https://gcc.gnu.org/bugzilla/show_bug.cgi?id=124086
--- Comment #3 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
So we had before:
_4 = VIEW_CONVERT_EXPR<complex long double>({ 0, -13, 0, 0, 0, 0, 0, 0 });
_5 = REALPART_EXPR <_4>;
Oh you know what, there is most likely padding which is why the VCE of a vector
constant is not folding.
