Skip to contentSkip to conversation

BPF packet scheduler

PATCH Superseded Geliang Tang 1 participant 6 messages Mar 17, 2022

Subsystems (6)

v1 of 18 Superseded Latest is v18 →

lore
Share & tools
lore
#0 lore raw

cover BPF packet scheduler

Addressed to the commends in the RFC version:
https://patchwork.kernel.org/project/mptcp/cover/[email protected]/

Depends on the bpf patches queued in the patchwork:

7541fe09cf06 selftests: bpf: add mptcp sched test
1fa592f2770e selftests: bpf: implement bpf_first scheduler 5bbd1d25b994 mptcp: add bpf_mptcp_sched_ops
ade0194842d1 mptcp: add struct mptcp_sched_ops
23bdbd0af9f4 mptcp: add a new sysctl scheduler
772471a34280 selftests: bpf: verify first subflow of mptcp_sock f3484ecf1bce Squash to "selftests: bpf: verify ca_name of struct mptcp_sock" 0cd058c7c489 selftests: bpf: verify ca_name of struct mptcp_sock 07a66e354126 Squash to "selftests: bpf: test bpf_skc_to_mptcp_sock" 168ee1d1a118 selftests: bpf: test bpf_skc_to_mptcp_sock
7a773e6af914 Squash to "selftests: bpf: add MPTCP test base" a1933407b912 bpf: add bpf_skc_to_mptcp_sock_proto
0095d8ca9a67 Revert "bpf: add 'bpf_mptcp_sock' structure and helper" 2856d8c8020a Revert "selftests: bpf: add bpf_mptcp_sock() verifier tests" df0e2a4c394f (origin/export, origin/HEAD) DO-NOT-MERGE: mptcp: enabled by default

Geliang Tang (5):
mptcp: add a new sysctl scheduler
mptcp: add struct mptcp_sched_ops
mptcp: add bpf_mptcp_sched_ops
selftests: bpf: implement bpf_first scheduler
selftests: bpf: add mptcp sched test

 Documentation/networking/mptcp-sysctl.rst     |   8 +
 include/net/mptcp.h                           |  15 ++
 kernel/bpf/bpf_struct_ops_types.h             |   4 +
 net/mptcp/Makefile                            |   2 +-
 net/mptcp/bpf.c                               | 102 +++++++++++
 net/mptcp/ctrl.c                              |  14 ++
 net/mptcp/protocol.c                          |  12 +-
 net/mptcp/protocol.h                          |   9 +
 net/mptcp/sched.c                             |  89 ++++++++++
 tools/testing/selftests/bpf/bpf_tcp_helpers.h |  14 ++
 .../bpf/prog_tests/bpf_mptcp_sched.c          | 158 ++++++++++++++++++
 tools/testing/selftests/bpf/progs/bpf_first.c |  30 ++++
 12 files changed, 452 insertions(+), 5 deletions(-)
 create mode 100644 net/mptcp/sched.c
 create mode 100644 tools/testing/selftests/bpf/prog_tests/bpf_mptcp_sched.c
 create mode 100644 tools/testing/selftests/bpf/progs/bpf_first.c
Signature · 1 line
2.34.1
#1 lore raw

PATCH 1/5 mptcp: add a new sysctl scheduler

This patch added a new sysctl, named scheduler, to support for selection of different schedulers.

---
 Documentation/networking/mptcp-sysctl.rst |  8 ++++++++
 include/net/mptcp.h                       |  2 ++
 net/mptcp/ctrl.c                          | 14 ++++++++++++++
 net/mptcp/protocol.h                      |  1 +
 4 files changed, 25 insertions(+)
Documentation/networking/mptcp-sysctl.rst +8 -0
index e263dfcc4b40..d9e69fdc7ea3 100644
@@ -75,3 +75,11 @@ stale_loss_cnt - INTEGER
 	This is a per-namespace sysctl.
 
