Run mptcp sched test in a dedicated netns.

Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
 tools/testing/selftests/bpf/prog_tests/mptcp.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/tools/testing/selftests/bpf/prog_tests/mptcp.c b/tools/testing/selftests/bpf/prog_tests/mptcp.c
index 210ab5c30962..88bef3dff24e 100644
--- a/tools/testing/selftests/bpf/prog_tests/mptcp.c
+++ b/tools/testing/selftests/bpf/prog_tests/mptcp.c
@@ -351,6 +351,7 @@ static void test_first(void)
 static void test_bkup(void)
 {
 	struct mptcp_bpf_bkup *bkup_skel;
+	struct nstoken *nstoken = NULL;
 	int server_fd, client_fd;
 	struct bpf_link *link;
 
@@ -364,7 +365,9 @@ static void test_bkup(void)
 		return;
 	}
 
-	sched_init("subflow backup", "bpf_bkup");
+	nstoken = sched_init("subflow backup", "bpf_bkup");
+	if (!ASSERT_OK_PTR(nstoken, "sched_init:bpf_bkup"))
+		goto fail;
 	server_fd = start_mptcp_server(AF_INET, ADDR_1, 0, 0);
 	client_fd = connect_to_fd(server_fd, 0);
 
@@ -374,7 +377,8 @@ static void test_bkup(void)
 
 	close(client_fd);
 	close(server_fd);
-	sched_cleanup();
+fail:
+	cleanup_netns(nstoken);
 	bpf_link__destroy(link);
 	mptcp_bpf_bkup__destroy(bkup_skel);
 }
-- 
2.35.3