["Benninghoff, John (RBC Dain)" <John.Benninghoff@xxxxxxxxxxx>]

Yep, here's the exact output of my custom OpenBSD port that does the compile:

$ FLAVOR="snmp adns gtk1" make
===>  Building for tethereal-0.10.9-snmp-adns-gtk1
...
(bunch of stuff removed)
...
Making all in mate
gmake[3]: Entering directory `/usr/ports/mystuff/net/ethereal/w-ethereal-0.10.9-snmp-adns-gtk1/build-i386-snmp-adns-gtk1/plugins/mate'
source='/usr/ports/mystuff/net/ethereal/w-ethereal-0.10.9-snmp-adns-gtk1/ethereal-0.10.9/plugins/mate/packet-mate.c' object='packet-mate.lo' libtool=yes \
depfile='.deps/packet-mate.Plo' tmpdepfile='.deps/packet-mate.TPlo' \
depmode=gcc /bin/sh /usr/ports/mystuff/net/ethereal/w-ethereal-0.10.9-snmp-adns-gtk1/ethereal-0.10.9/depcomp \
/usr/local/bin/libtool  --mode=compile cc -DHAVE_CONFIG_H -I. -I/usr/ports/mystuff/net/ethereal/w-ethereal-0.10.9-snmp-adns-gtk1/ethereal-0.10.9/plugins/mate -I../.. -I/usr/ports/mystuff/net/ethereal/w-ethereal-0.10.9-snmp-adns-gtk1/ethereal-0.10.9  -I/usr/local/include -I/usr/local/include -O2 -pipe -Dopenbsd3 -I. -I/usr/local/include  -DINET6 -D_U_="__attribute__((unused))" -Wall -W  -O2 -pipe  -I/usr/local/include -I/usr/local/include -I/usr/local/include/gtk-1.2 -I/usr/local/include/glib-1.2 -I/usr/X11R6/include -I/usr/local/include -O2 -pipe -Dopenbsd3 -I. -I/usr/local/include -c -o packet-mate.lo /usr/ports/mystuff/net/ethereal/w-ethereal-0.10.9-snmp-adns-gtk1/ethereal-0.10.9/plugins/mate/packet-mate.c
 cc -DHAVE_CONFIG_H -I. -I/usr/ports/mystuff/net/ethereal/w-ethereal-0.10.9-snmp-adns-gtk1/ethereal-0.10.9/plugins/mate -I../.. -I/usr/ports/mystuff/net/ethereal/w-ethereal-0.10.9-snmp-adns-gtk1/ethereal-0.10.9 -I/usr/local/include -I/usr/local/include -O2 -pipe -Dopenbsd3 -I. -I/usr/local/include -DINET6 "-D_U_=__attribute__((unused))" -Wall -W -O2 -pipe -I/usr/local/include -I/usr/local/include -I/usr/local/include/gtk-1.2 -I/usr/local/include/glib-1.2 -I/usr/X11R6/include -I/usr/local/include -O2 -pipe -Dopenbsd3 -I. -I/usr/local/include -c /usr/ports/mystuff/net/ethereal/w-ethereal-0.10.9-snmp-adns-gtk1/ethereal-0.10.9/plugins/mate/packet-mate.c -Wp,-MD,.deps/packet-mate.TPlo  -fPIC -DPIC -o .libs/packet-mate.o
In file included from /usr/ports/mystuff/net/ethereal/w-ethereal-0.10.9-snmp-adns-gtk1/ethereal-0.10.9/plugins/mate/packet-mate.c:33:
/usr/ports/mystuff/net/ethereal/w-ethereal-0.10.9-snmp-adns-gtk1/ethereal-0.10.9/plugins/mate/mate.h:327: syntax error before `,'
/usr/ports/mystuff/net/ethereal/w-ethereal-0.10.9-snmp-adns-gtk1/ethereal-0.10.9/plugins/mate/packet-mate.c: In function `init_mate':
/usr/ports/mystuff/net/ethereal/w-ethereal-0.10.9-snmp-adns-gtk1/ethereal-0.10.9/plugins/mate/packet-mate.c:244: `mate_packet' undeclared (first use in this function)
/usr/ports/mystuff/net/ethereal/w-ethereal-0.10.9-snmp-adns-gtk1/ethereal-0.10.9/plugins/mate/packet-mate.c:244: (Each undeclared identifier is reported only once
/usr/ports/mystuff/net/ethereal/w-ethereal-0.10.9-snmp-adns-gtk1/ethereal-0.10.9/plugins/mate/packet-mate.c:244: for each function it appears in.)
gmake[3]: *** [packet-mate.lo] Error 1
gmake[3]: Leaving directory `/usr/ports/mystuff/net/ethereal/w-ethereal-0.10.9-snmp-adns-gtk1/build-i386-snmp-adns-gtk1/plugins/mate'
gmake[2]: *** [all-recursive] Error 1
gmake[2]: Leaving directory `/usr/ports/mystuff/net/ethereal/w-ethereal-0.10.9-snmp-adns-gtk1/build-i386-snmp-adns-gtk1/plugins'
gmake[1]: *** [all-recursive] Error 1
gmake[1]: Leaving directory `/usr/ports/mystuff/net/ethereal/w-ethereal-0.10.9-snmp-adns-gtk1/build-i386-snmp-adns-gtk1'
gmake: *** [all] Error 2
*** Error code 2

Stop in /usr/ports/mystuff/net/ethereal (line 1729 of /usr/ports/infrastructure/mk/bsd.port.mk).
$

-----Original Message-----
From: Guy Harris [mailto:gharris@xxxxxxxxx]
Sent: Friday, January 28, 2005 1:03 PM
To: Benninghoff, John (RBC Dain)
Cc: LEGO; Ethereal development
Subject: Re: [Ethereal-dev] FYI - bug (with patch) for ethereal-0.10.9


Benninghoff, John (RBC Dain) wrote:

> The test.c program compiles fine on my system.

But the MATE code didn't?  That's odd - the declaration of "foo()" used 
"__attribute__((unused))" in a function prototype that lacked parameter 
names, and that's the construct that was, apparently, the source of the 
problem in the MATE code.

> I would guess that _U_ is a GCC3-ism?

_U_ is an Ethereal-ism (Joerg Mayer might have been the person who came 
up with it).  __attribute__ is a GCC-ism, but it's not a GCC3-ism - it 
was in GCC 2.x as well.

_U_ is our way of supporting the ability to declare function arguments 
as unused in GCC without using a *LOT* of #ifdefs and without making it 
impossible to compile Ethereal with compilers that don't support 
"__attribute__".  (Another alternative would be to define 
"__attribute__(x)" as nothing - I think that's why the GCC folk put the 
extra parentheses there, to let that trick work on compilers that don't 
support varargs macros - but _U_ is shorter than 
"__attribute__((unused))", so it also arguably clutters the code a bit less.


RBC Dain Rauscher does not accept buy, sell or cancel orders by e-mail, or any instructions by e-mail that would require your signature.  Information contained in this communication is not considered an official record of your account and does not supersede normal trade confirmations or statements.  Any information provided has been prepared from sources believed to be reliable but is not guaranteed, does not represent all available data necessary for making investment decisions and is for informational purposes only.

This e-mail may be privileged and/or confidential, and the sender does not waive any related rights and obligations.  Any distribution, use or copying of this e-mail or the information it contains by other than an intended recipient is unauthorized.  If you receive this e-mail in error, please advise me (by return e-mail or otherwise) immediately.

Information received by or sent from this system is subject to review by supervisory personnel, is retained and may be produced to regulatory authorities or others with a legal right to the information.