I am having an issue that has affected me since maybe February or March involving intermittent display corruption on Debian Testing/Forky it is now September, and I am on GNOME 50.4 and this issue is still not resolved and has possibly gotten worse.
It usually happens (inconsistently, maybe 1 in 5–10 times) when watching a video on my main monitor and something happens like going into the overview, app grid, receiving a notification, or just things changing on my second monitor, but sometimes I just randomly get a flash on my second monitor for seemingly no reason lately (possibly this: https://gitlab.gnome.org/GNOME/mutter/-/work_items/4698). The corruption is a brief flash (typically on my primary) of what was there before in a rainbow-y sort of manner where all the colors are shifted on one or both monitors. This doesn't get any worse or better when playing demanding games, and it does not show up in screen recordings. I just dealt with it for a while thinking GNOME 50 would fix it, but it hasn't and it's getting a bit distracting.
Specs:
# System Details Report
---
## Report details
- **Date generated:** 2026-09-06 00:10:23
## Hardware Information:
- **Hardware Model:** MSI MS-7845
- **Memory:** 32.0 GiB
- **Processor:** Intel® Core™ i5-4570 × 4
- **Graphics:** AMD Radeon™ RX 580 2048SP
- **Disk Capacity:** 1.3 TB
## Software Information:
- **Firmware Version:** V1.12
- **OS Name:** Debian GNU/Linux forky/sid
- **OS Build:** (null)
- **OS Type:** 64-bit
- **GNOME Shell Version:** 50.4
- **Kernel Version:** Linux 7.1.12+deb14-amd64
Mesa 26.1.6-1
Displays:
Primary - DELL 2407WFP: 1920x1200 at 59.95 Hz
Secondary - HP 2310: 1920x1080 at 60 Hz
Both are connected via Display Port to DVI as they are old monitors.
What I've tried over the past couple months:
Disabling GNOME Extensions and restarting - still present
Disabling overclocking and undervolting in LACT and restarting - still present
Setting GPU Clocks to highest or lowest - still present
amdgpu.aspm=0 amdgpu.runpm=0 in GRUB - still present
amdgpu.dcfeaturemask=0x2 in GRUB - still present
amdgpu.sg_display=0 in GRUB - still present
MUTTER_DEBUG_DISABLE_EXPL_SYNC=1 in environment - still present
Trying only one of both monitors - still present
Turning on reduce motion - still present
Booting a Debian 13 GNOME Live environment - not present but hesitant on downgrading. (and what if backported kernel reintroduces it?)
Changing display scaling to 125% and back to 100% - still present
Disabling night light - fixes flashes on second display but still present
MUTTER_DEBUG_DISABLE_DIRECT_SCANOUT=1 (in ~/.config/environment.d) - seemingly makes the corruption rarer? But still present on primary even combined with above
Booting an Ubuntu 26.04.1 Live Environment (with GNOME 50.1-0ubuntu1.2 and Linux 7.0) - still present
Booting a Kubuntu 26.04.1 Live Environment (With Plasma 6.6.6 and Linux 7.0) - Seemingly not present but unwilling to switch to Plasma.
Both MUTTER_DEBUG_DISABLE_DIRECT_SCANOUT=1 and MUTTER_DEBUG_FORCE_KMS_MODE=shadow in environment - still present
Booting a Fedora Workstation 44 Live Environment (With gnome-shell 50.4 and Linux 6.19.10) - present, especially with night light.
MUTTER_DEBUG_ENABLE_ATOMIC_KMS=0 (in ~/.config/environment.d) - still present (probably wrong spot to put it??)
I have seen nothing obvious (to me) in dmesg or journalctl about what could be going on.
Might be a kernel regression affecting Mutter, my specific GPU, or my strange monitor setup since I have seen almost no discussion related to my exact issue but I have no clue.
Do not suggest "just switch to Plasma or xyz distro/OS" it is **NOT** helpful.
Photo below is a video capture (because it lasts maybe a frame or two) recorded on my phone of how it looks on my primary monitor for a brief flash. This was from entering and exiting the Overview.
#Linux #Debian #GNOME #LinuxHelp #Troubleshooting