On Fri, 25 Jan 2008, David Young wrote:
> On Fri, Jan 25, 2008 at 04:02:15PM +0000, Steven M. Bellovin wrote:
>> drvctl hung when I tried using it to
>> detach one of the offending USB ports.)
>
> Where has drvctl hung in the kernel?  Can you drop into the kernel
> debugger and get a backtrace for drvctl's PID?

This can happen if the device's detach function assumes that the device is 
powered on, and tries to access an io or mem resource with the device in 
D3 state. It is likely that his USB initialization was not successful 
after resume from S3, and the devices were simply powered off.

Jared