 	Default: 4
+
+scheduler - STRING
+	Select the scheduler of your choice.
+
+	Support for selection of different schedulers. This is a per-namespace
+	sysctl.
+
+	Default: "default"
include/net/mptcp.h +2 -0
index 8b1afd6f5cc4..2dde95412fc3 100644
@@ -284,4 +284,6 @@ static inline int mptcpv6_init(void) { return 0; }
 static inline void mptcpv6_handle_mapped(struct sock *sk, bool mapped) { }
 #endif
 
+#define MPTCP_SCHED_NAME_MAX 16
+
 #endif /* __NET_MPTCP_H */
net/mptcp/ctrl.c +14 -0
index ae20b7d92e28..4e0811969d8c 100644
@@ -32,6 +32,7 @@ struct mptcp_pernet {
 	u8 checksum_enabled;
 	u8 allow_join_initial_addr_port;
 	u8 pm_type;
+	char scheduler[MPTCP_SCHED_NAME_MAX];
 };
 
 static struct mptcp_pernet *mptcp_get_pernet(const struct net *net)
@@ -69,6 +70,11 @@ int mptcp_get_pm_type(const struct net *net)
 	return mptcp_get_pernet(net)->pm_type;
 }
 
+char *mptcp_get_scheduler(struct net *net)
+{
+	return mptcp_get_pernet(net)->scheduler;
+}
+
 static void mptcp_pernet_set_defaults(struct mptcp_pernet *pernet)
 {
 	pernet->mptcp_enabled = 1;
@@ -77,6 +83,7 @@ static void mptcp_pernet_set_defaults(struct mptcp_pernet *pernet)
 	pernet->allow_join_initial_addr_port = 1;
 	pernet->stale_loss_cnt = 4;
 	pernet->pm_type = MPTCP_PM_TYPE_KERNEL;
+	strcpy(pernet->scheduler, "default");
 }
 
 #ifdef CONFIG_SYSCTL
@@ -128,6 +135,12 @@ static struct ctl_table mptcp_sysctl_table[] = {
 		.extra1       = SYSCTL_ZERO,
 		.extra2       = &mptcp_pm_type_max
 	},
+	{
+		.procname = "scheduler",
+		.maxlen	= MPTCP_SCHED_NAME_MAX,
+		.mode = 0644,
+		.proc_handler = proc_dostring,
+	},
 	{}
 };
 
@@ -149,6 +162,7 @@ static int mptcp_pernet_new_table(struct net *net, struct mptcp_pernet *pernet)
 	table[3].data = &pernet->allow_join_initial_addr_port;
 	table[4].data = &pernet->stale_loss_cnt;
 	table[5].data = &pernet->pm_type;
+	table[6].data = &pernet->scheduler;
 
 	hdr = register_net_sysctl(net, MPTCP_SYSCTL_PATH, table);
 	if (!hdr)
net/mptcp/protocol.h +1 -0
index c8bada4537e2..c36b9c40c314 100644
@@ -583,6 +583,7 @@ int mptcp_is_checksum_enabled(const struct net *net);
 int mptcp_allow_join_id0(const struct net *net);
 unsigned int mptcp_stale_loss_cnt(const struct net *net);
 int mptcp_get_pm_type(const struct net *net);
+char *mptcp_get_scheduler(struct net *net);
 void mptcp_subflow_fully_established(struct mptcp_subflow_context *subflow,
 				     struct mptcp_options_received *mp_opt);
 bool __mptcp_retransmit_pending_data(struct sock *sk);
Signature · 1 line
2.34.1
#2 lore raw

PATCH 2/5 mptcp: add struct mptcp_sched_ops

In reply to Geliang Tang

This patch added struct mptcp_sched_ops. And define the scheduler init, register and find functions.

---
 include/net/mptcp.h  | 13 +++++++
 net/mptcp/Makefile   |  2 +-
 net/mptcp/protocol.c | 12 ++++--
 net/mptcp/protocol.h |  8 ++++
 net/mptcp/sched.c    | 89 ++++++++++++++++++++++++++++++++++++++++++++
 5 files changed, 119 insertions(+), 5 deletions(-)
 create mode 100644 net/mptcp/sched.c
include/net/mptcp.h +13 -0
index 2dde95412fc3..bf7f4ba00424 100644
@@ -286,4 +286,17 @@ static inline void mptcpv6_handle_mapped(struct sock *sk, bool mapped) { }
 
