Re: AC_C_INLINE with -Werror llvm 3.4

[Kurt Schwehr]

Eric,

Thanks for the feedback.  I definitely had the feeling that it would be
impossible to get rid of all warnings for all compilers.

Two questions back at you:

1) Do you have a example of a source package that handles -Werror in a
clean way so that I can use it a s model try to add the same to these other
packages?

2) Does the suggested fix I gave make sense?  I think that means: a) It's
not going to break any environments without -Werror or equivalent for
compilers out there. b) It will generally reduce the number of
compilers/flag combinations that would issue a warning for this test?  If
this proposed fix is a bad idea, why is it bad?

Thanks,
-kurt


On Mon, Jun 9, 2014 at 8:17 AM, Eric Blake <address@hidden> wrote:

> On 06/08/2014 10:25 AM, Kurt Schwehr wrote:
> > I'm trying to clean hdf5 1.9.182 a bit and I'm using -Werror as my first
> > pass tool.  Just cleaning the messages that come up.  However, passing
> > -Werror as CPPFLAGS causes configure to die on a number of places.
>
> Don't do that.  It is not possible to make autoconf warning-free for all
> possible compilers.  So our advice has always been to probe for -Werror
> independently from other compiler flags, and add it in at the last
> minute, but NOT to expect configure to be running with -Werror in effect.
>
> That said, if you have a simple patch that reduces one case of false
> test results due to -Werror, we can still probably apply it.
>
> --
> Eric Blake   eblake redhat com    +1-919-301-3266
> Libvirt virtualization library http://libvirt.org
>
>


-- 
--
http://schwehr.org