> Not quite. The unit could be removed while mounted. The SCSI layer > doesn't know that the media might have been changed. So, while START_UNIT > is the SCSI layer's way of probing for media change, it's not the only way > that such a thing can be detected.
The media has no lock then ? I'd still implement the simple start_unit change. That will fix the 99% of events that actually matter. _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
