what is MT_PKTHDR? I guess MT_HEADER is meant instead.

As for the test, the real constraint is not to accidentally mix mbufs 
from different "sources" such as a datagram mbuf and a pfkey mbuf...
but since datagrams can be in different mbuf types, the test is bogus.


gnatios Souvatzis wrote:

> hi,
> 
> On Tue, Oct 05, 2004 at 11:33:15AM +0200, Pavel Cahyna wrote:
> 
> 
>>If the difference between MT_DATA and MT_PKTHDR is vaguely defined,
>>why not keep the test and get rid of MT_PKTHDR completely and replace
>>it with MT_DATA?
> 
> 
> Discussing, defining and implementing this will surely need more time than
> is left before 2.0.
> 
> 	-is