https://bugs.kde.org/show_bug.cgi?id=526811
Frederico Raylander <[email protected]> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |raylanderproductions@gmail. | |com --- Comment #7 from Frederico Raylander <[email protected]> --- I can confirm very similar behavior on Fedora 44 KDE Plasma Wayland, using Intel Meteor Lake graphics with the i915 driver. **Hardware and environment:** - Laptop: Samsung Galaxy Book4 Pro - GPU: Intel Meteor Lake-P [Intel Arc Graphics] (rev 08) - Kernel driver: i915 - Kernel: 7.2.10-200.fc44.x86_64 - OS: Fedora 44 KDE, Wayland - External monitor: 1920x1080 @ 74.97 Hz, connected via HDMI - Internal display: disabled while using the external monitor - Internal panel: 2880x1800 @ 120 Hz **Observed behavior:** With the default KWin configuration, mouse cursor movement on the external monitor exhibits noticeable micro-stuttering. The issue occurs with both an external mouse and the laptop's touchpad. The internal 120 Hz display feels smooth when used alone. **Troubleshooting performed:** 1. `KWIN_USE_OVERLAYS=0`: no noticeable improvement. 2. `KWIN_FORCE_SW_CURSOR=1`: cursor stuttering became slightly worse. 3. Switching the external monitor from 75 Hz to 60 Hz: stuttering remained noticeable, with the additional reduction in smoothness expected at 60 Hz. 4. `KWIN_DRM_NO_AMS=1`: dramatic improvement. The stuttering appears to be completely eliminated. I verified that `KWIN_DRM_NO_AMS=1` was present in the running `kwin_wayland` process environment. With legacy modesetting enabled, the remaining difference in cursor smoothness appears consistent with the normal refresh-rate difference between the external 75 Hz monitor and the internal 120 Hz display. **Conclusion:** This appears to reproduce a closely related issue on Fedora 44 using a 1080p 75 Hz HDMI monitor, suggesting that a 4K display is not necessary to experience this behavior. I have not measured frame pacing or FPS, so my observations are based on perceived cursor smoothness rather than quantitative measurements. Thank you for investigating this. The `KWIN_DRM_NO_AMS=1` workaround made a substantial difference in my daily desktop experience. -- You are receiving this mail because: You are watching all bug changes.
