https://gcc.gnu.org/bugzilla/show_bug.cgi?id=89247
--- Comment #9 from Richard Biener <rguenth at gcc dot gnu.org> --- (In reply to Jakub Jelinek from comment #8) > Or fold LOOP_VECTORIZED into 0 in fold_builtins pass just in case something > makes it through? > Could fix PR87446 too. Yeah, though the whole point was to see if we properly update during vectorization - if we simply "fold" later we lose this sanity check. I'm currently testing a fix in if-conversion itself.
