https://gcc.gnu.org/bugzilla/show_bug.cgi?id=122096
--- Comment #15 from Drea Pinski <pinskia at gcc dot gnu.org> --- (In reply to Haochen Jiang from comment #14) > The issue for 736.ocio_r is many frequently inlined function got > devirtualization, like _M_release(), as shown in dump. Those caused the > codesize increased a lot since the code piece is repeated. > > Should we adjust threshold here for -O2? I wonder if we could get some better speculatively targets instead ... That is get a smaller set in some cases instead of changing the max number of targets. I have not looked into that yet though.
