In message <Pine.BSF.4.51.0602171507150.95310@vegeta.city-net.com>, Matthew Org
ass writes:
>On 2006-02-16 smb@cs.columbia.edu wrote:
>> Page fault....
>>
>> usbd_alloc_xfer
>> usbd_do_request
>> usbd_do_request_flags
>> usbd_reset_port
>
>  This should have been:
>usbd_alloc_xfer
>usbd_do_request_flags_pipe
>usbd_do_request_flags
>usbd_do_request
>usbd_reset_port
>
>  If not, some stack corruption is involved.  Otherwise, it sounds like
>kern/32011.  The problem is that one of the data structures in the
>dev->bus->methods->allocx chain is freed early in detach.
>

I'll try to get another trace.

One problem is that I can't get a dump, since my RAM is too large for 
my swap area.  I obviously made a mistake when I configured the disk, 
but I can't easily correct that now.  If I build a kernel with
REALEXTMEM set to something less than my swap size, will I be able to 
get a dump?

		--Steven M. Bellovin, http://www.cs.columbia.edu/~smb