https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115533
--- Comment #24 from Richard Biener <rguenth at gcc dot gnu.org> --- (In reply to Alexander Monakov from comment #23) > I suggest it to close this a dup of PR 106902 if there are no better ideas. > > By the way, in both cases SLP introduces vectors in a loop where scalar > computations it's attempting to replace are not eliminable in the end (there > are loop-carried dependencies). This makes me think something is amiss in > profitability estimation. That's because of -fno-vect-cost-model, it wouldn't be vectorized otherwise.
