All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH mptcp-next v5 0/4] BPF round-robin scheduler
@ 2022-04-13 14:30 Geliang Tang
  2022-04-13 14:30 ` [PATCH mptcp-next v5 1/4] mptcp: add last_snd write access Geliang Tang
                   ` (3 more replies)
  0 siblings, 4 replies; 5+ messages in thread
From: Geliang Tang @ 2022-04-13 14:30 UTC (permalink / raw)
  To: mptcp; +Cc: Geliang Tang

v5:
 - update patch 2, use temporary storage instead.
 - update patch 3, use new helpers.
 - deponds on: "BPF packet scheduler" series v11.

v4:
 - add retrans argment for get_subflow()

v3:
 - add last_snd write access.
 - keep msk->last_snd setting in get_subflow().
 - deponds on: "BPF packet scheduler" series v10.

v2:
 - merge the squash-to patch.
 - implement bpf_mptcp_get_subflows helper, instead of
   bpf_mptcp_get_next_subflow.
 - deponds on: "BPF packet scheduler v9".

This patchset implements round-robin scheduler using BPF. Address to
some commends for the RFC version:

https://patchwork.kernel.org/project/mptcp/cover/cover.1631011068.git.geliangtang@xiaomi.com/

Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/75

Geliang Tang (4):
  mptcp: add last_snd write access
  mptcp: add bpf get_subflows helper
  selftests: bpf: add bpf_rr scheduler
  selftests: bpf: add bpf_rr test

 net/mptcp/bpf.c                               | 60 +++++++++++++++++++
 net/mptcp/protocol.h                          |  7 +++
 .../testing/selftests/bpf/bpf_mptcp_helpers.h | 13 ++++
 .../testing/selftests/bpf/prog_tests/mptcp.c  | 37 ++++++++++++
 .../selftests/bpf/progs/mptcp_bpf_rr.c        | 54 +++++++++++++++++
 5 files changed, 171 insertions(+)
 create mode 100644 tools/testing/selftests/bpf/progs/mptcp_bpf_rr.c

-- 
2.34.1


^ permalink raw reply	[flat|nested] 5+ messages in thread

* [PATCH mptcp-next v5 1/4] mptcp: add last_snd write access
  2022-04-13 14:30 [PATCH mptcp-next v5 0/4] BPF round-robin scheduler Geliang Tang
@ 2022-04-13 14:30 ` Geliang Tang
  2022-04-13 14:30 ` [PATCH mptcp-next v5 2/4] mptcp: add bpf get_subflows helper Geliang Tang
                   ` (2 subsequent siblings)
  3 siblings, 0 replies; 5+ messages in thread
From: Geliang Tang @ 2022-04-13 14:30 UTC (permalink / raw)
  To: mptcp; +Cc: Geliang Tang

This patch exports the member last_snd of struct mptcp_sock in
bpf_mptcp_helpers.h, and adds BPF write access to it.

Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
 net/mptcp/bpf.c                                 | 16 ++++++++++++++++
 tools/testing/selftests/bpf/bpf_mptcp_helpers.h |  1 +
 2 files changed, 17 insertions(+)

diff --git a/net/mptcp/bpf.c b/net/mptcp/bpf.c
index e849fc3fb6c5..bd3c50b07ab2 100644
--- a/net/mptcp/bpf.c
+++ b/net/mptcp/bpf.c
@@ -40,6 +40,7 @@ static int bpf_mptcp_sched_btf_struct_access(struct bpf_verifier_log *log,
 {
 	const struct btf_type *state;
 	u32 type_id;
+	size_t end;
 
 	if (atype == BPF_READ)
 		return btf_struct_access(log, btf, t, off, size, atype,
@@ -55,6 +56,21 @@ static int bpf_mptcp_sched_btf_struct_access(struct bpf_verifier_log *log,
 		return -EACCES;
 	}
 
+	switch (off) {
+	case offsetof(struct mptcp_sock, last_snd):
+		end = offsetofend(struct mptcp_sock, last_snd);
+		break;
+	default:
+		bpf_log(log, "no write support to mptcp_sock at off %d\n", off);
+		return -EACCES;
+	}
+
+	if (off + size > end) {
+		bpf_log(log, "access beyond mptcp_sock at off %u size %u ended at %lu",
+			off, size, end);
+		return -EACCES;
+	}
+
 	return NOT_INIT;
 }
 
diff --git a/tools/testing/selftests/bpf/bpf_mptcp_helpers.h b/tools/testing/selftests/bpf/bpf_mptcp_helpers.h
index 9a0e3e7766b0..4f82baee998c 100644
--- a/tools/testing/selftests/bpf/bpf_mptcp_helpers.h
+++ b/tools/testing/selftests/bpf/bpf_mptcp_helpers.h
@@ -9,6 +9,7 @@
 struct mptcp_sock {
 	struct inet_connection_sock	sk;
 
