[mirror_ubuntu-jammy-kernel.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"

/* path manager command handlers */

int mptcp_pm_announce_addr(struct mptcp_sock *msk,
			   const struct mptcp_addr_info *addr,
			   bool echo)
{
	pr_debug("msk=%p, local_id=%d", msk, addr->id);

	msk->pm.local = *addr;
	WRITE_ONCE(msk->pm.add_addr_echo, echo);
	WRITE_ONCE(msk->pm.add_addr_signal, true);
	return 0;
}

int mptcp_pm_remove_addr(struct mptcp_sock *msk, u8 local_id)
{
	return -ENOTSUPP;
}

int mptcp_pm_remove_subflow(struct mptcp_sock *msk, u8 remote_id)
{
	return -ENOTSUPP;
}

/* path manager event handlers */

void mptcp_pm_new_connection(struct mptcp_sock *msk, 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);
}

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

	pr_debug("msk=%p subflows=%d max=%d allow=%d", msk, pm->subflows,
		 pm->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);
	ret = pm->subflows < pm->subflows_max;
	if (ret && ++pm->subflows == pm->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);
	if (schedule_work(&msk->work))
		sock_hold((struct sock *)msk);
	return true;
}

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

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

	/* try to avoid acquiring the lock below */
	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_ESTABLISHED);

	spin_unlock_bh(&pm->lock);
}

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_subflow_context *subflow)
{
	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_closed(struct mptcp_sock *msk, u8 id)
{
	pr_debug("msk=%p", msk);
}

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));

	spin_lock_bh(&pm->lock);

	if (!READ_ONCE(pm->accept_addr))
		mptcp_pm_announce_addr(msk, addr, true);
	else if (mptcp_pm_schedule_work(msk, MPTCP_PM_ADD_ADDR_RECEIVED))
		pm->remote = *addr;

	spin_unlock_bh(&pm->lock);
}

void mptcp_pm_rm_addr_received(struct mptcp_sock *msk, u8 rm_id)
{
	struct mptcp_pm_data *pm = &msk->pm;

	pr_debug("msk=%p remote_id=%d", msk, rm_id);

	spin_lock_bh(&pm->lock);
	mptcp_pm_schedule_work(msk, MPTCP_PM_RM_ADDR_RECEIVED);
	pm->rm_id = rm_id;
	spin_unlock_bh(&pm->lock);
}

/* path manager helpers */

bool mptcp_pm_add_addr_signal(struct mptcp_sock *msk, unsigned int remaining,
			      struct mptcp_addr_info *saddr, bool *echo)
{
	int ret = false;

	spin_lock_bh(&msk->pm.lock);

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

	if (remaining < mptcp_add_addr_len(msk->pm.local.family))
		goto out_unlock;

	*saddr = msk->pm.local;
	*echo = READ_ONCE(msk->pm.add_addr_echo);
	WRITE_ONCE(msk->pm.add_addr_signal, false);
	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,
			     u8 *rm_id)
{
	int ret = false;

	spin_lock_bh(&msk->pm.lock);

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

	if (remaining < TCPOLEN_MPTCP_RM_ADDR_BASE)
		goto out_unlock;

	*rm_id = msk->pm.rm_id;
	WRITE_ONCE(msk->pm.rm_addr_signal, false);
	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_data_init(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_id = 0;
	WRITE_ONCE(msk->pm.work_pending, false);
	WRITE_ONCE(msk->pm.add_addr_signal, false);
	WRITE_ONCE(msk->pm.rm_addr_signal, false);
	WRITE_ONCE(msk->pm.accept_addr, false);
	WRITE_ONCE(msk->pm.accept_subflow, false);
	WRITE_ONCE(msk->pm.add_addr_echo, false);
	msk->pm.status = 0;

	spin_lock_init(&msk->pm.lock);

