> > Any ideas to do that without keeping the scn data around until the
> > final close?
>
> No, I think that's acceptable to do, as long as you don't try to do a
> USB accesses with it at that point in time :)
Upon further thought this is debateable.
It's obviously true if nobody without physical access can cause
a disconnect. But this is not the case. There's a possibility of a DOS
attack. Take a device with two interfaces. It just takes a task that opens
the first interface, then causes a device reset through usbfs rebinding
the second interface and this in turn with the second interface. Do this in
turns, fork a little and you get a lot of memory the OOM killer won't see.
Regards
Oliver
-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
