https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115533
--- Comment #26 from Alexander Monakov <amonakov at gcc dot gnu.org> --- (In reply to Richard Biener from comment #24) > > That's because of -fno-vect-cost-model, it wouldn't be vectorized otherwise. Thanks, I forgot. The testcase in PR 106902 was vectorized at plain -O3 before gcc-14, though. If the cost model is better about such pitfalls now, I'm happy.