References : http://lkml.org/lkml/2007/10/3/101 Submitter : Joerg Platte <lists@naasa.net>
that change probably wasn't intentional. There was some flux in the scsi scanning code at that time. if it isn't fixed already and you can reproduce it (3 years ago?) , please reopen this bug. i checked, and there were some fixes related to scsi scanning and also some reverts of some bad conversions from/to the "new" scsi hotplug model... i close it as unreproducible.
(In reply to comment #1) > i close it as unreproducible. at least i would if i could:)
Just for the record: I tried to reproduce it today and the bug seems to be fixed now. But it was there maybe 4 months ago.