On Tue, Jan 26, 2016 at 04:21:08PM +0100, Richard Biener wrote:
> > --- gcc/c/c-decl.c.jj 2016-01-21 00:41:47.000000000 +0100
> > +++ gcc/c/c-decl.c 2016-01-25 16:36:31.973504082 +0100
> > @@ -10741,11 +10741,19 @@ c_write_global_declarations_1 (tree glob
> > if (TREE_CODE (decl) == FUNCTION_DECL
> > && DECL_INITIAL (decl) == 0
> > && DECL_EXTERNAL (decl)
> > - && !TREE_PUBLIC (decl)
> > - && C_DECL_USED (decl))
> > + && !TREE_PUBLIC (decl))
> > {
> > - pedwarn (input_location, 0, "%q+F used but never defined", decl);
> > - TREE_NO_WARNING (decl) = 1;
> > + if (C_DECL_USED (decl))
> > + {
> > + pedwarn (input_location, 0, "%q+F used but never defined",
> > decl);
> > + TREE_NO_WARNING (decl) = 1;
> > + }
> > + /* For -Wunused-function push the unused statics into cgraph,
> > + so that check_global_declaration emits the warning. */
> > + else if (warn_unused_function
> > + && ! DECL_ARTIFICIAL (decl)
> > + && ! TREE_NO_WARNING (decl))
> > + cgraph_node::get_create (decl);
>
> Err, so why not warn here directly?
You mean check if it has a cgraph node (i.e. get instead of get_create) and
if it doesn't, warn? What I'm worried in that case is that it might have a
cgraph node created later on for whatever reason and that we'll get double
warning (from here and from cgraphunit.c (check_global_declaration)).
I can try it though.
Jakub
