[linux-block.git] / net / mptcp / pm.c

// SPDX-License-Identifier: GPL-2.0
/* Multipath TCP
 *
 * Copyright (c) 2019, Intel Corporation.
 */
#define pr_fmt(fmt) "MPTCP: " fmt

#include <linux/kernel.h>
#include <net/tcp.h>
#include <net/mptcp.h>
#include "protocol.h"

#include "mib.h"

/* path manager command handlers */

int mptcp_pm_announce_addr(struct mptcp_sock *msk,
			   const struct mptcp_addr_info *addr,
			   bool echo)
{
	u8 add_addr = READ_ONCE(msk->pm.addr_signal);

	pr_debug("msk=%p, local_id=%d, echo=%d", msk, addr->id, echo);

	lockdep_assert_held(&msk->pm.lock);

	if (add_addr &
	    (echo ? BIT(MPTCP_ADD_ADDR_ECHO) : BIT(MPTCP_ADD_ADDR_SIGNAL))) {
		pr_warn("addr_signal error, add_addr=%d, echo=%d", add_addr, echo);
		return -EINVAL;
	}

	if (echo) {
		msk->pm.remote = *addr;
		add_addr |= BIT(MPTCP_ADD_ADDR_ECHO);
	} else {
		msk->pm.local = *addr;
		add_addr |= BIT(MPTCP_ADD_ADDR_SIGNAL);
	}
	WRITE_ONCE(msk->pm.addr_signal, add_addr);
	return 0;
}

int mptcp_pm_remove_addr(struct mptcp_sock *msk, const struct mptcp_rm_list *rm_list)
{
	u8 rm_addr = READ_ONCE(msk->pm.addr_signal);

	pr_debug("msk=%p, rm_list_nr=%d", msk, rm_list->nr);

	if (rm_addr) {
		pr_warn("addr_signal error, rm_addr=%d", rm_addr);
		return -EINVAL;
	}

	msk->pm.rm_list_tx = *rm_list;
	rm_addr |= BIT(MPTCP_RM_ADDR_SIGNAL);
	WRITE_ONCE(msk->pm.addr_signal, rm_addr);
	mptcp_pm_nl_addr_send_ack(msk);
	return 0;
}

int mptcp_pm_remove_subflow(struct mptcp_sock *msk, const struct mptcp_rm_list *rm_list)
{
	pr_debug("msk=%p, rm_list_nr=%d", msk, rm_list->nr);

	spin_lock_bh(&msk->pm.lock);
	mptcp_pm_nl_rm_subflow_received(msk, rm_list);
	spin_unlock_bh(&msk->pm.lock);
	return 0;
}

/* path manager event handlers */

void mptcp_pm_new_connection(struct mptcp_sock *msk, const struct sock *ssk, int server_side)
{
	struct mptcp_pm_data *pm = &msk->pm;

	pr_debug("msk=%p, token=%u side=%d", msk, msk->token, server_side);

	WRITE_ONCE(pm->server_side, server_side);
	mptcp_event(MPTCP_EVENT_CREATED, msk, ssk, GFP_ATOMIC);
}

bool mptcp_pm_allow_new_subflow(struct mptcp_sock *msk)
{
	struct mptcp_pm_data *pm = &msk->pm;
	unsigned int subflows_max;
	int ret = 0;

	subflows_max = mptcp_pm_get_subflows_max(msk);

	pr_debug("msk=%p subflows=%d max=%d allow=%d", msk, pm->subflows,
		 subflows_max, READ_ONCE(pm->accept_subflow));

	/* try to avoid acquiring the lock below */
	if (!READ_ONCE(pm->accept_subflow))
		return false;

	spin_lock_bh(&pm->lock);
	if (READ_ONCE(pm->accept_subflow)) {
		ret = pm->subflows < subflows_max;
		if (ret && ++pm->subflows == subflows_max)
			WRITE_ONCE(pm->accept_subflow, false);
	}
	spin_unlock_bh(&pm->lock);

	return ret;
}

/* return true if the new status bit is currently cleared, that is, this event
 * can be server, eventually by an already scheduled work
 */
static bool mptcp_pm_schedule_work(struct mptcp_sock *msk,
				   enum mptcp_pm_status new_status)
{
	pr_debug("msk=%p status=%x new=%lx", msk, msk->pm.status,
		 BIT(new_status));
	if (msk->pm.status & BIT(new_status))
		return false;

	msk->pm.status |= BIT(new_status);
	mptcp_schedule_work((struct sock *)msk);
	return true;
}

void mptcp_pm_fully_established(struct mptcp_sock *msk, const struct sock *ssk, gfp_t gfp)
{
	struct mptcp_pm_data *pm = &msk->pm;
	bool announce = false;

	pr_debug("msk=%p", msk);

	spin_lock_bh(&pm->lock);