+	struct sock	*last_snd;
 	__u32		token;
 	struct sock	*first;
 	char		ca_name[TCP_CA_NAME_MAX];
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 5+ messages in thread

* [PATCH mptcp-next v5 2/4] mptcp: add bpf get_subflows helper
  2022-04-13 14:30 [PATCH mptcp-next v5 0/4] BPF round-robin scheduler Geliang Tang
  2022-04-13 14:30 ` [PATCH mptcp-next v5 1/4] mptcp: add last_snd write access Geliang Tang
@ 2022-04-13 14:30 ` Geliang Tang
  2022-04-13 14:30 ` [PATCH mptcp-next v5 3/4] selftests: bpf: add bpf_rr scheduler Geliang Tang
  2022-04-13 14:30 ` [PATCH mptcp-next v5 4/4] selftests: bpf: add bpf_rr test Geliang Tang
  3 siblings, 0 replies; 5+ messages in thread
From: Geliang Tang @ 2022-04-13 14:30 UTC (permalink / raw)
  To: mptcp; +Cc: Geliang Tang

This patch implements a new helper bpf_mptcp_get_subflows() to get all the
subflows of the given mptcp_sock, it returns the number of suflows. Add
a new member subflows in struct mptcp_sock as a pointers array of all the
subflows.

Register this helper in bpf_mptcp_sched_kfunc_init() to make sure it can
be accessed from the BPF context.

Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
 net/mptcp/bpf.c      | 44 ++++++++++++++++++++++++++++++++++++++++++++
 net/mptcp/protocol.h |  7 +++++++
 2 files changed, 51 insertions(+)

diff --git a/net/mptcp/bpf.c b/net/mptcp/bpf.c
index bd3c50b07ab2..823e5af1a2f4 100644
--- a/net/mptcp/bpf.c
+++ b/net/mptcp/bpf.c
@@ -157,6 +157,23 @@ struct bpf_struct_ops bpf_mptcp_sched_ops = {
 	.name		= "mptcp_sched_ops",
 };
 
+BTF_SET_START(bpf_mptcp_sched_kfunc_ids)
+BTF_ID(func, bpf_mptcp_get_subflows_array)
+BTF_ID(func, bpf_mptcp_put_subflows_array)
+BTF_SET_END(bpf_mptcp_sched_kfunc_ids)
+
+static const struct btf_kfunc_id_set bpf_mptcp_sched_kfunc_set = {
+	.owner		= THIS_MODULE,
+	.check_set	= &bpf_mptcp_sched_kfunc_ids,
+};
+
+static int __init bpf_mptcp_sched_kfunc_init(void)
+{
+	return register_btf_kfunc_id_set(BPF_PROG_TYPE_STRUCT_OPS,
+					 &bpf_mptcp_sched_kfunc_set);
+}
+late_initcall(bpf_mptcp_sched_kfunc_init);
+
 struct mptcp_sock *bpf_mptcp_sock_from_subflow(struct sock *sk)
 {
 	if (sk && sk_fullsock(sk) && sk->sk_protocol == IPPROTO_TCP && sk_is_mptcp(sk))
@@ -165,3 +182,30 @@ struct mptcp_sock *bpf_mptcp_sock_from_subflow(struct sock *sk)
 	return NULL;
 }
 EXPORT_SYMBOL(bpf_mptcp_sock_from_subflow);
+
+struct mptcp_subflows_array *bpf_mptcp_get_subflows_array(struct mptcp_sock *msk)
+{
+	struct mptcp_subflow_context *subflow;
+	struct mptcp_subflows_array *array;
+
+	array = kzalloc(sizeof(*array), GFP_KERNEL);
+	if (!array)
+		return array;
+
+	mptcp_for_each_subflow(msk, subflow)
+		array->subflows[array->nr++] = subflow;
+
+	return array;
+}
+EXPORT_SYMBOL(bpf_mptcp_get_subflows_array);
+
+void bpf_mptcp_put_subflows_array(struct mptcp_subflows_array *array)
+{
+	int i;
+
+	for (i = 0; i < MPTCP_SUBFLOWS_MAX; i++)
+		array->subflows[i] = NULL;
+	array->nr = 0;
+	kfree(array);
+}
+EXPORT_SYMBOL(bpf_mptcp_put_subflows_array);
diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h
index a58c5c84e72b..5581b37028ee 100644
--- a/net/mptcp/protocol.h
+++ b/net/mptcp/protocol.h
@@ -490,6 +490,13 @@ struct mptcp_subflow_context {
 	struct	rcu_head rcu;
 };
 
