https://gcc.gnu.org/bugzilla/show_bug.cgi?id=119377
--- Comment #24 from Robert Dubner <rdubner at gcc dot gnu.org> --- It's my belief that the last thing I talked about, which had to do with valgrind warning about the read access of uninitialized alignment padding bytes in a cbl_declarative_t structure, has been resolved. (A memcpy of the entire structure has been eliminated.) This discussion covered a lot of ground, and some of it is on architectures to which I do not, as yet, have access. Is there any reason to keep this PR open?