Just to triage the bug: this hits the (maybe over eager?) warnings: WARN_ON(!wm_changed). One would need to understand why the WM values have changed and why I didn't expected that to happen there.
Just to triage the bug: this hits the (maybe over eager?) warnings: WARN_ON( !wm_changed) . One would need to understand why the WM values have changed and why I didn't expected that to happen there.