https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71445
--- Comment #17 from Jakub Jelinek <jakub at gcc dot gnu.org> --- Well, you should then revert your compiler-rt patch (and most likely double check the signed vs. unsigned types, in the patch you've committed there were various signedness (but for big endian even size) issues). And, somebody from the sanitizer folks should have look at the write_msghdr/recgmsg sanitizer wrapper issues I've raised.
