Here's one case this lemma might be useful. Suppose we want to
programmatically generate namespaces for e.g. locks:

```
Definition lockN (l : loc) := nroot .@ "lock" .@ l.
```

Then to know that two such namespaces are disjoint, we need to know
that the corresponding locations are distinct. For that we use
the lemma here introduced.

```
Lemma ne l1 l2 v1 v2 :
  l1 ↦ v1 -∗ l2 ↦ v2 -∗ ⌜l1 ≠ l2⌝.
Proof. iApply mapsto_mapsto_ne. (* goal ¬ ✓ 2%Qp *) by intros []. Qed.
```