https://gcc.gnu.org/bugzilla/show_bug.cgi?id=124454
--- Comment #2 from Jakub Jelinek <jakub at gcc dot gnu.org> --- Anyway, seems the cfg is in some non-cleaned-up state at that point, as pre_and_rev_post_order_compute_fn for a function with n_basic_blocks_for_fn (fn) 29 is about to return 20 (but before doing that it accesses rev_post_order array on the uninitialized members and crashes even before reaching the assert later on.
