[Bug c++/58950] [4.9 Regression] Missing "statement has no effect"

[glisse at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58950

--- Comment #4 from Marc Glisse <glisse at gcc dot gnu.org> ---
(In reply to Manuel López-Ibáñez from comment #3)
> You could decorate __builtin_shuffle with attribute "warn_unused_result". It
> is not the same, but it probably more precise. It would also warn for:
> __builtin_shuffle(x,++m), while -Wunused-value would not.

Would it? The front-ends replace __builtin_shuffle with VEC_PERM_EXPR quite
early. I am not even sure we have a function type to which we could attach the
attribute.