https://gcc.gnu.org/bugzilla/show_bug.cgi?id=127953
--- Comment #3 from Drea Pinski <pinskia at gcc dot gnu.org> --- (In reply to Gero from comment #2) > However, shifts using __int128 are not vectorized. But that's exactly the > point. Are you sure? If it is not then file a bug, the x86 backend does promote to use the sse registers in many cases.
