https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71084
--- Comment #6 from Richard Biener <rguenth at gcc dot gnu.org> --- (In reply to Ilya Enkovich from comment #5) > (In reply to Richard Biener from comment #3) > > fwprop_init calcuates dominator info so it's likely df that runs cfg-cleanup > > and doesn't expect it to wreck dominator info. > > > > Probably rightfully so. > > So should we improve cse_cfg_altered computation taking cfg_cleanup calls > into account or free dominance info in cfg_cleanup instead? The former.