ports/85256: fwtk port broken in 5.4

[Christoph Weber-Fahr]

>Number:         85256
>Category:       ports
>Synopsis:       fwtk port broken in 5.4
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    freebsd-ports-bugs
>State:          open
>Quarter:        
>Keywords:       
>Date-Required:
>Class:          sw-bug
>Submitter-Id:   current-users
>Arrival-Date:   Tue Aug 23 21:30:10 GMT 2005
>Closed-Date:
>Last-Modified:
>Originator:     Christoph Weber-Fahr
>Release:        FreeBSD 5.4-RELEASE-p2 i386
>Organization:
Arcor AG
>Environment:
System: FreeBSD tnd37.tnd.arcor.net 5.4-RELEASE-p2 FreeBSD 5.4-RELEASE-p2 #1: Fri Jul 8 21:00:21 CEST 2005 root at tnd37.tnd.arcor.net:/usr/obj/usr/src/sys/HPDC7100 i386

>Description:
	the fwtk port fails to compile. 
	I found several issues, and suspect there might be some more
	- it tries to build its library with "-lskey" which fails, apparently 
	  lacking libskey.a 
	- when you remove, for the sake of testing, skey from AUXLIBS and 
	  auth.h, it later falls prey to the "gcc 3.4 maintainers send 
	  varargs users to hell" syndrome. Error message:

	/usr/include/varargs.h:34:2: #error "<varargs.h> is obsolete with this version of GCC."
	/usr/include/varargs.h:35:2: #error "Change your code to use <stdarg.h> instead."

	- when you try to compile it using the gcc295 port, it exchanges this error
	  to a screenful of warnings about how vargs.h redefines stuff from stdio.h,
	  and later it fails with

	sig.c: In function `handle_sigpid':
	sig.c:37: storage size of `wstatus' isn't known
	*** Error code 1

	Stop in /usr/ports/security/fwtk/work/fwtk/x-gw.

	Thats the point where I gave up. Apparently there are some issues with 
	fwtk on freebsd5.

	FWIW, the jumbo patch (enabled with an extra make option, as decsribed in
	the make file) also fails because there is no md5 file for that.


>How-To-Repeat:

	cd /usr/ports/security/fwtk
	make


>Fix:

	The varargs thing may or may not be fixable by including <machine/varargs.h>
	but I don't understand enough of the underlying compiler issues to know
	if this is a cosmetic or a real fix. In any case, you only run against other 
	issues then.



>Release-Note:
>Audit-Trail:
>Unformatted: