https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113907
--- Comment #35 from rguenther at suse dot de <rguenther at suse dot de> --- On Thu, 15 Feb 2024, rguenther at suse dot de wrote: > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113907 > > --- Comment #34 from rguenther at suse dot de <rguenther at suse dot de> --- > On Thu, 15 Feb 2024, jakub at gcc dot gnu.org wrote: > > > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113907 > > > > --- Comment #32 from Jakub Jelinek <jakub at gcc dot gnu.org> --- > > (In reply to Jan Hubicka from comment #31) > [...] > > > Not streaming value ranges is an omission on my side (I mistakely assumed > > > we > > > do stream them). We ought to stream them, since otherwise we will lose > > > propagated return value ranges in partitioned programs, which is pity. > > > > Not sure if it is a good idea to start streaming them now in stage4, but > > sure, > > Please don't do that now ;) > > (we've not had ranges early originally, and even nonzero bits were not > computed) > > But yes, IPA CP jump functions with value-ranges might be a problem. > I think it does instantiate them as local ranges, does it? We > have streaming support for ranges, we just don't stream the auxiliary > data for all SSA names (like also not points-to info). Also remember we like to have a fix that's easily backportable, and that's probably going to be resetting the info. We can do something more fancy for GCC 15
