The problem still persists. I've tried kernel 4.19.16 from Debian testing. The bug somehow lies in the logic for detecting a disk change. If I access the floppy the motor won't turn off. But if I force a disk change (ejecting and immediately inserting the floppy again) the motor turns off within about two seconds - as it should.
Also, if I practically disable the disk change checks with "floppycontrol -C 2147483647" the motor turns off normally. So I suspect that frequently checking for disk changes without the floppy being changed keeps the motor in its current state.
The problem still persists. I've tried kernel 4.19.16 from Debian testing. The bug somehow lies in the logic for detecting a disk change. If I access the floppy the motor won't turn off. But if I force a disk change (ejecting and immediately inserting the floppy again) the motor turns off within about two seconds - as it should.
Also, if I practically disable the disk change checks with "floppycontrol -C 2147483647" the motor turns off normally. So I suspect that frequently checking for disk changes without the floppy being changed keeps the motor in its current state.