 #define MPTCP_SCHED_NAME_MAX 16
 
+struct mptcp_sched_ops {
+	struct sock *	(*get_subflow)(struct mptcp_sock *msk);
+
+	char			name[MPTCP_SCHED_NAME_MAX];
+	struct module		*owner;
+	struct list_head	list;
+
+	/* initialize private data (optional) */
+	void (*init)(struct sock *sk);
+	/* cleanup private data  (optional) */
+	void (*release)(struct sock *sk);
+} ____cacheline_aligned_in_smp;
+
 #endif /* __NET_MPTCP_H */
net/mptcp/Makefile +1 -1
index 0a0608b6b4b4..aa5c10d1b80a 100644
@@ -3,7 +3,7 @@ obj-$(CONFIG_MPTCP) += mptcp.o
 ccflags-y += -DDEBUG
 
 mptcp-y := protocol.o subflow.o options.o token.o crypto.o ctrl.o pm.o diag.o \
-	   mib.o pm_netlink.o sockopt.o
+	   mib.o pm_netlink.o sockopt.o sched.o
 
 obj-$(CONFIG_SYN_COOKIES) += syncookies.o
 obj-$(CONFIG_INET_MPTCP_DIAG) += mptcp_diag.o
net/mptcp/protocol.c +8 -4
index fbb14dfe62b3..2250b1c11e9b 100644
@@ -1424,7 +1424,7 @@ bool mptcp_subflow_active(struct mptcp_subflow_context *subflow)
  * returns the subflow that will transmit the next DSS
  * additionally updates the rtx timeout
  */
-static struct sock *mptcp_subflow_get_send(struct mptcp_sock *msk)
+struct sock *mptcp_subflow_get_send(struct mptcp_sock *msk)
 {
 	struct subflow_send_info send_info[SSK_MODE_MAX];
 	struct mptcp_subflow_context *subflow;
@@ -1567,7 +1567,7 @@ void __mptcp_push_pending(struct sock *sk, unsigned int flags)
 			int ret = 0;
 
 			prev_ssk = ssk;
-			ssk = mptcp_subflow_get_send(msk);
+			ssk = mptcp_get_subflow(msk);
 
 			/* First check. If the ssk has changed since
 			 * the last round, release prev_ssk
@@ -1634,7 +1634,7 @@ static void __mptcp_subflow_push_pending(struct sock *sk, struct sock *ssk)
 			 * check for a different subflow usage only after
 			 * spooling the first chunk of data
 			 */
-			xmit_ssk = first ? ssk : mptcp_subflow_get_send(mptcp_sk(sk));
+			xmit_ssk = first ? ssk : mptcp_get_subflow(mptcp_sk(sk));
 			if (!xmit_ssk)
 				goto out;
 			if (xmit_ssk != ssk) {
@@ -2609,6 +2609,7 @@ static int mptcp_init_sock(struct sock *sk)
 	 * propagate the correct value
 	 */
 	mptcp_ca_reset(sk);
+	mptcp_sched_data_init(sk);
 
 	sk_sockets_allocated_inc(sk);
 	sk->sk_rcvbuf = sock_net(sk)->ipv4.sysctl_tcp_rmem[1];
@@ -2760,6 +2761,7 @@ static void __mptcp_destroy_sock(struct sock *sk)
 	sk_stop_timer(sk, &msk->sk.icsk_retransmit_timer);
 	sk_stop_timer(sk, &sk->sk_timer);
 	msk->pm.status = 0;
+	msk->sched = NULL;
 
 	/* clears msk->subflow, allowing the following loop to close
 	 * even the initial subflow
@@ -2933,6 +2935,7 @@ struct sock *mptcp_sk_clone(const struct sock *sk,
 	msk->snd_una = msk->write_seq;
 	msk->wnd_end = msk->snd_nxt + req->rsk_rcv_wnd;
 	msk->setsockopt_seq = mptcp_sk(sk)->setsockopt_seq;
+	msk->sched = mptcp_sk(sk)->sched;
 
 	if (mp_opt->suboptions & OPTIONS_MPTCP_MPC) {
 		msk->can_ack = true;
@@ -3070,7 +3073,7 @@ void __mptcp_check_push(struct sock *sk, struct sock *ssk)
 		return;
 
