https://gcc.gnu.org/bugzilla/show_bug.cgi?id=78968
--- Comment #3 from Jonathan Wakely <redi at gcc dot gnu.org> --- (In reply to h2+bugs from comment #0) > This is a bug in gcc. The libsupc++ configuration assumes that there are > exactly two possibilities: either libc is glibc and implements > __cxa_thread_atexit() using __cxa_thread_atexit_impl(), or libsupc++ must > provide an implementation on its own. Although this isn't quite right, there's no assumption of glibc. The two possibilities are that only __cxa_thread_atexit_impl exists, or that neither __cxa_thread_atexit_impl nor __cxa_thread_atexit exists. There's no assumption about the provenance of __cxa_thread_atexit_impl.