	/* mptcp_pm_fully_established() can be invoked by multiple
	 * racing paths - accept() and check_fully_established()
	 * be sure to serve this event only once.
	 */
	if (READ_ONCE(pm->work_pending) &&
	    !(msk->pm.status & BIT(MPTCP_PM_ALREADY_ESTABLISHED)))
		mptcp_pm_schedule_work(msk, MPTCP_PM_ESTABLISHED);

	if ((msk->pm.status & BIT(MPTCP_PM_ALREADY_ESTABLISHED)) == 0)
		announce = true;

	msk->pm.status |= BIT(MPTCP_PM_ALREADY_ESTABLISHED);
	spin_unlock_bh(&pm->lock);

	if (announce)
		mptcp_event(MPTCP_EVENT_ESTABLISHED, msk, ssk, gfp);
}

void mptcp_pm_connection_closed(struct mptcp_sock *msk)
{
	pr_debug("msk=%p", msk);
}

void mptcp_pm_subflow_established(struct mptcp_sock *msk)
{
	struct mptcp_pm_data *pm = &msk->pm;

	pr_debug("msk=%p", msk);

	if (!READ_ONCE(pm->work_pending))
		return;

	spin_lock_bh(&pm->lock);

	if (READ_ONCE(pm->work_pending))
		mptcp_pm_schedule_work(msk, MPTCP_PM_SUBFLOW_ESTABLISHED);

	spin_unlock_bh(&pm->lock);
}

void mptcp_pm_subflow_check_next(struct mptcp_sock *msk, const struct sock *ssk,
				 const struct mptcp_subflow_context *subflow)
{
	struct mptcp_pm_data *pm = &msk->pm;
	bool update_subflows;

	update_subflows = (ssk->sk_state == TCP_CLOSE) &&
			  (subflow->request_join || subflow->mp_join);
	if (!READ_ONCE(pm->work_pending) && !update_subflows)
		return;

	spin_lock_bh(&pm->lock);
	if (update_subflows)
		pm->subflows--;

	/* Even if this subflow is not really established, tell the PM to try
	 * to pick the next ones, if possible.
	 */
	if (mptcp_pm_nl_check_work_pending(msk))
		mptcp_pm_schedule_work(msk, MPTCP_PM_SUBFLOW_ESTABLISHED);

	spin_unlock_bh(&pm->lock);
}

void mptcp_pm_add_addr_received(struct mptcp_sock *msk,
				const struct mptcp_addr_info *addr)
{
	struct mptcp_pm_data *pm = &msk->pm;

	pr_debug("msk=%p remote_id=%d accept=%d", msk, addr->id,
		 READ_ONCE(pm->accept_addr));

	mptcp_event_addr_announced(msk, addr);

	spin_lock_bh(&pm->lock);

	if (!READ_ONCE(pm->accept_addr)) {
		mptcp_pm_announce_addr(msk, addr, true);
		mptcp_pm_add_addr_send_ack(msk);
	} else if (mptcp_pm_schedule_work(msk, MPTCP_PM_ADD_ADDR_RECEIVED)) {
		pm->remote = *addr;
	} else {
		__MPTCP_INC_STATS(sock_net((struct sock *)msk), MPTCP_MIB_ADDADDRDROP);
	}

	spin_unlock_bh(&pm->lock);
}

void mptcp_pm_add_addr_echoed(struct mptcp_sock *msk,
			      const struct mptcp_addr_info *addr)
{
	struct mptcp_pm_data *pm = &msk->pm;

	pr_debug("msk=%p", msk);

	spin_lock_bh(&pm->lock);

	if (mptcp_lookup_anno_list_by_saddr(msk, addr) && READ_ONCE(pm->work_pending))
		mptcp_pm_schedule_work(msk, MPTCP_PM_SUBFLOW_ESTABLISHED);

	spin_unlock_bh(&pm->lock);
}

void mptcp_pm_add_addr_send_ack(struct mptcp_sock *msk)
{
	if (!mptcp_pm_should_add_signal(msk))
		return;

	mptcp_pm_schedule_work(msk, MPTCP_PM_ADD_ADDR_SEND_ACK);
}

void mptcp_pm_rm_addr_received(struct mptcp_sock *msk,
			       const struct mptcp_rm_list *rm_list)
{
	struct mptcp_pm_data *pm = &msk->pm;
	u8 i;

	pr_debug("msk=%p remote_ids_nr=%d", msk, rm_list->nr);

	for (i = 0; i < rm_list->nr; i++)
		mptcp_event_addr_removed(msk, rm_list->ids[i]);

