Mark, On Tue, Nov 14, 2023 at 8:57 AM Mark Wielaard <m...@klomp.org> wrote:
> I am slightly surprised our testsuite didn't catch this. We do have -- > enable-sanitize-undefined which does build everything with -- > sanitize=undefined. Which should enable -fsanitize=pointer-overflow. > But I just tried (with gcc) and it seems to not trigger. This was exposed by Clang (and a close to HEAD Clang at that -- I am not sure whether released Clang also catches this). > Looks good. Applied. Thanks! Appreciate the speedy reviews. -- Paul Pluzhnikov
