Re: [PATCH] fix some clang warnings

[John Keeping <john@keeping.me.uk>]

On Wed, Jan 16, 2013 at 10:15:58AM -0800, Jeff King wrote:
> On Wed, Jan 16, 2013 at 06:12:03PM +0000, John Keeping wrote:
> 
> > On Wed, Jan 16, 2013 at 10:00:42AM -0800, Jeff King wrote:
> > > It is not about the macro itself, but rather the callsites that do not
> > > return error, but call it for its printing side effect. It seems that
> > > clang -Wunused-value is OK with unused values from functions being
> > > discarded, but not with constants. So:
> > > 
> > >   int foo();
> > >   void bar()
> > >   {
> > >     foo(); /* ok */
> > >     1; /* not ok */
> > >     (foo(), 1); /* not ok */
> > >   }
> > > 
> > > The first one is OK (I think it would fall under -Wunused-result under
> > > either compiler). The middle one is an obvious error, and caught by both
> > > compilers. The last one is OK by gcc, but clang complains.
> > 
> > I wonder if this would be changed in clang - the change in [1] is
> > superficially similar.
> > 
> > [1] http://llvm.org/bugs/show_bug.cgi?id=13747
> 
> Yeah, I think it is exactly the same issue, and the fix they mention
> there would apply to us, too.
> 
> Is it worth applying this at all, then? Or should we apply it but limit
> it with a clang version macro (they mention r163034, but I do not know
> if it is in a released version yet, nor what macros are available to
> inspect the version)?

That maps to revision 06b3a06007 in their git repository [1], which is
contained in remotes/origin/release_32 so I think that change should be
in release 3.2, where I still see the warning (although that's not using
a clang built from that source), so I don't think that the fix for that
bug removes the warning in this case.

[1] http://llvm.org/git/clang.git