On Sun, May 02, 2004 at 01:51:09PM +1000, Darren Reed wrote:
> 
> The patch you posted undoes a change I made to bpf_mtap() that
> has catchpacket call either memcpy() or bpf_memcpy(), depending
> on whether or not the entire packet is in a single mbuf.

Darren and I talked; we agree that I did not actually undo this
change.

> bpf_measure() should really be a m_length() function and put in
> either mbuf.h or uipc_mbuf.c, not bpf.c, along with a prototype
> in the appropriate place.

I will move it to mbuf.h and I will add a prototype.

I will not move m_length to uipc_mbuf.c right now, because it was
previously inline code, and I cannot predict the impact of making it
non-inline.

> And, btw, a prototype for bpf_deliver() is also missing.

Will fix.

> i.e. your patch needs more work before integrating into NetBSD.

Ooops, I already committed.  I think that Darren, Matt, and I agree
that with changes, the code is acceptable.  I will send a second patch
to the list.

> Also, see what Sam said about its raison d'etre.

I know its raison d'etre, thanks; sorry I was not as concise as Sam.

Dave

-- 
David Young             OJC Technologies
dyoung@ojctech.com      Urbana, IL * (217) 278-3933