https://gcc.gnu.org/bugzilla/show_bug.cgi?id=125977
Richard Biener <rguenth at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |jakub at gcc dot gnu.org
--- Comment #3 from Richard Biener <rguenth at gcc dot gnu.org> ---
Hmm, I wouldn't expect any range analysis at -O0, so it's probably a bug that
we arrive with that set during RTL expansion.
We have
enable_ranger (fun);
lab_rtx_for_bb = new hash_map<basic_block, rtx_code_label *>;
head_end_for_bb.create (last_basic_block_for_fn (fun));
FOR_BB_BETWEEN (bb, init_block->next_bb, EXIT_BLOCK_PTR_FOR_FN (fun),
next_bb)
bb = expand_gimple_basic_block (bb, var_ret_seq);
disable_ranger (fun);
in cfgexpand.cc, I think we want to guard enable/disable_ranger with if
(optimize)?
