Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 21 Jan 1997 16:31:59 +1100 (EST)
From:      proff@suburbia.net
To:        kuku@gilberto.physik.rwth-aachen.de (Christoph Kukulies)
Cc:        hackers@freebsd.org
Subject:   Re: FreeBSD 3.0-current port
Message-ID:  <19970121053159.6549.qmail@suburbia.net>
In-Reply-To: <199701201925.UAA21346@gilberto.physik.rwth-aachen.de> from Christoph Kukulies at "Jan 20, 97 08:25:42 pm"

next in thread | previous in thread | raw e-mail | index | archive | help
> Is anyone doing that at present? 
> I started out but ran into some difficulties/glitches:
> 
> 
> 1)  if.h in FreeBSD 3.0 has a struct member timeval something.
>     This results in having to include <sys/time.h> in front of every if.h
>     inclusion.

I have agreement from FreeBSD-core that all include files should be self
resolving. The solution in this case is to modify if.h and send
in a patch as a PR.

> 2)  ipl.h contains a macro IPL_VERSION
> 
>     #define IPL_VERSION     "IP Filter v3.1.4 - 10/1/97"
> 
>     which is used in mln_ipl.c in the MOD_DEV macro
>     top build some variable names (MOD_DECL).
>     
>     This leads to very odd names - uncompilable.

I can't recall how I got around this one, but 7 months
ago some of the FreeBSD-current MOD_* macros were
definitely inconsistent.

> 3)  some clash occurs with the
>     extern struct cdevsw cdevsw[]; in same file
>    
>     It clashes with the definition of /usr/include/sys/conf.h:
>     extern struct cdevsw *cdevsw[];

FreeBSD uses dynamically allocated cdevsw's. imho ipfilter should
not be manipulating the cdevsw directly at all (there are macros/
functions in the kernel to do this). Darren's rational (correct me
if I'm wrong) behind scanning through cdevsw directly was to ensure
free major/minor number pairs.

> So far I have come by now but I'm stuck at the moment. Anyone else
> got it to compile? 

Darren has recently been given commit access to FreeBSD-current in
order to integrate ipfilter. From from my preliminary code-review
there are no significant changes to be made appart from the addition
of divert sockets (and my socket-credential code <grin>). Hopefully
we will see something soon.

Cheers,
Julian <proff@iq.org>



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19970121053159.6549.qmail>