	spin_lock_bh(&pm->lock);
	if (mptcp_pm_schedule_work(msk, MPTCP_PM_RM_ADDR_RECEIVED))
		pm->rm_list_rx = *rm_list;
	else
		__MPTCP_INC_STATS(sock_net((struct sock *)msk), MPTCP_MIB_RMADDRDROP);
	spin_unlock_bh(&pm->lock);
}

void mptcp_pm_mp_prio_received(struct sock *ssk, u8 bkup)
{
	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(ssk);
	struct sock *sk = subflow->conn;
	struct mptcp_sock *msk;

	pr_debug("subflow->backup=%d, bkup=%d\n", subflow->backup, bkup);
	msk = mptcp_sk(sk);
	if (subflow->backup != bkup) {
		subflow->backup = bkup;
		mptcp_data_lock(sk);
		if (!sock_owned_by_user(sk))
			msk->last_snd = NULL;
		else
			__set_bit(MPTCP_RESET_SCHEDULER,  &msk->cb_flags);
		mptcp_data_unlock(sk);
	}

	mptcp_event(MPTCP_EVENT_SUB_PRIORITY, msk, ssk, GFP_ATOMIC);
}

void mptcp_pm_mp_fail_received(struct sock *sk, u64 fail_seq)
{
	pr_debug("fail_seq=%llu", fail_seq);
}

/* path manager helpers */

bool mptcp_pm_add_addr_signal(struct mptcp_sock *msk, const struct sk_buff *skb,
			      unsigned int opt_size, unsigned int remaining,
			      struct mptcp_addr_info *addr, bool *echo,
			      bool *drop_other_suboptions)
{
	int ret = false;
	u8 add_addr;
	u8 family;
	bool port;

	spin_lock_bh(&msk->pm.lock);

	/* double check after the lock is acquired */
	if (!mptcp_pm_should_add_signal(msk))
		goto out_unlock;

	/* always drop every other options for pure ack ADD_ADDR; this is a
	 * plain dup-ack from TCP perspective. The other MPTCP-relevant info,
	 * if any, will be carried by the 'original' TCP ack
	 */
	if (skb && skb_is_tcp_pure_ack(skb)) {
		remaining += opt_size;
		*drop_other_suboptions = true;
	}

	*echo = mptcp_pm_should_add_signal_echo(msk);
	port = !!(*echo ? msk->pm.remote.port : msk->pm.local.port);

	family = *echo ? msk->pm.remote.family : msk->pm.local.family;
	if (remaining < mptcp_add_addr_len(family, *echo, port))
		goto out_unlock;

	if (*echo) {
		*addr = msk->pm.remote;
		add_addr = msk->pm.addr_signal & ~BIT(MPTCP_ADD_ADDR_ECHO);
	} else {
		*addr = msk->pm.local;
		add_addr = msk->pm.addr_signal & ~BIT(MPTCP_ADD_ADDR_SIGNAL);
	}
	WRITE_ONCE(msk->pm.addr_signal, add_addr);
	ret = true;

out_unlock:
	spin_unlock_bh(&msk->pm.lock);
	return ret;
}

bool mptcp_pm_rm_addr_signal(struct mptcp_sock *msk, unsigned int remaining,
			     struct mptcp_rm_list *rm_list)
{
	int ret = false, len;
	u8 rm_addr;

	spin_lock_bh(&msk->pm.lock);

	/* double check after the lock is acquired */
	if (!mptcp_pm_should_rm_signal(msk))
		goto out_unlock;

	rm_addr = msk->pm.addr_signal & ~BIT(MPTCP_RM_ADDR_SIGNAL);
	len = mptcp_rm_addr_len(&msk->pm.rm_list_tx);
	if (len < 0) {
		WRITE_ONCE(msk->pm.addr_signal, rm_addr);
		goto out_unlock;
	}
	if (remaining < len)
		goto out_unlock;

	*rm_list = msk->pm.rm_list_tx;
	WRITE_ONCE(msk->pm.addr_signal, rm_addr);
	ret = true;

out_unlock:
	spin_unlock_bh(&msk->pm.lock);
	return ret;
}

int mptcp_pm_get_local_id(struct mptcp_sock *msk, struct sock_common *skc)
{
	return mptcp_pm_nl_get_local_id(msk, skc);
}

void mptcp_pm_subflow_chk_stale(const struct mptcp_sock *msk, struct sock *ssk)
{
	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(ssk);
	u32 rcv_tstamp = READ_ONCE(tcp_sk(ssk)->rcv_tstamp);

