http://gcc.gnu.org/bugzilla/show_bug.cgi?id=59600
--- Comment #4 from Andrew Pinski <pinskia at gcc dot gnu.org> --- (In reply to Andrew Pinski from comment #3) > (In reply to Yury Gribov from comment #1) > > Created attachment 31515 [details] > > Draft patch > > > Why can't you just set DECL_UNINLINABLE on the function decl in > handle_no_sanitize_undefined_attribute in c-common.c just like how > handle_noinline_attribute handles it? Or better yet handle the case where the attributes diff inside function_attribute_inlinable_p so you don't lose that much optimizations only when the attributes differ.