this file from the kernel git repository, and after updating my initramfs, that error goes away, but the system display still doesn't show up. I have no idea if this missing firmware is a red herring, or if my attempts to fix it were simply incomplete/incorrect.

I have also tried similar steps upgrading all the way to unstable, figuring maybe something hasn't been released down to testing yet, however I get the same result.

My GPU is a RX 7900 XTX.

I am willing to post whatever further system details needed, however, in order to get any useful console output, I need to be back on a fresh stable install.

[mrmazda]

I've noticed threads not so long ago that a missing firmware warning turned out to be true, and that fetching correct or fixed firmware file or kernel from a non-standard source was the only fix. I often see missing firmware messages on Bookworm and Trixie kernel installations, but it's never actually been missing for any of my GPUs. You may need to do a search for an available gc_11_0_0_mes_2.bin.

Input/output from inxi -GSaz --vs pasted here might be useful if you can find a way to run it from within an X session without booting using nomodeset or amdgpu.modeset=0, not so much if you must use a vt or can't employ kernel modesetting (KMS).

[nanoshot]

Like I said in my first post, I've already tried going and retrieving the missing .bin file from the linux kernel. but it didn't seem to fix the issue. I can try again if you have recommendations for specific steps, in case I did something wrong.
At the time I was on a repair boot mounted into the system and simply did a wget <kernel rawfile url> followed by a update-initramfs -u, perhaps those steps are incorrect?.

Because some information is better than nothing, this is the result of running inxi -GSaz --vs on a pre-upgrade stable install for the same hardware.

Code: Select all

root@loki:~# inxi -GSaz --vs
inxi 3.3.33-00 (2024-02-06)
System:
  Kernel: 6.1.0-18-amd64 arch: x86_64 bits: 64 compiler: gcc v: 12.2.0
    clocksource: tsc avail: hpet,acpi_pm
    parameters: BOOT_IMAGE=/@rootfs/boot/vmlinuz-6.1.0-18-amd64
    root=UUID=7a784327-1594-475a-a0ea-5ce6d232ec23 ro
    rootflags=subvol=@rootfs quiet
  Desktop: KDE Plasma v: 5.27.5 tk: Qt v: N/A wm: kwin_x11 dm: SDDM
    Distro: Debian GNU/Linux 12 (bookworm)
Graphics:
  Device-1: AMD Navi 31 [Radeon RX 7900 XT/7900 XTX] vendor: Micro-Star MSI
    driver: amdgpu v: kernel arch: RDNA-3 code: Navi-3x process: TSMC n5 (5nm)
    built: 2022+ pcie: gen: 4 speed: 16 GT/s lanes: 16 ports: active: HDMI-A-1
    empty: DP-1,DP-2,DP-3 bus-ID: 03:00.0 chip-ID: 1002:744c class-ID: 0300
  Device-2: AMD Raphael vendor: ASUSTeK driver: amdgpu v: kernel
    arch: RDNA-2 code: Navi-2x process: TSMC n7 (7nm) built: 2020-22 pcie:
    gen: 4 speed: 16 GT/s lanes: 16 ports: active: none empty: DP-4,HDMI-A-2
    bus-ID: 13:00.0 chip-ID: 1002:164e class-ID: 0300 temp: 32.0 C
  Display: x11 server: X.Org v: 1.21.1.7 with: Xwayland v: 22.1.9
    compositor: kwin_x11 driver: X: loaded: amdgpu
    unloaded: fbdev,modesetting,radeon,vesa dri: radeonsi gpu: amdgpu
    display-ID: :0 screens: 1
  Screen-1: 0 s-res: 3440x1440 s-dpi: 96 s-size: 910x381mm (35.83x15.00")
    s-diag: 987mm (38.84")
  Monitor-1: HDMI-A-1 mapped: HDMI-A-0 model: Acer XR341CK serial: <filter>
    built: 2016 res: 3440x1440 hz: 60 dpi: 107 gamma: 1.2
    size: 819x363mm (32.24x14.29") diag: 896mm (35.3") modes: max: 3440x1440
    min: 720x400
  API: EGL v: 1.5 hw: drv: amd radeonsi platforms: device: 0 drv: radeonsi
    device: 1 drv: radeonsi device: 2 drv: swrast gbm: drv: radeonsi
    surfaceless: drv: radeonsi x11: drv: radeonsi inactive: wayland
  API: OpenGL v: 4.6 compat-v: 4.5 vendor: amd mesa v: 22.3.6 glx-v: 1.4
    direct-render: yes renderer: AMD Radeon Graphics (gfx1100 LLVM 15.0.6 DRM
    3.49 6.1.0-18-amd64) device-ID: 1002:744c memory: 23.44 GiB unified: no

[strapicarus]

Hi, since last week, doing an install or update from the repository sid/testing, may break something in the system. I read in some post that sid/testing is going through a big time transition or transaction, plus xz's investigation will cause delays in sid repository being stabilized.

This is my opinion based on my experience last week, my gpu drivers reside in the sid repository and a upgrade just start to held packages back to protect the system, but i dint like it, so i start making apt hell --only-upgrade, --reinstall, purge, install, ect. finishing in a stable clean install.

[nanoshot]

I don't think it's related to the latest testing/unstable instability due to the time libraries. I have done some further testing on my own and installing just any backported kernel on debian stable also causes the issue. I have tried with 6.5 and 6.6, both of which cause display to fail on reboot. I booted into the old kernel to rollback to restore system functionality.

[strapicarus]

Well, i have done what you want todo a couple of times the pass 7 months;

1. Instal Debian stable.
2. edit sources.list to sid and update upgrade.
3. install linux-header firmwaremisc non free and gpu drivers.

All worked fine till last week, the kernels i know it worked was 6.5, 6.6 and 6.7.

Disclaimer: i haven't try the last 2 days. im not a expert

[mrmazda]

I did an upgrade of Noble last night with a 12 year old NVidia Fermi and got similar experience with the fresh 6.8.0-22 kernel - black screens with a random scattering of red and blue pixels on launch of DM that doesn't happen with prior kernel 6.8.0-11. Booting 22 to multi-user.target works normally. All my Trixies remain on 6.6.15, and I have no Sids. My newest AMD GPU is GCN #2 10 years old, working fine with Trixie's 6.6.15, but with TDE. KDE is not installed.

[nanoshot]

So as of today, I did a new upgrade to Unstable and now my displays work. I'm not entirely sure what was updated upstream between monday and today to fix the problem, but it does appear to be fixed. I still have an issue with my audio, but that persisted on stable as well. I'll be doing some research and opening a thread on that later.

[strapicarus]

i also have installed gpu drivers today, but couldn't install wine32, i use pulseaudio, maybe yours is pipewire?.

[DeBonzo]

I had a similar experience install testing on an Asus laptop with an Nvidia gpu. Turns out that the nouveau driver was crashing. Installing the propriety Nvidia driver fixed it. Calling journalctl --system might show you what happened.