Optimize shallow_clone for Bytes::split_{off,to} (#92)

If `shallow_clone` is called with `&mut self`, and `Bytes` contains
`Vec`, then expensive CAS can be avoided, because no other thread
have references to this `Bytes` object.

Bench `split_off_and_drop` difference:

Before the diff:

```
test split_off_and_drop             ... bench:      91,858 ns/iter (+/- 17,401)
```

With the diff:

```
test split_off_and_drop             ... bench:      81,162 ns/iter (+/- 17,603)
```

benches/bytes.rs

@@ -29,6 +29,18 @@ fn alloc_big(b: &mut Bencher) {
     })
 }
 
+#[bench]
+fn split_off_and_drop(b: &mut Bencher) {
+    b.iter(|| {
+        for _ in 0..1024 {
+            let v = vec![10; 200];
+            let mut b = Bytes::from(v);
+            test::black_box(b.split_off(100));
+            test::black_box(b);
+        }
+    })
+}
+
 #[bench]
 fn deref_unique(b: &mut Bencher) {
     let mut buf = BytesMut::with_capacity(4096);