selftests/bpf: Don't hard-code root cgroup id

Commit 40430452 ("kernfs: use 64bit inos if ino_t is 64bit") changed
the way cgroup ids are exposed to the userspace. Instead of assuming
fixed root id, let's query it.

Fixes: 40430452 ("kernfs: use 64bit inos if ino_t is 64bit")
Signed-off-by: NStanislav Fomichev <sdf@google.com>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20191202200143.250793-1-sdf@google.com

tools/testing/selftests/bpf/test_skb_cgroup_id_user.c

@@ -120,7 +120,7 @@ int check_ancestor_cgroup_ids(int prog_id)
 	int err = 0;
 	int map_fd;
 
-	expected_ids[0] = 0x100000001;	/* root cgroup */
+	expected_ids[0] = get_cgroup_id("/..");	/* root cgroup */
 	expected_ids[1] = get_cgroup_id("");
 	expected_ids[2] = get_cgroup_id(CGROUP_PATH);
 	expected_ids[3] = 0; /* non-existent cgroup */