[linux-block.git] / include / net / request_sock.h

/*
 * NET		Generic infrastructure for Network protocols.
 *
 *		Definitions for request_sock
 *
 * Authors:	Arnaldo Carvalho de Melo <acme@conectiva.com.br>
 *
 * 		From code originally in include/net/tcp.h
 *
 *		This program is free software; you can redistribute it and/or
 *		modify it under the terms of the GNU General Public License
 *		as published by the Free Software Foundation; either version
 *		2 of the License, or (at your option) any later version.
 */
#ifndef _REQUEST_SOCK_H
#define _REQUEST_SOCK_H

#include <linux/slab.h>
#include <linux/spinlock.h>
#include <linux/types.h>
#include <linux/bug.h>
#include <linux/refcount.h>

#include <net/sock.h>

struct request_sock;
struct sk_buff;
struct dst_entry;
struct proto;

struct request_sock_ops {
	int		family;
	unsigned int	obj_size;
	struct kmem_cache	*slab;
	char		*slab_name;
	int		(*rtx_syn_ack)(const struct sock *sk,
				       struct request_sock *req);
	void		(*send_ack)(const struct sock *sk, struct sk_buff *skb,
				    struct request_sock *req);
	void		(*send_reset)(const struct sock *sk,
				      struct sk_buff *skb);
	void		(*destructor)(struct request_sock *req);
	void		(*syn_ack_timeout)(const struct request_sock *req);
};

int inet_rtx_syn_ack(const struct sock *parent, struct request_sock *req);

/* struct request_sock - mini sock to represent a connection request
 */
struct request_sock {
	struct sock_common		__req_common;
#define rsk_refcnt			__req_common.skc_refcnt
#define rsk_hash			__req_common.skc_hash
#define rsk_listener			__req_common.skc_listener
#define rsk_window_clamp		__req_common.skc_window_clamp
#define rsk_rcv_wnd			__req_common.skc_rcv_wnd

	struct request_sock		*dl_next;
	u16				mss;
	u8				num_retrans; /* number of retransmits */
	u8				cookie_ts:1; /* syncookie: encode tcpopts in timestamp */
	u8				num_timeout:7; /* number of timeouts */
	u32				ts_recent;
	struct timer_list		rsk_timer;
	const struct request_sock_ops	*rsk_ops;
	struct sock			*sk;
	u32				*saved_syn;
	u32				secid;
	u32				peer_secid;
};

static inline struct request_sock *inet_reqsk(const struct sock *sk)
{
	return (struct request_sock *)sk;
}

static inline struct sock *req_to_sk(struct request_sock *req)
{
	return (struct sock *)req;
}

static inline struct request_sock *
reqsk_alloc(const struct request_sock_ops *ops, struct sock *sk_listener,
	    bool attach_listener)
{
	struct request_sock *req;

	req = kmem_cache_alloc(ops->slab, GFP_ATOMIC | __GFP_NOWARN);
	if (!req)
		return NULL;
	req->rsk_listener = NULL;
	if (attach_listener) {
		if (unlikely(!refcount_inc_not_zero(&sk_listener->sk_refcnt))) {
			kmem_cache_free(ops->slab, req);
			return NULL;
		}
		req->rsk_listener = sk_listener;
	}
	req->rsk_ops = ops;
	req_to_sk(req)->sk_prot = sk_listener->sk_prot;
	sk_node_init(&req_to_sk(req)->sk_node);
	sk_tx_queue_clear(req_to_sk(req));
	req->saved_syn = NULL;
	refcount_set(&req->rsk_refcnt, 0);

	return req;
}

static inline void reqsk_free(struct request_sock *req)
{
	/* temporary debugging */
	WARN_ON_ONCE(refcount_read(&req->rsk_refcnt) != 0);

	req->rsk_ops->destructor(req);
	if (req->rsk_listener)
		sock_put(req->rsk_listener);
	kfree(req->saved_syn);
	kmem_cache_free(req->rsk_ops->slab, req);
}

static inline void reqsk_put(struct request_sock *req)
{
	if (refcount_dec_and_test(&req->rsk_refcnt))
		reqsk_free(req);
}

/*
 * For a TCP Fast Open listener -
 *	lock - protects the access to all the reqsk, which is co-owned by
 *		the listener and the child socket.
 *	qlen - pending TFO requests (still in TCP_SYN_RECV).
 *	max_qlen - max TFO reqs allowed before TFO is disabled.
 *
 *	XXX (TFO) - ideally these fields can be made as part of "listen_sock"
 *	structure above. But there is some implementation difficulty due to
 *	listen_sock being part of request_sock_queue hence will be freed when
 *	a listener is stopped. But TFO related fields may continue to be
 *	accessed even after a listener is closed, until its sk_refcnt drops
 *	to 0 implying no more outstanding TFO reqs. One solution is to keep
 *	listen_opt around until	sk_refcnt drops to 0. But there is some other
 *	complexity that needs to be resolved. E.g., a listener can be disabled
 *	temporarily through shutdown()->tcp_disconnect(), and re-enabled later.
 */
struct fastopen_queue {
	struct request_sock	*rskq_rst_head; /* Keep track of past TFO */
	struct request_sock	*rskq_rst_tail; /* requests that caused RST.
						 * This is part of the defense
						 * against spoofing attack.
						 */
	spinlock_t	lock;
	int		qlen;		/* # of pending (TCP_SYN_RECV) reqs */
	int		max_qlen;	/* != 0 iff TFO is currently enabled */

	struct tcp_fastopen_context __rcu *ctx; /* cipher context for cookie */
};

