Thanks for your reply, but what is the root cause of that fatal condition.
I mean, we have mounted two modules on io-pkt:
- one ip filter
- one virtual driver
I suppose that something there cause this problem, but up to now I dont' know where.

As you see the problem appear with IP traffic involving two interface (a real one and the virtual one) and so the fast 
forward optimization .. but only on high rate, otherwise all stuff works fine.

Note that before this event I see a lot of this other traces on sloginfo
Jun 08 12:28:19    5    14     0 ient 810ac9c from 0 to 1
Jun 08 12:28:20    5    14     0 ient 810ac9c from 1 to 0
Jun 08 12:28:20    5    14     0 ient 8241c9c from 0 to 1

Thanks for any idea or detail about the root cause of this core.