+#define MPTCP_SUBFLOWS_MAX	8
+
+struct mptcp_subflows_array {
+	u8 nr;
+	struct mptcp_subflow_context *subflows[MPTCP_SUBFLOWS_MAX];
+};
+
 static inline struct mptcp_subflow_context *
 mptcp_subflow_ctx(const struct sock *sk)
 {
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 5+ messages in thread

* [PATCH mptcp-next v5 3/4] selftests: bpf: add bpf_rr scheduler
  2022-04-13 14:30 [PATCH mptcp-next v5 0/4] BPF round-robin scheduler Geliang Tang
  2022-04-13 14:30 ` [PATCH mptcp-next v5 1/4] mptcp: add last_snd write access Geliang Tang
  2022-04-13 14:30 ` [PATCH mptcp-next v5 2/4] mptcp: add bpf get_subflows helper Geliang Tang
@ 2022-04-13 14:30 ` Geliang Tang
  2022-04-13 14:30 ` [PATCH mptcp-next v5 4/4] selftests: bpf: add bpf_rr test Geliang Tang
  3 siblings, 0 replies; 5+ messages in thread
From: Geliang Tang @ 2022-04-13 14:30 UTC (permalink / raw)
  To: mptcp; +Cc: Geliang Tang

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.

Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
 .../testing/selftests/bpf/bpf_mptcp_helpers.h | 12 +++++
 .../selftests/bpf/progs/mptcp_bpf_rr.c        | 54 +++++++++++++++++++
 2 files changed, 66 insertions(+)
 create mode 100644 tools/testing/selftests/bpf/progs/mptcp_bpf_rr.c

diff --git a/tools/testing/selftests/bpf/bpf_mptcp_helpers.h b/tools/testing/selftests/bpf/bpf_mptcp_helpers.h
index 4f82baee998c..140ead91e2c6 100644
--- a/tools/testing/selftests/bpf/bpf_mptcp_helpers.h
+++ b/tools/testing/selftests/bpf/bpf_mptcp_helpers.h
@@ -27,4 +27,16 @@ struct mptcp_sched_ops {
 	void *owner;
 };
 
+struct mptcp_subflow_context {
+	__u32	token;
+	struct	sock *tcp_sock;	    /* tcp sk backpointer */
+} __attribute__((preserve_access_index));
+
+#define MPTCP_SUBFLOWS_MAX	8
+
+struct mptcp_subflows_array {
+	__u8 nr;
+	struct mptcp_subflow_context *subflows[MPTCP_SUBFLOWS_MAX];
+};
+
 #endif
diff --git a/tools/testing/selftests/bpf/progs/mptcp_bpf_rr.c b/tools/testing/selftests/bpf/progs/mptcp_bpf_rr.c
new file mode 100644
index 000000000000..531ab12a93b1
--- /dev/null
+++ b/tools/testing/selftests/bpf/progs/mptcp_bpf_rr.c
@@ -0,0 +1,54 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2022, SUSE. */
+
+#include <linux/bpf.h>
+#include <linux/stddef.h>
+#include <linux/tcp.h>
+#include "bpf_mptcp_helpers.h"
+
+char _license[] SEC("license") = "GPL";
+
+extern struct mptcp_subflows_array *
+bpf_mptcp_get_subflows_array(struct mptcp_sock *msk) __ksym;
+extern void bpf_mptcp_put_subflows_array(struct mptcp_subflows_array *array) __ksym;
+
+SEC("struct_ops/mptcp_sched_rr_init")
+void BPF_PROG(mptcp_sched_rr_init, struct mptcp_sock *msk)
+{
+}
+
+SEC("struct_ops/mptcp_sched_rr_release")
+void BPF_PROG(mptcp_sched_rr_release, struct mptcp_sock *msk)
+{
+}
+
+struct sock *BPF_STRUCT_OPS(bpf_rr_get_subflow, struct mptcp_sock *msk, bool retrans)
+{
+	struct mptcp_subflow_context *subflow;
+	struct mptcp_subflows_array *array;
+	struct sock *ssk = msk->first;
+
+	array = bpf_mptcp_get_subflows_array(msk);
+	for (int i = 0; i < MPTCP_SUBFLOWS_MAX; i++) {
+		if (i == array->nr)
+			break;
+
+		subflow = array->subflows[i];
+		if (subflow->tcp_sock != msk->last_snd) {
+			ssk = subflow->tcp_sock;
+			break;
+		}
+	}
+	bpf_mptcp_put_subflows_array(array);
+
+	msk->last_snd = ssk;
+	return ssk;
+}
+
+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",
+};
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 5+ messages in thread

