http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53176
--- Comment #11 from rsandifo at gcc dot gnu.org <rsandifo at gcc dot gnu.org> 2012-05-02 22:36:28 UTC --- (In reply to comment #8) > Actually, I think that's what rtlanal.c:rtx_cost should be changed to do, for > the cases where the targetm.rtx_costs function returns false. Yeah, I'd wondered about that too. We've tended to make targets fend for themselves when it comes to doubleword ANDs, ADDs, etc., but I suppose that doesn't mean we should...
