On Mon, Jul 28, 2025 at 7:53 AM Alfie Richards <alfie.richa...@arm.com> wrote: > > Hi, > > Small fix to resolve an UBSAN diagnostic. > > Reg tested for Aarch64 > > Thanks, > Alfie > > -- >8 -- > > This adds a nullptr check to fix a regression where it is possible to call > `memcmp (NULL, NULL, 0)` which is UB.
This is obvious at least to me. Also `memcmp (NULL, NULL, 0)` is not UB always; C26 makes it well defined. nonnull_if_nonzero is on the builtins these days but many libc (older glibc included) still uses nonnull which has priority over it. Thanks, Andrew Pinski > > This should fix the bootstrap-ubsan build. > > gcc/ChangeLog: > PR middle-end/121261 > * vec.h: Add null ptr check. > --- > gcc/vec.h | 4 ++++ > 1 file changed, 4 insertions(+) > > diff --git a/gcc/vec.h b/gcc/vec.h > index 9604edb1c3c..1277a881f1c 100644 > --- a/gcc/vec.h > +++ b/gcc/vec.h > @@ -2514,6 +2514,10 @@ public: > return false; > if (lhs.size () != rhs.size ()) > return false; > + /* Case where either is a NULL pointer and therefore, as both are valid, > + both are empty slices with length 0. */ > + if (lhs.begin () == NULL || rhs.begin () == NULL) > + return true; > return memcmp (lhs.begin (), rhs.begin (), lhs.size ()) == 0; > } > > -- > 2.34.1 >