	/* keep track of rtx periods with no progress */
	if (!subflow->stale_count) {
		subflow->stale_rcv_tstamp = rcv_tstamp;
		subflow->stale_count++;
	} else if (subflow->stale_rcv_tstamp == rcv_tstamp) {
		if (subflow->stale_count < U8_MAX)
			subflow->stale_count++;
		mptcp_pm_nl_subflow_chk_stale(msk, ssk);
	} else {
		subflow->stale_count = 0;
		mptcp_subflow_set_active(subflow);
	}
}

void mptcp_pm_data_reset(struct mptcp_sock *msk)
{
	msk->pm.add_addr_signaled = 0;
	msk->pm.add_addr_accepted = 0;
	msk->pm.local_addr_used = 0;
	msk->pm.subflows = 0;
	msk->pm.rm_list_tx.nr = 0;
	msk->pm.rm_list_rx.nr = 0;
	WRITE_ONCE(msk->pm.work_pending, false);
	WRITE_ONCE(msk->pm.addr_signal, 0);
	WRITE_ONCE(msk->pm.accept_addr, false);
	WRITE_ONCE(msk->pm.accept_subflow, false);
	WRITE_ONCE(msk->pm.remote_deny_join_id0, false);
	msk->pm.status = 0;
	bitmap_fill(msk->pm.id_avail_bitmap, MPTCP_PM_MAX_ADDR_ID + 1);

	mptcp_pm_nl_data_init(msk);
}

void mptcp_pm_data_init(struct mptcp_sock *msk)
{
	spin_lock_init(&msk->pm.lock);
	INIT_LIST_HEAD(&msk->pm.anno_list);
	mptcp_pm_data_reset(msk);
}

