https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69729
--- Comment #3 from Ilya Enkovich <ienkovich at gcc dot gnu.org> --- (In reply to Jan Hubicka from comment #2) > Yes, the patch is meant to disbale streaming of instrumentation thunks, so > we do not output two function bodies for one assembler name into LTO files. > Can we possibly just fix the conditional instead of fully reverting the > patch? Instrumentation thunks have no body, we call release_body for all of them.