 	if (!sock_owned_by_user(sk)) {
-		struct sock *xmit_ssk = mptcp_subflow_get_send(mptcp_sk(sk));
+		struct sock *xmit_ssk = mptcp_get_subflow(mptcp_sk(sk));
 
 		if (xmit_ssk == ssk)
 			__mptcp_subflow_push_pending(sk, ssk);
@@ -3743,6 +3746,7 @@ void __init mptcp_proto_init(void)
 
 	mptcp_subflow_init();
 	mptcp_pm_init();
+	mptcp_sched_init();
 	mptcp_token_init();
 
 	if (proto_register(&mptcp_prot, MPTCP_USE_SLAB) != 0)
net/mptcp/protocol.h +8 -0
index c36b9c40c314..33720856bfed 100644
@@ -286,6 +286,7 @@ struct mptcp_sock {
 	struct socket	*subflow; /* outgoing connect/listener/!mp_capable */
 	struct sock	*first;
 	struct mptcp_pm_data	pm;
+	struct mptcp_sched_ops	*sched;
 	struct {
 		u32	space;	/* bytes copied in last measurement window */
 		u32	copied; /* bytes copied in this measurement window */
@@ -584,6 +585,13 @@ int mptcp_allow_join_id0(const struct net *net);
 unsigned int mptcp_stale_loss_cnt(const struct net *net);
 int mptcp_get_pm_type(const struct net *net);
 char *mptcp_get_scheduler(struct net *net);
+int mptcp_register_scheduler(struct mptcp_sched_ops *sched);
+void mptcp_unregister_scheduler(struct mptcp_sched_ops *sched);
+struct mptcp_sched_ops *mptcp_sched_find(const char *name);
+void mptcp_sched_data_init(struct sock *sk);
+void mptcp_sched_init(void);
+struct sock *mptcp_get_subflow(struct mptcp_sock *msk);
+struct sock *mptcp_subflow_get_send(struct mptcp_sock *msk);
 void mptcp_subflow_fully_established(struct mptcp_subflow_context *subflow,
 				     struct mptcp_options_received *mp_opt);
 bool __mptcp_retransmit_pending_data(struct sock *sk);
net/mptcp/sched.c new file +89 -0

89 lines not on this page · raw message

Signature · 1 line
2.34.1
#3 lore raw

PATCH 3/5 mptcp: add bpf_mptcp_sched_ops

In reply to Geliang Tang

This patch implemented a new struct bpf_struct_ops, bpf_mptcp_sched_ops. Register and unregister the mptcp scheduler in .reg and .unreg.

---
 kernel/bpf/bpf_struct_ops_types.h |   4 ++
 net/mptcp/bpf.c                   | 102 ++++++++++++++++++++++++++++++
 2 files changed, 106 insertions(+)
kernel/bpf/bpf_struct_ops_types.h +4 -0
index 5678a9ddf817..5a6b0c0d8d3d 100644
@@ -8,5 +8,9 @@ BPF_STRUCT_OPS_TYPE(bpf_dummy_ops)
 #ifdef CONFIG_INET
 #include <net/tcp.h>
 BPF_STRUCT_OPS_TYPE(tcp_congestion_ops)
+#ifdef CONFIG_MPTCP
+#include <net/mptcp.h>
+BPF_STRUCT_OPS_TYPE(mptcp_sched_ops)
+#endif
 #endif
 #endif
net/mptcp/bpf.c +102 -0

110 lines not on this page · raw message

Signature · 1 line
2.34.1
#4 lore raw

PATCH 4/5 selftests: bpf: implement bpf_first scheduler

In reply to Geliang Tang

This patch implemented the simplest MPTCP scheduler, named bpf_first, which always picks the first subflow to send data.