void __init mptcp_pm_init(void)
{
	mptcp_pm_nl_init();
}
Commit	Line	Data
1b1c7a0e PK	1	// SPDX-License-Identifier: GPL-2.0
	2	/* Multipath TCP
	3	*
	4	* Copyright (c) 2019, Intel Corporation.
	5	*/
c85adced GT	6	#define pr_fmt(fmt) "MPTCP: " fmt
c85adced GT	7
1b1c7a0e PK	8	#include <linux/kernel.h>
	9	#include <net/tcp.h>
	10	#include <net/mptcp.h>
	11	#include "protocol.h"
	12
fc1b4e3b PA	13	#include "mib.h"
fc1b4e3b PA	14
1b1c7a0e PK	15	/* path manager command handlers */
	16
	17	int mptcp_pm_announce_addr(struct mptcp_sock *msk,
6a6c05a8	18	const struct mptcp_addr_info *addr,
f7efc777	19	bool echo)
1b1c7a0e	20	{
13ad9f01	21	u8 add_addr = READ_ONCE(msk->pm.addr_signal);
d91d322a	22
18fc1a92	23	pr_debug("msk=%p, local_id=%d, echo=%d", msk, addr->id, echo);
926bdeab	24
3abc05d9 FW	25	lockdep_assert_held(&msk->pm.lock);
3abc05d9 FW	26
18fc1a92 YL	27	if (add_addr &
	28	(echo ? BIT(MPTCP_ADD_ADDR_ECHO) : BIT(MPTCP_ADD_ADDR_SIGNAL))) {
	29	pr_warn("addr_signal error, add_addr=%d, echo=%d", add_addr, echo);
42842a42 GT	30	return -EINVAL;
	31	}
	32
18fc1a92 YL	33	if (echo) {
18fc1a92 YL	34	msk->pm.remote = *addr;
d91d322a	35	add_addr \|= BIT(MPTCP_ADD_ADDR_ECHO);
18fc1a92 YL	36	} else {
	37	msk->pm.local = *addr;
	38	add_addr \|= BIT(MPTCP_ADD_ADDR_SIGNAL);
	39	}
13ad9f01	40	WRITE_ONCE(msk->pm.addr_signal, add_addr);
926bdeab	41	return 0;
1b1c7a0e PK	42	}
1b1c7a0e PK	43
cbde2787	44	int mptcp_pm_remove_addr(struct mptcp_sock msk, const struct mptcp_rm_list rm_list)
1b1c7a0e	45	{
13ad9f01	46	u8 rm_addr = READ_ONCE(msk->pm.addr_signal);
42842a42	47
cbde2787	48	pr_debug("msk=%p, rm_list_nr=%d", msk, rm_list->nr);
b6c08380	49
42842a42 GT	50	if (rm_addr) {
	51	pr_warn("addr_signal error, rm_addr=%d", rm_addr);
	52	return -EINVAL;
	53	}
	54
cbde2787	55	msk->pm.rm_list_tx = *rm_list;
42842a42	56	rm_addr \|= BIT(MPTCP_RM_ADDR_SIGNAL);
13ad9f01	57	WRITE_ONCE(msk->pm.addr_signal, rm_addr);
b46a0238	58	mptcp_pm_nl_addr_send_ack(msk);
b6c08380	59	return 0;
1b1c7a0e PK	60	}
1b1c7a0e PK	61
ddd14bb8	62	int mptcp_pm_remove_subflow(struct mptcp_sock msk, const struct mptcp_rm_list rm_list)
1b1c7a0e	63	{
ddd14bb8	64	pr_debug("msk=%p, rm_list_nr=%d", msk, rm_list->nr);
0ee4261a GT	65
0ee4261a GT	66	spin_lock_bh(&msk->pm.lock);
ddd14bb8	67	mptcp_pm_nl_rm_subflow_received(msk, rm_list);
0ee4261a GT	68	spin_unlock_bh(&msk->pm.lock);
0ee4261a GT	69	return 0;
1b1c7a0e PK	70	}
	71
	72	/* path manager event handlers */
	73
6c714f1b	74	void mptcp_pm_new_connection(struct mptcp_sock msk, const struct sock ssk, int server_side)
1b1c7a0e PK	75	{
	76	struct mptcp_pm_data *pm = &msk->pm;
	77
	78	pr_debug("msk=%p, token=%u side=%d", msk, msk->token, server_side);
	79
	80	WRITE_ONCE(pm->server_side, server_side);
b911c97c	81	mptcp_event(MPTCP_EVENT_CREATED, msk, ssk, GFP_ATOMIC);
1b1c7a0e PK	82	}
	83
	84	bool mptcp_pm_allow_new_subflow(struct mptcp_sock *msk)
	85	{
926bdeab	86	struct mptcp_pm_data *pm = &msk->pm;
a914e586	87	unsigned int subflows_max;
f58f065a	88	int ret = 0;
926bdeab	89
a914e586 GT	90	subflows_max = mptcp_pm_get_subflows_max(msk);
a914e586 GT	91
926bdeab	92	pr_debug("msk=%p subflows=%d max=%d allow=%d", msk, pm->subflows,
a914e586	93	subflows_max, READ_ONCE(pm->accept_subflow));
926bdeab PK	94
	95	/* try to avoid acquiring the lock below */
	96	if (!READ_ONCE(pm->accept_subflow))
	97	return false;
	98
	99	spin_lock_bh(&pm->lock);
f58f065a	100	if (READ_ONCE(pm->accept_subflow)) {
a914e586 GT	101	ret = pm->subflows < subflows_max;
a914e586 GT	102	if (ret && ++pm->subflows == subflows_max)
f58f065a GT	103	WRITE_ONCE(pm->accept_subflow, false);
f58f065a GT	104	}
926bdeab PK	105	spin_unlock_bh(&pm->lock);
	106
