https://gcc.gnu.org/bugzilla/show_bug.cgi?id=126142
--- Comment #2 from ptomsich at gcc dot gnu.org --- The IVCANON pointer is somewhat ironic: our first prototype of the rewrite did live there, but we had to move it because of pass-ordering. We will structure the series accordingly and treat this PR as a tracker for the the dead-helper-cycle + induction-index variant of PR88259. That said, I believe that the intent of this PR matches what you want.
