https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69433
--- Comment #4 from Andrew Pinski <pinskia at gcc dot gnu.org> --- >IMO we should look into why this optimization doesn't happen before PRE (why >not FRE for instance?). Because the reads from s are only partially redundant (PRE) and not fully (vs FRE). :)
