https://bugs.kde.org/show_bug.cgi?id=515736
[email protected] changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |[email protected] --- Comment #1 from [email protected] --- I think I've been hitting this same bug, just from a different angle. Mine isn't a fullscreen toggle — it's resume from suspend. Setup: video playing in Haruna, close the lid, open it back up, crash within a couple seconds. Happened twice, about a week apart, and both times gdb gave me the exact same backtrace: QQmlIncubationController::incubateFor → QQmlObjectCreator::finalize → QQmlBinding::update → AOTCompiledContext::getObjectLookup → [org.kde.desktop QML, liborg_kde_desktop.so] → QQuickItem::setImplicitSize → libqqc2desktopstyleplugin.so → QQuickItem::geometryChange → QQmlNotifier::emitNotify → QQmlBoundSignalExpression::evaluate → AOTCompiledContext::callQmlContextPropertyLookup ← null deref here That's the same core path as the trace already on this bug — callQmlContextPropertyLookup, QQmlObjectCreator::finalize, setImplicitSize, the same ScrollBar/qqc2-desktop-style code. It splits off after that, so I don't think it's literally the same stack, but it looks like the same root cause hit through a different door. In my case the crash lines up exactly (same second, checked against journalctl) with KWin failing to reset the display on wake: kwin_wayland_drm: Atomic modeset test failed! Invalid argument which tears down and rebuilds the outputs. So instead of a fullscreen toggle yanking the geometry out from under a ScrollBar mid-incubation, it's the screen itself disappearing and coming back. Different trigger, same mechanism. If that's right, this probably isn't really a "crashes on fullscreen toggle" bug so much as "crashes on any sudden geometry change while a qqc2-desktop-style control is mid-init" — fullscreen toggle just happens to be the easiest way to trigger it. For what it's worth I'm on a newer stack than the original report: Haruna 1.7.1, Qt 6.10.2, Mesa 26.0.3, Intel i915. Nothing GPU or codec related shows up anywhere in either trace, it's all QML/style layer. Can attach the coredumps if that's useful. -- You are receiving this mail because: You are watching all bug changes.