---
 tools/testing/selftests/bpf/bpf_tcp_helpers.h | 14 +++++++++
 tools/testing/selftests/bpf/progs/bpf_first.c | 30 +++++++++++++++++++
 2 files changed, 44 insertions(+)
 create mode 100644 tools/testing/selftests/bpf/progs/bpf_first.c
tools/testing/selftests/bpf/bpf_tcp_helpers.h +14 -0
index f92357597e63..886d6fa100ce 100644
@@ -197,6 +197,20 @@ struct tcp_congestion_ops {
 	void *owner;
 };
 
+#define MPTCP_SCHED_NAME_MAX 16
+
+struct mptcp_sched_ops {
+	char name[MPTCP_SCHED_NAME_MAX];
+
+	/* initialize private data (optional) */
+	void (*init)(struct sock *sk);
+	/* cleanup private data  (optional) */
+	void (*release)(struct sock *sk);
+
+	struct sock *(*get_subflow)(struct sock *sk);
+	void *owner;
+};
+
 #define min(a, b) ((a) < (b) ? (a) : (b))
 #define max(a, b) ((a) > (b) ? (a) : (b))
 #define min_not_zero(x, y) ({			\
tools/testing/selftests/bpf/progs/bpf_first.c new file +30 -0
new file mode 100644
index 000000000000..c30a7cd3b932
@@ -0,0 +1,30 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* Copyright (c) 2022, SUSE. */
+
+#include <linux/bpf.h>
+#include "bpf_tcp_helpers.h"
+
+char _license[] SEC("license") = "GPL";
+
+char fmt[] = "%s";
+
+/* "struct_ops/" prefix is a requirement */
+SEC("struct_ops/mptcp_sched_first_init")
+void BPF_PROG(mptcp_sched_first_init, struct sock *sk)
+{
+	bpf_trace_printk(fmt, sizeof(fmt), __func__);
+}
+
+struct sock *BPF_STRUCT_OPS(bpf_first_get_subflow, struct sock *sk)
+{
+	struct mptcp_sock *msk = (struct mptcp_sock *)sk;
+
+	return msk->first;
+}
+
+SEC(".struct_ops")
+struct mptcp_sched_ops first = {
+	.init 		= (void *)mptcp_sched_first_init,
+	.get_subflow	= (void *)bpf_first_get_subflow,
+	.name		= "bpf_first",
+};
Signature · 1 line
2.34.1
#5 lore raw

PATCH 5/5 selftests: bpf: add mptcp sched test

In reply to Geliang Tang

This patch added the MPTCP sched testcases. Use sysctl to set net.mptcp.scheduler in the tests.