	mptcp_pm_nl_data_init(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
1b1c7a0e PK	13	/* path manager command handlers */
	14
	15	int mptcp_pm_announce_addr(struct mptcp_sock *msk,
6a6c05a8 GT	16	const struct mptcp_addr_info *addr,
6a6c05a8 GT	17	bool echo)
1b1c7a0e	18	{
926bdeab PK	19	pr_debug("msk=%p, local_id=%d", msk, addr->id);
	20
	21	msk->pm.local = *addr;
6a6c05a8	22	WRITE_ONCE(msk->pm.add_addr_echo, echo);
f643b803	23	WRITE_ONCE(msk->pm.add_addr_signal, true);
926bdeab	24	return 0;
1b1c7a0e PK	25	}
	26
	27	int mptcp_pm_remove_addr(struct mptcp_sock *msk, u8 local_id)
	28	{
	29	return -ENOTSUPP;
	30	}
	31
	32	int mptcp_pm_remove_subflow(struct mptcp_sock *msk, u8 remote_id)
	33	{
	34	return -ENOTSUPP;
	35	}
	36
	37	/* path manager event handlers */
	38
	39	void mptcp_pm_new_connection(struct mptcp_sock *msk, int server_side)
	40	{
	41	struct mptcp_pm_data *pm = &msk->pm;
	42
	43	pr_debug("msk=%p, token=%u side=%d", msk, msk->token, server_side);
	44
	45	WRITE_ONCE(pm->server_side, server_side);
	46	}
	47
	48	bool mptcp_pm_allow_new_subflow(struct mptcp_sock *msk)
	49	{
926bdeab PK	50	struct mptcp_pm_data *pm = &msk->pm;
	51	int ret;
	52
	53	pr_debug("msk=%p subflows=%d max=%d allow=%d", msk, pm->subflows,
	54	pm->subflows_max, READ_ONCE(pm->accept_subflow));
	55
	56	/* try to avoid acquiring the lock below */
	57	if (!READ_ONCE(pm->accept_subflow))
	58	return false;
	59
	60	spin_lock_bh(&pm->lock);
	61	ret = pm->subflows < pm->subflows_max;
	62	if (ret && ++pm->subflows == pm->subflows_max)
	63	WRITE_ONCE(pm->accept_subflow, false);
	64	spin_unlock_bh(&pm->lock);
	65
	66	return ret;
	67	}
	68
	69	/* return true if the new status bit is currently cleared, that is, this event
	70	* can be server, eventually by an already scheduled work
	71	*/
	72	static bool mptcp_pm_schedule_work(struct mptcp_sock *msk,
	73	enum mptcp_pm_status new_status)
	74	{
	75	pr_debug("msk=%p status=%x new=%lx", msk, msk->pm.status,
	76	BIT(new_status));
	77	if (msk->pm.status & BIT(new_status))
	78	return false;
	79
	80	msk->pm.status \|= BIT(new_status);
b416268b	81	if (schedule_work(&msk->work))
926bdeab PK	82	sock_hold((struct sock *)msk);
926bdeab PK	83	return true;
1b1c7a0e PK	84	}
	85
	86	void mptcp_pm_fully_established(struct mptcp_sock *msk)
	87	{
926bdeab PK	88	struct mptcp_pm_data *pm = &msk->pm;
926bdeab PK	89
1b1c7a0e	90	pr_debug("msk=%p", msk);
926bdeab PK	91
	92	/* try to avoid acquiring the lock below */
	93	if (!READ_ONCE(pm->work_pending))
	94	return;
	95
	96	spin_lock_bh(&pm->lock);
	97
	98	if (READ_ONCE(pm->work_pending))
	99	mptcp_pm_schedule_work(msk, MPTCP_PM_ESTABLISHED);
	100
	101	spin_unlock_bh(&pm->lock);
1b1c7a0e PK	102	}
	103
	104	void mptcp_pm_connection_closed(struct mptcp_sock *msk)
	105	{
	106	pr_debug("msk=%p", msk);
	107	}
	108
	109	void mptcp_pm_subflow_established(struct mptcp_sock *msk,
	110	struct mptcp_subflow_context *subflow)
	111	{
926bdeab PK	112	struct mptcp_pm_data *pm = &msk->pm;
926bdeab PK	113
1b1c7a0e	114	pr_debug("msk=%p", msk);
926bdeab PK	115
	116	if (!READ_ONCE(pm->work_pending))
	117	return;
	118
	119	spin_lock_bh(&pm->lock);
	120
	121	if (READ_ONCE(pm->work_pending))
	122	mptcp_pm_schedule_work(msk, MPTCP_PM_SUBFLOW_ESTABLISHED);
	123
	124	spin_unlock_bh(&pm->lock);
1b1c7a0e PK	125	}
	126
	127	void mptcp_pm_subflow_closed(struct mptcp_sock *msk, u8 id)
	128	{
	129	pr_debug("msk=%p", msk);
