[Bug tree-optimization/119683] [13/14/15/16 Regression] recalculating the return value which was already in the register right before function return

[jakub at gcc dot gnu.org via Gcc-bugs]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=119683

--- Comment #5 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Note, guess this is also related to PR122733, where I've been wondering where
such an optimization should be done.  match.pd would be best if there was the
possibility to at least sometimes (during vrp*) to query the local range rather
than the global one.