	107	return ret;
	108	}
	109
	110	/* return true if the new status bit is currently cleared, that is, this event
	111	* can be server, eventually by an already scheduled work
	112	*/
	113	static bool mptcp_pm_schedule_work(struct mptcp_sock *msk,
	114	enum mptcp_pm_status new_status)
	115	{
	116	pr_debug("msk=%p status=%x new=%lx", msk, msk->pm.status,
	117	BIT(new_status));
	118	if (msk->pm.status & BIT(new_status))
	119	return false;
	120
	121	msk->pm.status \|= BIT(new_status);
ba8f48f7	122	mptcp_schedule_work((struct sock *)msk);
926bdeab	123	return true;
1b1c7a0e PK	124	}
1b1c7a0e PK	125
6c714f1b	126	void mptcp_pm_fully_established(struct mptcp_sock msk, const struct sock ssk, gfp_t gfp)
1b1c7a0e	127	{
926bdeab	128	struct mptcp_pm_data *pm = &msk->pm;
b911c97c	129	bool announce = false;
926bdeab	130
1b1c7a0e	131	pr_debug("msk=%p", msk);
926bdeab	132
926bdeab PK	133	spin_lock_bh(&pm->lock);
926bdeab PK	134
5b950ff4 PA	135	/* mptcp_pm_fully_established() can be invoked by multiple
	136	* racing paths - accept() and check_fully_established()
	137	* be sure to serve this event only once.
	138	*/
	139	if (READ_ONCE(pm->work_pending) &&
	140	!(msk->pm.status & BIT(MPTCP_PM_ALREADY_ESTABLISHED)))
926bdeab PK	141	mptcp_pm_schedule_work(msk, MPTCP_PM_ESTABLISHED);
926bdeab PK	142
b911c97c FW	143	if ((msk->pm.status & BIT(MPTCP_PM_ALREADY_ESTABLISHED)) == 0)
	144	announce = true;
	145
	146	msk->pm.status \|= BIT(MPTCP_PM_ALREADY_ESTABLISHED);
926bdeab	147	spin_unlock_bh(&pm->lock);
b911c97c FW	148
	149	if (announce)
	150	mptcp_event(MPTCP_EVENT_ESTABLISHED, msk, ssk, gfp);
1b1c7a0e PK	151	}
	152
	153	void mptcp_pm_connection_closed(struct mptcp_sock *msk)
	154	{
	155	pr_debug("msk=%p", msk);
	156	}
	157
62535200	158	void mptcp_pm_subflow_established(struct mptcp_sock *msk)
1b1c7a0e	159	{
926bdeab PK	160	struct mptcp_pm_data *pm = &msk->pm;
926bdeab PK	161
1b1c7a0e	162	pr_debug("msk=%p", msk);
926bdeab PK	163
	164	if (!READ_ONCE(pm->work_pending))
	165	return;
	166
	167	spin_lock_bh(&pm->lock);
	168
	169	if (READ_ONCE(pm->work_pending))
	170	mptcp_pm_schedule_work(msk, MPTCP_PM_SUBFLOW_ESTABLISHED);
	171
	172	spin_unlock_bh(&pm->lock);
1b1c7a0e PK	173	}
1b1c7a0e PK	174
a88c9e49 PA	175	void mptcp_pm_subflow_check_next(struct mptcp_sock msk, const struct sock ssk,
a88c9e49 PA	176	const struct mptcp_subflow_context *subflow)
1b1c7a0e	177	{
a88c9e49 PA	178	struct mptcp_pm_data *pm = &msk->pm;
	179	bool update_subflows;
	180
	181	update_subflows = (ssk->sk_state == TCP_CLOSE) &&
	182	(subflow->request_join \|\| subflow->mp_join);
	183	if (!READ_ONCE(pm->work_pending) && !update_subflows)
	184	return;
	185
	186	spin_lock_bh(&pm->lock);
	187	if (update_subflows)
	188	pm->subflows--;
	189
	190	/* Even if this subflow is not really established, tell the PM to try
	191	* to pick the next ones, if possible.
	192	*/
	193	if (mptcp_pm_nl_check_work_pending(msk))
	194	mptcp_pm_schedule_work(msk, MPTCP_PM_SUBFLOW_ESTABLISHED);
	195
	196	spin_unlock_bh(&pm->lock);
1b1c7a0e PK	197	}
	198
	199	void mptcp_pm_add_addr_received(struct mptcp_sock *msk,
	200	const struct mptcp_addr_info *addr)
	201	{
926bdeab PK	202	struct mptcp_pm_data *pm = &msk->pm;
	203
	204	pr_debug("msk=%p remote_id=%d accept=%d", msk, addr->id,
	205	READ_ONCE(pm->accept_addr));
	206
b911c97c FW	207	mptcp_event_addr_announced(msk, addr);
b911c97c FW	208
926bdeab PK	209	spin_lock_bh(&pm->lock);
926bdeab PK	210
84dfe367	211	if (!READ_ONCE(pm->accept_addr)) {
f7efc777	212	mptcp_pm_announce_addr(msk, addr, true);
84dfe367 GT	213	mptcp_pm_add_addr_send_ack(msk);
84dfe367 GT	214	} else if (mptcp_pm_schedule_work(msk, MPTCP_PM_ADD_ADDR_RECEIVED)) {
