Easily fixable nge(4) issue?

[Yar Tikhiy]

Folks,

Has anybody looked at kern/86618?  It looks like a piece of cake.
Yuriy, the originator, tells that nge(4) panics his RELENG_6 and
HEAD systems, but adding M_ZERO to contigmalloc() flags at one spot
in if_nge.c is enough to get rid of the trouble (patch included.)
Yuriy and yours truly were communicating over a VLAN issue and he
drew my attention to the PR, but unfortunately I don't have any
nge(4) hardware.  If there is consensus that the problem description
and fix are correct, I can back the solution with my commit bit.

-- 
Yar