MFC of ether_input() changes

[Andre Oppermann]

Bruce M Simpson wrote:
> Actually, I thought the change which moved the VLAN tag out of the mbuf 
> tag pool and into the mbuf packet header had also been MFCed. It has not.

It would be an API+ABI change to MFC it.  That's why I haven't done it.

> As CURRENT is the branch normally used for feature development it is 
> probably best I don't MFC this unless the VLAN tag change is MFCed also.
> 
> Therefore there is not a lot of point in merging this change apart from 
> to benefit from the code cleanup which M_PROMISC offers, so I'll back 
> off for now.

-- 
Andre