selftests/bpf: Add bpf_rr scheduler & test

This patch implements the round-robin BPF MPTCP scheduler, named bpf_rr,
which always picks the next available subflow to send data. If no such
next subflow available, picks the first one.

Using MPTCP_SCHED_TEST macro to add a new test for this bpf_rr
scheduler, the arguments "1 1" means data has been sent on both net
devices. Run this test by RUN_MPTCP_TEST macro.

Signed-off-by: Geliang Tang <[email protected]>
Reviewed-by: Mat Martineau <[email protected]>
Reviewed-by: Matthieu Baerts (NGI0) <[email protected]>

Author: Geliang Tang

tools/testing/selftests/bpf/prog_tests/mptcp.c

@@ -14,6 +14,7 @@
 #include "mptcp_bpf_iters.skel.h"
 #include "mptcp_bpf_first.skel.h"
 #include "mptcp_bpf_bkup.skel.h"
+#include "mptcp_bpf_rr.skel.h"
 
 #define NS_TEST "mptcp_ns"
 #define ADDR_1	"10.0.1.1"
@@ -714,6 +715,18 @@ static void test_bkup(void)
 	mptcp_bpf_bkup__destroy(skel);
 }
 
+static void test_rr(void)
+{
+	struct mptcp_bpf_rr *skel;
+
+	skel = mptcp_bpf_rr__open_and_load();
+	if (!ASSERT_OK_PTR(skel, "open_and_load: rr"))
+		return;
+
+	test_bpf_sched(skel->obj, "rr", WITH_DATA, WITH_DATA);
+	mptcp_bpf_rr__destroy(skel);
+}
+
 void test_mptcp(void)
 {
 	if (test__start_subtest("base"))
@@ -730,4 +743,6 @@ void test_mptcp(void)
 		test_first();
 	if (test__start_subtest("bkup"))
 		test_bkup();
+	if (test__start_subtest("rr"))
+		test_rr();
 }

tools/testing/selftests/bpf/progs/mptcp_bpf_rr.c

@@ -0,0 +1,78 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2022, SUSE. */
+
+#include "mptcp_bpf.h"
+#include <bpf/bpf_tracing.h>
+
+char _license[] SEC("license") = "GPL";
+
+struct mptcp_rr_storage {
+	struct sock *last_snd;
+};
+
+struct {
+	__uint(type, BPF_MAP_TYPE_SK_STORAGE);
+	__uint(map_flags, BPF_F_NO_PREALLOC);
+	__type(key, int);
+	__type(value, struct mptcp_rr_storage);
+} mptcp_rr_map SEC(".maps");
+
+SEC("struct_ops")
+void BPF_PROG(mptcp_sched_rr_init, struct mptcp_sock *msk)
+{
+	bpf_sk_storage_get(&mptcp_rr_map, msk, 0,
+			   BPF_LOCAL_STORAGE_GET_F_CREATE);
+}
+
+SEC("struct_ops")
+void BPF_PROG(mptcp_sched_rr_release, struct mptcp_sock *msk)
+{
+	bpf_sk_storage_delete(&mptcp_rr_map, msk);
+}
+
+SEC("struct_ops")
+int BPF_PROG(bpf_rr_get_subflow, struct mptcp_sock *msk,
+	     struct mptcp_sched_data *data)
+{
+	struct mptcp_subflow_context *subflow;
+	struct mptcp_rr_storage *ptr;
+	struct sock *last_snd = NULL;
+	int nr = 0;
+
+	ptr = bpf_sk_storage_get(&mptcp_rr_map, msk, 0,
+				 BPF_LOCAL_STORAGE_GET_F_CREATE);
+	if (!ptr)
+		return -1;
+
+	last_snd = ptr->last_snd;
+
+	for (int i = 0; i < data->subflows && i < MPTCP_SUBFLOWS_MAX; i++) {
+		subflow = bpf_mptcp_subflow_ctx_by_pos(data, i);
+		if (!last_snd || !subflow)
+			break;
+
+		if (mptcp_subflow_tcp_sock(subflow) == last_snd) {
+			if (i + 1 == MPTCP_SUBFLOWS_MAX ||
+			    !bpf_mptcp_subflow_ctx_by_pos(data, i + 1))
+				break;
+
+			nr = i + 1;
+			break;
+		}
+	}
+
+	subflow = bpf_mptcp_subflow_ctx_by_pos(data, nr);
+	if (!subflow)
+		return -1;
+	mptcp_subflow_set_scheduled(subflow, true);
+	ptr->last_snd = mptcp_subflow_tcp_sock(subflow);
+	return 0;
+}
+
+SEC(".struct_ops")
+struct mptcp_sched_ops rr = {
+	.init		= (void *)mptcp_sched_rr_init,
+	.release	= (void *)mptcp_sched_rr_release,
+	.get_subflow	= (void *)bpf_rr_get_subflow,
+	.name		= "bpf_rr",
+};