https://bugs.kde.org/show_bug.cgi?id=517696
--- Comment #8 from Itsame <[email protected]> --- Hi, closing this as I've dug into this more today. The forced connector probes without DRM master turned out to be harmless, the kernel demotes them to read-only and they don't touch hardware. The actual signal loss is caused by the display manager tearing down the CRTC entirely between compositor transitions, leaving the monitor with no signal during the gap. QD-OLED panels are slow to re-lock, which makes the issue very visible. To future people stumbling upon this: manually matching monitor settings (Dynamic range, resolution etc) between the greeter and session didn't remove but greatly reduced the delay. I'm opening a bug against plasma-login-manager instead, since the issue is in how the display manager handles the compositor handoff. Thanks for your time looking at this! -- You are receiving this mail because: You are watching all bug changes.