	130	}
	131
	132	void mptcp_pm_add_addr_received(struct mptcp_sock *msk,
	133	const struct mptcp_addr_info *addr)
	134	{
926bdeab PK	135	struct mptcp_pm_data *pm = &msk->pm;
	136
	137	pr_debug("msk=%p remote_id=%d accept=%d", msk, addr->id,
	138	READ_ONCE(pm->accept_addr));
	139
926bdeab PK	140	spin_lock_bh(&pm->lock);
926bdeab PK	141
6a6c05a8 GT	142	if (!READ_ONCE(pm->accept_addr))
	143	mptcp_pm_announce_addr(msk, addr, true);
	144	else if (mptcp_pm_schedule_work(msk, MPTCP_PM_ADD_ADDR_RECEIVED))
926bdeab PK	145	pm->remote = *addr;
	146
	147	spin_unlock_bh(&pm->lock);
1b1c7a0e PK	148	}
1b1c7a0e PK	149
d0876b22 GT	150	void mptcp_pm_rm_addr_received(struct mptcp_sock *msk, u8 rm_id)
	151	{
	152	struct mptcp_pm_data *pm = &msk->pm;
	153
	154	pr_debug("msk=%p remote_id=%d", msk, rm_id);
	155
	156	spin_lock_bh(&pm->lock);
	157	mptcp_pm_schedule_work(msk, MPTCP_PM_RM_ADDR_RECEIVED);
	158	pm->rm_id = rm_id;
	159	spin_unlock_bh(&pm->lock);
	160	}
	161
1b1c7a0e PK	162	/* path manager helpers */
1b1c7a0e PK	163
f643b803	164	bool mptcp_pm_add_addr_signal(struct mptcp_sock *msk, unsigned int remaining,
6a6c05a8	165	struct mptcp_addr_info saddr, bool echo)
1b1c7a0e	166	{
926bdeab PK	167	int ret = false;
	168
	169	spin_lock_bh(&msk->pm.lock);
	170
	171	/* double check after the lock is acquired */
f643b803	172	if (!mptcp_pm_should_add_signal(msk))
926bdeab PK	173	goto out_unlock;
	174
	175	if (remaining < mptcp_add_addr_len(msk->pm.local.family))
	176	goto out_unlock;
	177
	178	*saddr = msk->pm.local;
6a6c05a8	179	*echo = READ_ONCE(msk->pm.add_addr_echo);
f643b803	180	WRITE_ONCE(msk->pm.add_addr_signal, false);
926bdeab PK	181	ret = true;
	182
	183	out_unlock:
	184	spin_unlock_bh(&msk->pm.lock);
	185	return ret;
1b1c7a0e PK	186	}
1b1c7a0e PK	187
5cb104ae GT	188	bool mptcp_pm_rm_addr_signal(struct mptcp_sock *msk, unsigned int remaining,
	189	u8 *rm_id)
	190	{
	191	int ret = false;
	192
	193	spin_lock_bh(&msk->pm.lock);
	194
	195	/* double check after the lock is acquired */
	196	if (!mptcp_pm_should_rm_signal(msk))
	197	goto out_unlock;
	198
	199	if (remaining < TCPOLEN_MPTCP_RM_ADDR_BASE)
	200	goto out_unlock;
	201
	202	*rm_id = msk->pm.rm_id;
	203	WRITE_ONCE(msk->pm.rm_addr_signal, false);
	204	ret = true;
	205
	206	out_unlock:
	207	spin_unlock_bh(&msk->pm.lock);
	208	return ret;
	209	}
	210
1b1c7a0e PK	211	int mptcp_pm_get_local_id(struct mptcp_sock msk, struct sock_common skc)
1b1c7a0e PK	212	{
01cacb00	213	return mptcp_pm_nl_get_local_id(msk, skc);
1b1c7a0e PK	214	}
1b1c7a0e PK	215
1b1c7a0e PK	216	void mptcp_pm_data_init(struct mptcp_sock *msk)
	217	{
	218	msk->pm.add_addr_signaled = 0;
	219	msk->pm.add_addr_accepted = 0;
	220	msk->pm.local_addr_used = 0;
	221	msk->pm.subflows = 0;
5cb104ae	222	msk->pm.rm_id = 0;
1b1c7a0e	223	WRITE_ONCE(msk->pm.work_pending, false);
f643b803	224	WRITE_ONCE(msk->pm.add_addr_signal, false);
5cb104ae	225	WRITE_ONCE(msk->pm.rm_addr_signal, false);
1b1c7a0e PK	226	WRITE_ONCE(msk->pm.accept_addr, false);
1b1c7a0e PK	227	WRITE_ONCE(msk->pm.accept_subflow, false);
6a6c05a8	228	WRITE_ONCE(msk->pm.add_addr_echo, false);
1b1c7a0e PK	229	msk->pm.status = 0;
	230
	231	spin_lock_init(&msk->pm.lock);
01cacb00 PA	232
01cacb00 PA	233	mptcp_pm_nl_data_init(msk);
1b1c7a0e PK	234	}
1b1c7a0e PK	235
d39dceca	236	void __init mptcp_pm_init(void)
1b1c7a0e	237	{
01cacb00	238	mptcp_pm_nl_init();
1b1c7a0e	239	}