926bdeab	215	pm->remote = *addr;
f73c1194 PA	216	} else {
f73c1194 PA	217	__MPTCP_INC_STATS(sock_net((struct sock *)msk), MPTCP_MIB_ADDADDRDROP);
84dfe367	218	}
926bdeab PK	219
926bdeab PK	220	spin_unlock_bh(&pm->lock);
84dfe367 GT	221	}
84dfe367 GT	222
557963c3	223	void mptcp_pm_add_addr_echoed(struct mptcp_sock *msk,
90d93088	224	const struct mptcp_addr_info *addr)
557963c3 GT	225	{
	226	struct mptcp_pm_data *pm = &msk->pm;
	227
	228	pr_debug("msk=%p", msk);
	229
	230	spin_lock_bh(&pm->lock);
	231
	232	if (mptcp_lookup_anno_list_by_saddr(msk, addr) && READ_ONCE(pm->work_pending))
	233	mptcp_pm_schedule_work(msk, MPTCP_PM_SUBFLOW_ESTABLISHED);
	234
	235	spin_unlock_bh(&pm->lock);
	236	}
	237
84dfe367 GT	238	void mptcp_pm_add_addr_send_ack(struct mptcp_sock *msk)
84dfe367 GT	239	{
b5a7acd3	240	if (!mptcp_pm_should_add_signal(msk))
84dfe367 GT	241	return;
	242
	243	mptcp_pm_schedule_work(msk, MPTCP_PM_ADD_ADDR_SEND_ACK);
1b1c7a0e PK	244	}
1b1c7a0e PK	245
5c4a824d GT	246	void mptcp_pm_rm_addr_received(struct mptcp_sock *msk,
5c4a824d GT	247	const struct mptcp_rm_list *rm_list)
d0876b22 GT	248	{
d0876b22 GT	249	struct mptcp_pm_data *pm = &msk->pm;
5c4a824d	250	u8 i;
d0876b22	251
5c4a824d	252	pr_debug("msk=%p remote_ids_nr=%d", msk, rm_list->nr);
d0876b22	253
5c4a824d GT	254	for (i = 0; i < rm_list->nr; i++)
5c4a824d GT	255	mptcp_event_addr_removed(msk, rm_list->ids[i]);
b911c97c	256
d0876b22	257	spin_lock_bh(&pm->lock);
f73c1194 PA	258	if (mptcp_pm_schedule_work(msk, MPTCP_PM_RM_ADDR_RECEIVED))
	259	pm->rm_list_rx = *rm_list;
	260	else
	261	__MPTCP_INC_STATS(sock_net((struct sock *)msk), MPTCP_MIB_RMADDRDROP);
d0876b22 GT	262	spin_unlock_bh(&pm->lock);
	263	}
	264
43f5b111	265	void mptcp_pm_mp_prio_received(struct sock *ssk, u8 bkup)
40453a5c	266	{
43f5b111 PA	267	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(ssk);
	268	struct sock *sk = subflow->conn;
	269	struct mptcp_sock *msk;
40453a5c GT	270
40453a5c GT	271	pr_debug("subflow->backup=%d, bkup=%d\n", subflow->backup, bkup);
43f5b111 PA	272	msk = mptcp_sk(sk);
	273	if (subflow->backup != bkup) {
	274	subflow->backup = bkup;
	275	mptcp_data_lock(sk);
	276	if (!sock_owned_by_user(sk))
	277	msk->last_snd = NULL;
	278	else
	279	__set_bit(MPTCP_RESET_SCHEDULER, &msk->cb_flags);
	280	mptcp_data_unlock(sk);
	281	}
b911c97c	282
43f5b111	283	mptcp_event(MPTCP_EVENT_SUB_PRIORITY, msk, ssk, GFP_ATOMIC);
40453a5c GT	284	}
40453a5c GT	285
5580d41b GT	286	void mptcp_pm_mp_fail_received(struct sock *sk, u64 fail_seq)
	287	{
	288	pr_debug("fail_seq=%llu", fail_seq);
	289	}
	290
1b1c7a0e PK	291	/* path manager helpers */
1b1c7a0e PK	292
90d93088	293	bool mptcp_pm_add_addr_signal(struct mptcp_sock msk, const struct sk_buff skb,
1f5e9e2f	294	unsigned int opt_size, unsigned int remaining,
f462a446	295	struct mptcp_addr_info addr, bool echo,
af7939f3	296	bool *drop_other_suboptions)
1b1c7a0e	297	{
926bdeab	298	int ret = false;
119c0220	299	u8 add_addr;
f462a446	300	u8 family;
af7939f3	301	bool port;
926bdeab PK	302
	303	spin_lock_bh(&msk->pm.lock);
	304
	305	/* double check after the lock is acquired */
f643b803	306	if (!mptcp_pm_should_add_signal(msk))
926bdeab PK	307	goto out_unlock;
926bdeab PK	308
1f5e9e2f YL	309	/* always drop every other options for pure ack ADD_ADDR; this is a
	310	* plain dup-ack from TCP perspective. The other MPTCP-relevant info,
	311	* if any, will be carried by the 'original' TCP ack
	312	*/
	313	if (skb && skb_is_tcp_pure_ack(skb)) {
	314	remaining += opt_size;
	315	*drop_other_suboptions = true;
	316	}
	317
d91d322a	318	*echo = mptcp_pm_should_add_signal_echo(msk);
af7939f3	319	port = !!(*echo ? msk->pm.remote.port : msk->pm.local.port);
