Second GPU doesn't work after big update

I am having issues with the Wynsdey update that I haven’t seen posted elsewhere.

I posted about it here but I think it wasn’t the right place.

  • I have six monitors split across two GPUs. Four plugged in to AMD RV730 GL FirePro V3750 and two plugged in to AMD Cape Verde XT [Radeon HD 7770/8760 / R7 250X]
  • One of my two GPUs (I think it is AMD RV730 GL FirePro V3750) no longer drives video output even though it is detected by the OS.
  • If I drop to a tty, then the monitors attached to the non-working GPU start working, and the other monitors attached to the other GPU freeze
  • I’m also getting corrupted display issues if the computer has been booted for a while. Corrupted cursor square:

All monitors and gpus were working perfectly fine before this last major update.

`$ inxi -bz`
System:
  Kernel: 6.6.32-1-MANJARO arch: x86_64 bits: 64
  Desktop: GNOME v: 46.2 Distro: Manjaro Linux
Machine:
  Type: Desktop Mobo: ASUSTeK model: TUF GAMING X570-PLUS (WI-FI) v: Rev X.0x
    serial: <superuser required> UEFI: American Megatrends v: 4403
    date: 04/27/2022
CPU:
  Info: 12-core AMD Ryzen 9 5900X [MT MCP] speed (MHz): avg: 2565
    min/max: 2200/4950
Graphics:
  Device-1: AMD RV730 GL [FirePro V3750] driver: radeon v: kernel
  Device-2: AMD Cape Verde XT [Radeon HD 7770/8760 / R7 250X] driver: radeon
    v: kernel
  Device-3: Logitech Webcam C930e driver: snd-usb-audio,uvcvideo type: USB
  Display: wayland server: X.Org v: 24.1 with: Xwayland v: 24.1.0
    compositor: gnome-shell driver: X: loaded: radeon unloaded: modesetting
    dri: radeonsi,r600 gpu: radeon resolution: 1: 1920x1080~60Hz
    2: 1920x1080~60Hz
  API: EGL Message: EGL data requires eglinfo. Check --recommends.
Network:
  Device-1: Intel Wi-Fi 5 Wireless-AC 9x6x [Thunder Peak] driver: iwlwifi
  Device-2: Realtek RTL8111/8168/8211/8411 PCI Express Gigabit Ethernet
    driver: r8169
Drives:
  Local Storage: total: 51.87 TiB used: 10.23 TiB (19.7%)
Info:
  Memory: total: 64 GiB note: est. available: 62.7 GiB used: 8.43 GiB (13.4%)
  Processes: 606 Uptime: 2h 24m Shell: Zsh inxi: 3.3.34
sudo mhwd -l
> 0000:0b:00.0 (0300:1002:683d) Display controller ATI Technologies Inc:
--------------------------------------------------------------------------------
                  NAME               VERSION          FREEDRIVER           TYPE
--------------------------------------------------------------------------------
           video-linux            2024.05.06                true            PCI
     video-modesetting            2020.01.13                true            PCI
            video-vesa            2017.03.12                true            PCI


> 0000:07:00.0 (0200:10ec:8168) Network controller Realtek Semiconductor Co., Ltd.:
--------------------------------------------------------------------------------
                  NAME               VERSION          FREEDRIVER           TYPE
--------------------------------------------------------------------------------
         network-r8168            2023.09.12                true            PCI


> 0000:05:00.0 (0300:1002:949f) Display controller ATI Technologies Inc:
--------------------------------------------------------------------------------
                  NAME               VERSION          FREEDRIVER           TYPE
--------------------------------------------------------------------------------
           video-linux            2024.05.06                true            PCI
     video-modesetting            2020.01.13                true            PCI
            video-vesa            2017.03.12                true            PCI

journalctl:

gnome-shell:6746: GLib-CRICITAL **: g_close(fd:0) failed with EBADF. The tracking of file descriptors got messed up
grome-shell:6746: libmutterWARNING **: Failed to open gpu '/dev/dri/card0': No suitable mode setting backend found
lspci -k | grep -EA3 'VGA|3D|Display'
05:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] RV730 GL [FirePro V3750]
	Subsystem: Advanced Micro Devices, Inc. [AMD/ATI] Device 2542
	Kernel driver in use: radeon
	Kernel modules: radeon, amdgpu
--
0b:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Cape Verde XT [Radeon HD 7770/8760 / R7 250X]
	Subsystem: PC Partner Limited / Sapphire Technology Device e214
	Kernel driver in use: radeon
	Kernel modules: radeon, amdgpu

If I append amdgpu.modeset=1 radeon.modeset=0 to the linux line on my grub settings on boot, then the other gpu kinda works in that 2/4 monitors have display, but it is duplicate and at 1024x768 resolution. I cannot extend between the two displays. The other two displays on that gpu are black and the remaining two displays on the other gpu are also black.

/dev/dri/card1 works perfectly fine, that is the GPU that is working. Why is the other one failing? I appreciate any help!

Update: I can login to a different user account that uses x11 and everything works. This is wayland related?

Or there is something different in the two user configurations. Create another user and test Wayland there, to see if it’s user configuration related or not.

Thank you for your reply. I created a new user and logged in with and without Wayland.

  • Wayland: Broken
  • X11: All GPUs/monitors work correctly

I don’t know what else to suggest. Apart from use X11 if that works.

You have such an unusual system (your RV730 is 15 years old and is only supported by the radeon driver) that I suspect it’s going to be difficult for anyone to replicate or diagnose your issue.

Thank you anyways. I’m happy to have stumbled upon the X11 solution. Not ideal, but at least I can have a working system while I either wait for an update that fixes this or continue troubleshooting in my free time.

I haven’t noticed any major issues with going back to X11 other than mpv dropping a lot of frames, but VLC seems to work fine.

And yes, it’s a really old card, I don’t play video games so driving multiple monitors is really all I need, thankfully.