I'm not convinced it was turning off the atapi attribute that did it.
I've seen something similar to this, and it seems to sometimes work,
and sometimes not.

    -- Lennart

Simon Burge wrote:

>On Mon, Jul 15, 2002 at 12:34:30PM +1000, Simon Burge wrote:
>
>  
>
>>>Number:         17599
>>>Category:       kern
>>>Synopsis:       panic when disconnecting a USB camera from a USB hub.
>>>      
>>>
>
>So, turning off the atapi attribute in files.usb for umass allows this
>to work flawlessly:
>
>	umass0 at uhub1 port 4 configuration 1 interface 0
>	umass0: Sony Sony DSC, rev 1.10/4.10, addr 4
>	umass0: using RBC over CBI
>	scsibus0 at umass0: 2 targets, 1 luns per target
>	scsibus0: waiting 2 seconds for devices to settle...
>	sd0 at scsibus0 target 1 lun 0: <Sony, Sony DSC, 4.10> SCSI0 0/direct removable
>	sd0: mode sense (4) returned nonsense; using fictitious geometry
>	sd0: 63424 KB, 61 cyl, 64 head, 32 sec, 512 bytes/sect x 126848 sectors
>	umass0: at uhub1 port 4 (addr 4) disconnected
>	sd0 detached
>	scsibus0 detached
>	umass0 detached
>
>Now that I think of it, I may not have tried the a -current kernel
>(since the scsi and atapi attributes were added) with the camera plugged
>directly into the laptop - I used to have only something like:
>
>	umass* at uhub?
>	scsibus* at umass?
>	sd* at scsibus?
>
>There's already some Sony quirks in umass_quirks.c; I'm not sure how/if
>to tell this device to not use atapi and use scsi only.  Here is the device
>info from "usbdevs":
>
>	  port 4 addr 4: full speed, self powered, config 1, Sony DSC(0x0010), Sony(0x054c), rev 4.10
>
>Simon.
>--
>Simon Burge                                   <simonb@wasabisystems.com>
>NetBSD Development, Support and Service:   http://www.wasabisystems.com/
>
>  
>