This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Question on -Werror usage in Makefiles...
> From: Andreas Jaeger <aj@suse.de>
>
> "Kaveh R. Ghazi" <ghazi@caip.rutgers.edu> writes:
>
> > Sorry, I think your approach is wrong. You're adding strict warnings
> > (-pedantic) to target files which are only compiled by gcc and thus
> > should be able to use gcc-isms. Then you're working on needlessly
> > fixing these warnings and/or also proposing adding -Wno-error to new
> > files which used to compile cleanly.
> >
> > I think you should consider another approach. I'm going to guess that
> > the new warnings from SYSCALLS.c are all of the "old-style"
> > declaration type? If so, you may want to simply add
> > -Wno-old-style-declarations to the rule which compiles that file.
> > Should be a one line fix.
>
> It is - I send it previously. But I was surprised that the
> warn=-Wno-error black-listing does not work...
Let's ignore the black-listing trick for now on that file, it doesn't
seem to work and the Makefile is ugly enough. Stick with the Makefile
rule instead.
Ah, I see now that you did try adding something to the rule - except
you added -Wno-error. That'll stop it from erroring, but you'll still
get a gillion lines of (IMO useless) warning diagnostics.
On solaris2, the only warnings are "function declaration isn't a
prototype". So you should only need to add -Wno-strict-prototypes.
The file is just a horking big set of system prototypes, and there are
some for which we don't know (or can't guarantee) the arguments, so I
feel that disabling that warning is the correct approach.
--
Kaveh R. Ghazi ghazi@caip.rutgers.edu