---
 .../bpf/prog_tests/bpf_mptcp_sched.c          | 158 ++++++++++++++++++
 1 file changed, 158 insertions(+)
 create mode 100644 tools/testing/selftests/bpf/prog_tests/bpf_mptcp_sched.c
tools/testing/selftests/bpf/prog_tests/bpf_mptcp_sched.c new file +158 -0
new file mode 100644
index 000000000000..b351ae260903
@@ -0,0 +1,158 @@
+// SPDX-License-Identifier: GPL-2.0
+
+#include <linux/err.h>
+#include <netinet/tcp.h>
+#include <test_progs.h>
+#include "network_helpers.h"
+#include "bpf_first.skel.h"
+
+#ifndef IPPROTO_MPTCP
+#define IPPROTO_MPTCP 262
+#endif
+
+#define min(a, b) ((a) < (b) ? (a) : (b))
+
+static const unsigned int total_bytes = 10 * 1024 * 1024;
+static int stop, duration;
+
+static void *server(void *arg)
+{
+	int lfd = (int)(long)arg, err = 0, fd;
+	ssize_t nr_sent = 0, bytes = 0;
+	char batch[1500];
+
+	fd = accept(lfd, NULL, NULL);
+	while (fd == -1) {
+		if (errno == EINTR)
+			continue;
+		err = -errno;
+		goto done;
+	}
+
+	if (settimeo(fd, 0)) {
+		err = -errno;
+		goto done;
+	}
+
+	while (bytes < total_bytes && !READ_ONCE(stop)) {
+		nr_sent = send(fd, &batch,
+			       min(total_bytes - bytes, sizeof(batch)), 0);
+		if (nr_sent == -1 && errno == EINTR)
+			continue;
+		if (nr_sent == -1) {
+			err = -errno;
+			break;
+		}
+		bytes += nr_sent;
+	}
+
+	CHECK(bytes != total_bytes, "send", "%zd != %u nr_sent:%zd errno:%d\n",
+	      bytes, total_bytes, nr_sent, errno);
+
+done:
+	if (fd >= 0)
+		close(fd);
+	if (err) {
+		WRITE_ONCE(stop, 1);
+		return ERR_PTR(err);
+	}
+	return NULL;
+}
+
+static void do_test(const char *mptcp_sched)
+{
+	struct sockaddr_in6 sa6 = {};
+	ssize_t nr_recv = 0, bytes = 0;
+	int lfd = -1, fd = -1;
+	pthread_t srv_thread;
+	socklen_t addrlen = sizeof(sa6);
+	void *thread_ret;
+	char batch[1500];
+	char cmd[256];
+	int err;
+
+	WRITE_ONCE(stop, 0);
+
+	snprintf(cmd, sizeof(cmd), "sysctl -q net.mptcp.scheduler=%s", mptcp_sched);
+	system(cmd);
+
+	lfd = socket(AF_INET6, SOCK_STREAM, IPPROTO_MPTCP);
+	if (CHECK(lfd == -1, "socket", "errno:%d\n", errno))
+		return;
+	fd = socket(AF_INET6, SOCK_STREAM, IPPROTO_MPTCP);
+	if (CHECK(fd == -1, "socket", "errno:%d\n", errno)) {
+		close(lfd);
+		return;
+	}
+
+	/* bind, listen and start server thread to accept */
+	sa6.sin6_family = AF_INET6;
+	sa6.sin6_addr = in6addr_loopback;
+	err = bind(lfd, (struct sockaddr *)&sa6, addrlen);
+	if (CHECK(err == -1, "bind", "errno:%d\n", errno))
+		goto done;
+	err = getsockname(lfd, (struct sockaddr *)&sa6, &addrlen);
+	if (CHECK(err == -1, "getsockname", "errno:%d\n", errno))
+		goto done;
+	err = listen(lfd, 1);
+	if (CHECK(err == -1, "listen", "errno:%d\n", errno))
+		goto done;
+
+	/* connect to server */
+	err = connect(fd, (struct sockaddr *)&sa6, addrlen);
+	if (CHECK(err == -1, "connect", "errno:%d\n", errno))
+		goto done;
+
+	err = pthread_create(&srv_thread, NULL, server, (void *)(long)lfd);
+	if (CHECK(err != 0, "pthread_create", "err:%d errno:%d\n", err, errno))
+		goto done;
+
+	/* recv total_bytes */
+	while (bytes < total_bytes && !READ_ONCE(stop)) {
+		nr_recv = recv(fd, &batch,
+			       min(total_bytes - bytes, sizeof(batch)), 0);
+		if (nr_recv == -1 && errno == EINTR)
+			continue;
+		if (nr_recv == -1)
+			break;
+		bytes += nr_recv;
+	}
+
+	CHECK(bytes != total_bytes, "recv", "%zd != %u nr_recv:%zd errno:%d\n",
+	      bytes, total_bytes, nr_recv, errno);
+
+	WRITE_ONCE(stop, 1);
+	pthread_join(srv_thread, &thread_ret);
+	CHECK(IS_ERR(thread_ret), "pthread_join", "thread_ret:%ld",
+	      PTR_ERR(thread_ret));
+done:
+	close(lfd);
+	close(fd);
+}
+
+static void test_first(void)
+{
+	struct bpf_first *first_skel;
+	struct bpf_link *link;
+
+	first_skel = bpf_first__open_and_load();
+	if (CHECK(!first_skel, "bpf_first__open_and_load", "failed\n"))
+		return;
+
+	link = bpf_map__attach_struct_ops(first_skel->maps.first);
+	if (!ASSERT_OK_PTR(link, "bpf_map__attach_struct_ops")) {
+		bpf_first__destroy(first_skel);
+		return;
+	}
+
+	do_test("bpf_first");
+
+	bpf_link__destroy(link);
+	bpf_first__destroy(first_skel);
+}
+
+void test_bpf_mptcp_sched(void)
+{
+	if (test__start_subtest("first"))
+		test_first();
+}
Signature · 1 line
2.34.1