The second half of the bug is that the kernel does not send an EJECT_REQUEST uevent any more when the hardware button is pressed. That definitively used to work in the past, but apparently something changed. I'll have a closer look at this.
The second half of the bug is that the kernel does not send an EJECT_REQUEST uevent any more when the hardware button is pressed. That definitively used to work in the past, but apparently something changed. I'll have a closer look at this.