On Thu, 5 Dec 2013 20:01:24 -0500, Daniel Johnson 
<daniel.johnso...@gmail.com> wrote:
> On Dec 5, 2013, at 2:25 PM, Remi Mommsen <remigius.momm...@cern.ch> wrote:
>
> > All this did not help. However, I tracked it to the following issue:
> > > configure at line 11523 checks if ldap_set_rebind_proc takes 2 or 
> 3 arguments. However, the CFLAGS are extended on line 11530 to mark 
> all warnings as errors. This causes the "clang: warning: argument 
> unused during compilation" which also happens here to be promoted to 
> an error, which in turn causes the compile to fail on line 15551. 
> Thus, the 'ac_cv_ldap_set_rebind_proc_style' gets set to two instead 
> of three. Calling clang with '-Qunused-arguments' quiets the warning 
> about unused arguments. I.e. the following change to 
> libaprutil.0-shlibs.info solves the issue for me:
> > > diff -u -r1.7 libaprutil.0-shlibs.info
> > --- libaprutil.0-shlibs.info 27 Nov 2013 03:06:10 -0000 1.7
> > +++ libaprutil.0-shlibs.info 5 Dec 2013 19:21:46 -0000
> > @@ -36,6 +36,7 @@
> > PatchScript: <<
> >   ### Fix layout
> >   perl -pi -e 's,/usr/local,%p,g' config.layout
> > +  perl -pi -e 's,-Werror,-Werror -Qunused-arguments,g' configure
> >   %{default_script}
> > <<
> > > > The question of course remains why it only fails for me ):
> > > Remi
>
> Yeah, I have no idea. I had someone else try and also couldn't 
> duplicate it on 10.8 or 10.9. I've added your patch since it 
> certainly can't hurt anything. I'm still curious as to what's causing 
> this though. "clang: warning: argument unused during compilation" 
> seems to be popping up more frequently now. This is the third time 
> I've seen something related to it in the past couple days. 

That's the generic "unknown flag" diagnostic (wording suggests it's 
something like passing a linker-stage flag to the source-compile stage, 
but it could just as well be a totally bogus flag, altogether. (a 
totally bogus flag)

Various upstreams might be trying to guess what flags to use based on 
heuristics of -v output or pathnames rather than testing them directly. 
The more our compiler looks like "some other one" but doesn't support 
the exact same -W* and other flags, and the more other compilers add 
new flags that they do support, the more these warnings appear for us. 

dan

  --
Daniel Macks
dma...@netspace.org



------------------------------------------------------------------------------
Sponsored by Intel(R) XDK 
Develop, test and display web and hybrid apps with a single code base.
Download it for free now!
http://pubads.g.doubleclick.net/gampad/clk?id=111408631&iu=/4140/ostg.clktrk
_______________________________________________
Fink-users mailing list
Fink-users@lists.sourceforge.net
List archive:
http://news.gmane.org/gmane.os.macosx.fink.user
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-users

Reply via email to