* [PATCH mptcp-next v5 4/4] selftests: bpf: add bpf_rr test
  2022-04-13 14:30 [PATCH mptcp-next v5 0/4] BPF round-robin scheduler Geliang Tang
                   ` (2 preceding siblings ...)
  2022-04-13 14:30 ` [PATCH mptcp-next v5 3/4] selftests: bpf: add bpf_rr scheduler Geliang Tang
@ 2022-04-13 14:30 ` Geliang Tang
  3 siblings, 0 replies; 5+ messages in thread
From: Geliang Tang @ 2022-04-13 14:30 UTC (permalink / raw)
  To: mptcp; +Cc: Geliang Tang

This patch adds the round-robin BPF MPTCP scheduler test. Use sysctl to
set net.mptcp.scheduler to use this sched. Add a veth net device to
simulate the multiple addresses case. Use 'ip mptcp endpoint' command to
add this new endpoint to PM netlink.

Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
 .../testing/selftests/bpf/prog_tests/mptcp.c  | 37 +++++++++++++++++++
 1 file changed, 37 insertions(+)

diff --git a/tools/testing/selftests/bpf/prog_tests/mptcp.c b/tools/testing/selftests/bpf/prog_tests/mptcp.c
index 01679077521d..20f382da6812 100644
--- a/tools/testing/selftests/bpf/prog_tests/mptcp.c
+++ b/tools/testing/selftests/bpf/prog_tests/mptcp.c
@@ -6,6 +6,7 @@
 #include "cgroup_helpers.h"
 #include "network_helpers.h"
 #include "mptcp_bpf_first.skel.h"
+#include "mptcp_bpf_rr.skel.h"
 
 #ifndef TCP_CA_NAME_MAX
 #define TCP_CA_NAME_MAX	16
@@ -362,10 +363,46 @@ static void test_first(void)
 	mptcp_bpf_first__destroy(first_skel);
 }
 
+static void test_rr(void)
+{
+	struct mptcp_bpf_rr *rr_skel;
+	int server_fd, client_fd;
+	struct bpf_link *link;
+
+	rr_skel = mptcp_bpf_rr__open_and_load();
+	if (CHECK(!rr_skel, "bpf_rr__open_and_load", "failed\n"))
+		return;
+
+	link = bpf_map__attach_struct_ops(rr_skel->maps.rr);
+	if (!ASSERT_OK_PTR(link, "bpf_map__attach_struct_ops")) {
+		mptcp_bpf_rr__destroy(rr_skel);
+		return;
+	}
+
+	system("ip link add veth1 type veth; \
+		ip addr add 10.0.1.1/24 dev veth1; \
+		ip link set veth1 up");
+	system("ip mptcp endpoint add 10.0.1.1 subflow");
+	system("sysctl -q net.mptcp.scheduler=bpf_rr");
+	server_fd = start_mptcp_server(AF_INET, NULL, 0, 0);
+	client_fd = connect_to_mptcp_fd(server_fd, 0);
+
+	send_data(server_fd, client_fd);
+
+	close(client_fd);
+	close(server_fd);
+	system("ip mptcp endpoint flush");
+	system("ip link del veth1");
+	bpf_link__destroy(link);
+	mptcp_bpf_rr__destroy(rr_skel);
+}
+
 void test_mptcp(void)
 {
 	if (test__start_subtest("base"))
 		test_base();
 	if (test__start_subtest("first"))
 		test_first();
+	if (test__start_subtest("rr"))
+		test_rr();
 }
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2022-04-13 14:31 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-04-13 14:30 [PATCH mptcp-next v5 0/4] BPF round-robin scheduler Geliang Tang
2022-04-13 14:30 ` [PATCH mptcp-next v5 1/4] mptcp: add last_snd write access Geliang Tang
2022-04-13 14:30 ` [PATCH mptcp-next v5 2/4] mptcp: add bpf get_subflows helper Geliang Tang
2022-04-13 14:30 ` [PATCH mptcp-next v5 3/4] selftests: bpf: add bpf_rr scheduler Geliang Tang
2022-04-13 14:30 ` [PATCH mptcp-next v5 4/4] selftests: bpf: add bpf_rr test Geliang Tang

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.