diff --git a/library/alloc/src/ffi/c_str.rs b/library/alloc/src/ffi/c_str.rs
index ae61b1f1e8ed5..be21d8c722d78 100644
--- a/library/alloc/src/ffi/c_str.rs
+++ b/library/alloc/src/ffi/c_str.rs
@@ -436,9 +436,9 @@ impl CString {
     ///
     /// unsafe {
     ///     assert_eq!(b'f', *ptr as u8);
-    ///     assert_eq!(b'o', *ptr.offset(1) as u8);
-    ///     assert_eq!(b'o', *ptr.offset(2) as u8);
-    ///     assert_eq!(b'\0', *ptr.offset(3) as u8);
+    ///     assert_eq!(b'o', *ptr.add(1) as u8);
+    ///     assert_eq!(b'o', *ptr.add(2) as u8);
+    ///     assert_eq!(b'\0', *ptr.add(3) as u8);
     ///
     ///     // retake pointer to free memory
     ///     let _ = CString::from_raw(ptr);
diff --git a/library/alloc/src/vec/mod.rs b/library/alloc/src/vec/mod.rs
index fa9f2131c0c1d..fe4dcafe14c32 100644
--- a/library/alloc/src/vec/mod.rs
+++ b/library/alloc/src/vec/mod.rs
@@ -542,8 +542,8 @@ impl<T> Vec<T> {
     ///
     /// unsafe {
     ///     // Overwrite memory with 4, 5, 6
-    ///     for i in 0..len as isize {
-    ///         ptr::write(p.offset(i), 4 + i);
+    ///     for i in 0..len {
+    ///         ptr::write(p.add(i), 4 + i);
     ///     }
     ///
     ///     // Put everything back together into a Vec
@@ -702,8 +702,8 @@ impl<T, A: Allocator> Vec<T, A> {
     ///
     /// unsafe {
     ///     // Overwrite memory with 4, 5, 6
-    ///     for i in 0..len as isize {
-    ///         ptr::write(p.offset(i), 4 + i);
+    ///     for i in 0..len {
+    ///         ptr::write(p.add(i), 4 + i);
     ///     }
     ///
     ///     // Put everything back together into a Vec
diff --git a/library/core/src/intrinsics.rs b/library/core/src/intrinsics.rs
index 15467e0191dbf..3d7b875eac15d 100644
--- a/library/core/src/intrinsics.rs
+++ b/library/core/src/intrinsics.rs
@@ -2209,9 +2209,9 @@ pub(crate) fn is_nonoverlapping<T>(src: *const T, dst: *const T, count: usize) -
 ///     dst.reserve(src_len);
 ///
 ///     unsafe {
-///         // The call to offset is always safe because `Vec` will never
+///         // The call to add is always safe because `Vec` will never
 ///         // allocate more than `isize::MAX` bytes.
-///         let dst_ptr = dst.as_mut_ptr().offset(dst_len as isize);
+///         let dst_ptr = dst.as_mut_ptr().add(dst_len);
 ///         let src_ptr = src.as_ptr();
 ///
 ///         // Truncate `src` without dropping its contents. We do this first,
diff --git a/library/core/src/sync/atomic.rs b/library/core/src/sync/atomic.rs
index 40ca9abd6bdc9..3c96290fc537e 100644
--- a/library/core/src/sync/atomic.rs
+++ b/library/core/src/sync/atomic.rs
@@ -1554,8 +1554,8 @@ impl<T> AtomicPtr<T> {
     /// Offsets the pointer's address by adding `val` *bytes*, returning the
     /// previous pointer.
     ///
-    /// This is equivalent to using [`wrapping_add`] and [`cast`] to atomically
-    /// perform `ptr = ptr.cast::<u8>().wrapping_add(val).cast::<T>()`.
+    /// This is equivalent to using [`wrapping_byte_add`] to atomically
+    /// perform `ptr = ptr.wrapping_byte_add(val)`.
     ///
     /// `fetch_byte_add` takes an [`Ordering`] argument which describes the
     /// memory ordering of this operation. All ordering modes are possible. Note
@@ -1565,8 +1565,7 @@ impl<T> AtomicPtr<T> {
     /// **Note**: This method is only available on platforms that support atomic
     /// operations on [`AtomicPtr`].
     ///
-    /// [`wrapping_add`]: pointer::wrapping_add
-    /// [`cast`]: pointer::cast
+    /// [`wrapping_byte_add`]: pointer::wrapping_byte_add
     ///
     /// # Examples
     ///
@@ -1591,8 +1590,8 @@ impl<T> AtomicPtr<T> {
     /// Offsets the pointer's address by subtracting `val` *bytes*, returning the
     /// previous pointer.
     ///
-    /// This is equivalent to using [`wrapping_sub`] and [`cast`] to atomically
-    /// perform `ptr = ptr.cast::<u8>().wrapping_sub(val).cast::<T>()`.
+    /// This is equivalent to using [`wrapping_byte_sub`] to atomically
+    /// perform `ptr = ptr.wrapping_byte_sub(val)`.
     ///
     /// `fetch_byte_sub` takes an [`Ordering`] argument which describes the
     /// memory ordering of this operation. All ordering modes are possible. Note
@@ -1602,8 +1601,7 @@ impl<T> AtomicPtr<T> {
     /// **Note**: This method is only available on platforms that support atomic
     /// operations on [`AtomicPtr`].
     ///
-    /// [`wrapping_sub`]: pointer::wrapping_sub
-    /// [`cast`]: pointer::cast
+    /// [`wrapping_byte_sub`]: pointer::wrapping_byte_sub
     ///
     /// # Examples
     ///