456afe01	320
f462a446	321	family = *echo ? msk->pm.remote.family : msk->pm.local.family;
af7939f3	322	if (remaining < mptcp_add_addr_len(family, *echo, port))
926bdeab PK	323	goto out_unlock;
926bdeab PK	324
f462a446 YL	325	if (*echo) {
f462a446 YL	326	*addr = msk->pm.remote;
119c0220	327	add_addr = msk->pm.addr_signal & ~BIT(MPTCP_ADD_ADDR_ECHO);
f462a446 YL	328	} else {
f462a446 YL	329	*addr = msk->pm.local;
119c0220	330	add_addr = msk->pm.addr_signal & ~BIT(MPTCP_ADD_ADDR_SIGNAL);
f462a446	331	}
119c0220	332	WRITE_ONCE(msk->pm.addr_signal, add_addr);
926bdeab PK	333	ret = true;
	334
	335	out_unlock:
	336	spin_unlock_bh(&msk->pm.lock);
	337	return ret;
1b1c7a0e PK	338	}
1b1c7a0e PK	339
5cb104ae	340	bool mptcp_pm_rm_addr_signal(struct mptcp_sock *msk, unsigned int remaining,
6445e17a	341	struct mptcp_rm_list *rm_list)
5cb104ae	342	{
cbde2787	343	int ret = false, len;
119c0220	344	u8 rm_addr;
5cb104ae GT	345
	346	spin_lock_bh(&msk->pm.lock);
	347
	348	/* double check after the lock is acquired */
	349	if (!mptcp_pm_should_rm_signal(msk))
	350	goto out_unlock;
	351
119c0220	352	rm_addr = msk->pm.addr_signal & ~BIT(MPTCP_RM_ADDR_SIGNAL);
cbde2787 GT	353	len = mptcp_rm_addr_len(&msk->pm.rm_list_tx);
cbde2787 GT	354	if (len < 0) {
119c0220	355	WRITE_ONCE(msk->pm.addr_signal, rm_addr);
cbde2787 GT	356	goto out_unlock;
	357	}
	358	if (remaining < len)
5cb104ae GT	359	goto out_unlock;
5cb104ae GT	360
cbde2787	361	*rm_list = msk->pm.rm_list_tx;
119c0220	362	WRITE_ONCE(msk->pm.addr_signal, rm_addr);
5cb104ae GT	363	ret = true;
	364
	365	out_unlock:
	366	spin_unlock_bh(&msk->pm.lock);
	367	return ret;
	368	}
	369
1b1c7a0e PK	370	int mptcp_pm_get_local_id(struct mptcp_sock msk, struct sock_common skc)
1b1c7a0e PK	371	{
01cacb00	372	return mptcp_pm_nl_get_local_id(msk, skc);
1b1c7a0e PK	373	}
1b1c7a0e PK	374
71b7dec2 PA	375	void mptcp_pm_subflow_chk_stale(const struct mptcp_sock msk, struct sock ssk)
	376	{
	377	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(ssk);
	378	u32 rcv_tstamp = READ_ONCE(tcp_sk(ssk)->rcv_tstamp);
	379
	380	/* keep track of rtx periods with no progress */
	381	if (!subflow->stale_count) {
	382	subflow->stale_rcv_tstamp = rcv_tstamp;
	383	subflow->stale_count++;
	384	} else if (subflow->stale_rcv_tstamp == rcv_tstamp) {
	385	if (subflow->stale_count < U8_MAX)
	386	subflow->stale_count++;
ff5a0b42	387	mptcp_pm_nl_subflow_chk_stale(msk, ssk);
71b7dec2 PA	388	} else {
71b7dec2 PA	389	subflow->stale_count = 0;
ff5a0b42	390	mptcp_subflow_set_active(subflow);
71b7dec2 PA	391	}
	392	}
	393
b29fcfb5	394	void mptcp_pm_data_reset(struct mptcp_sock *msk)
1b1c7a0e PK	395	{
	396	msk->pm.add_addr_signaled = 0;
	397	msk->pm.add_addr_accepted = 0;
	398	msk->pm.local_addr_used = 0;
	399	msk->pm.subflows = 0;
cbde2787	400	msk->pm.rm_list_tx.nr = 0;
b5c55f33	401	msk->pm.rm_list_rx.nr = 0;
1b1c7a0e	402	WRITE_ONCE(msk->pm.work_pending, false);
13ad9f01	403	WRITE_ONCE(msk->pm.addr_signal, 0);
1b1c7a0e PK	404	WRITE_ONCE(msk->pm.accept_addr, false);
1b1c7a0e PK	405	WRITE_ONCE(msk->pm.accept_subflow, false);
df377be3	406	WRITE_ONCE(msk->pm.remote_deny_join_id0, false);
1b1c7a0e	407	msk->pm.status = 0;
86e39e04	408	bitmap_fill(msk->pm.id_avail_bitmap, MPTCP_PM_MAX_ADDR_ID + 1);
1b1c7a0e	409
b29fcfb5 PA	410	mptcp_pm_nl_data_init(msk);
	411	}
	412
	413	void mptcp_pm_data_init(struct mptcp_sock *msk)
	414	{
1b1c7a0e	415	spin_lock_init(&msk->pm.lock);
b6c08380	416	INIT_LIST_HEAD(&msk->pm.anno_list);
b29fcfb5	417	mptcp_pm_data_reset(msk);
1b1c7a0e PK	418	}
1b1c7a0e PK	419
d39dceca	420	void __init mptcp_pm_init(void)
1b1c7a0e	421	{
01cacb00	422	mptcp_pm_nl_init();
1b1c7a0e	423	}