/** struct request_sock_queue - queue of request_socks
 *
 * @rskq_accept_head - FIFO head of established children
 * @rskq_accept_tail - FIFO tail of established children
 * @rskq_defer_accept - User waits for some data after accept()
 *
 */
struct request_sock_queue {
	spinlock_t		rskq_lock;
	u8			rskq_defer_accept;

	u32			synflood_warned;
	atomic_t		qlen;
	atomic_t		young;

	struct request_sock	*rskq_accept_head;
	struct request_sock	*rskq_accept_tail;
	struct fastopen_queue	fastopenq;  /* Check max_qlen != 0 to determine
					     * if TFO is enabled.
					     */
};

void reqsk_queue_alloc(struct request_sock_queue *queue);

void reqsk_fastopen_remove(struct sock *sk, struct request_sock *req,
			   bool reset);

static inline bool reqsk_queue_empty(const struct request_sock_queue *queue)
{
	return queue->rskq_accept_head == NULL;
}

static inline struct request_sock *reqsk_queue_remove(struct request_sock_queue *queue,
						      struct sock *parent)
{
	struct request_sock *req;

	spin_lock_bh(&queue->rskq_lock);
	req = queue->rskq_accept_head;
	if (req) {
		sk_acceptq_removed(parent);
		queue->rskq_accept_head = req->dl_next;
		if (queue->rskq_accept_head == NULL)
			queue->rskq_accept_tail = NULL;
	}
	spin_unlock_bh(&queue->rskq_lock);
	return req;
}

static inline void reqsk_queue_removed(struct request_sock_queue *queue,
				       const struct request_sock *req)
{
	if (req->num_timeout == 0)
		atomic_dec(&queue->young);
	atomic_dec(&queue->qlen);
}

static inline void reqsk_queue_added(struct request_sock_queue *queue)
{
	atomic_inc(&queue->young);
	atomic_inc(&queue->qlen);
}

static inline int reqsk_queue_len(const struct request_sock_queue *queue)
{
	return atomic_read(&queue->qlen);
}

static inline int reqsk_queue_len_young(const struct request_sock_queue *queue)
{
	return atomic_read(&queue->young);
}

#endif /* _REQUEST_SOCK_H */
Commit	Line	Data
	1	/*
	2	* NET Generic infrastructure for Network protocols.
	3	*
	4	* Definitions for request_sock
	5	*
	6	* Authors: Arnaldo Carvalho de Melo <acme@conectiva.com.br>
	7	*
	8	* From code originally in include/net/tcp.h
	9	*
	10	* This program is free software; you can redistribute it and/or
	11	* modify it under the terms of the GNU General Public License
	12	* as published by the Free Software Foundation; either version
	13	* 2 of the License, or (at your option) any later version.
	14	*/
	15	#ifndef _REQUEST_SOCK_H
	16	#define _REQUEST_SOCK_H
	17
	18	#include <linux/slab.h>
	19	#include <linux/spinlock.h>
	20	#include <linux/types.h>
	21	#include <linux/bug.h>
	22	#include <linux/refcount.h>
	23
	24	#include <net/sock.h>
	25
	26	struct request_sock;
	27	struct sk_buff;
	28	struct dst_entry;
	29	struct proto;
	30
	31	struct request_sock_ops {
	32	int family;
	33	unsigned int obj_size;
	34	struct kmem_cache *slab;
	35	char *slab_name;
	36	int (rtx_syn_ack)(const struct sock sk,
	37	struct request_sock *req);
	38	void (send_ack)(const struct sock sk, struct sk_buff *skb,
	39	struct request_sock *req);
	40	void (send_reset)(const struct sock sk,
	41	struct sk_buff *skb);
	42	void (destructor)(struct request_sock req);
	43	void (syn_ack_timeout)(const struct request_sock req);
	44	};
	45
	46	int inet_rtx_syn_ack(const struct sock parent, struct request_sock req);
	47
	48	/* struct request_sock - mini sock to represent a connection request
	49	*/
	50	struct request_sock {
	51	struct sock_common __req_common;
	52	#define rsk_refcnt __req_common.skc_refcnt
	53	#define rsk_hash __req_common.skc_hash
	54	#define rsk_listener __req_common.skc_listener
	55	#define rsk_window_clamp __req_common.skc_window_clamp
	56	#define rsk_rcv_wnd __req_common.skc_rcv_wnd
	57
	58	struct request_sock *dl_next;
	59	u16 mss;
	60	u8 num_retrans; /* number of retransmits */
	61	u8 cookie_ts:1; /* syncookie: encode tcpopts in timestamp */
	62	u8 num_timeout:7; /* number of timeouts */
	63	u32 ts_recent;
	64	struct timer_list rsk_timer;
	65	const struct request_sock_ops *rsk_ops;
	66	struct sock *sk;
	67	u32 *saved_syn;
	68	u32 secid;
	69	u32 peer_secid;
	70	};
	71
	72	static inline struct request_sock inet_reqsk(const struct sock sk)
	73	{
	74	return (struct request_sock *)sk;
	75	}
	76
	77	static inline struct sock req_to_sk(struct request_sock req)
	78	{
	79	return (struct sock *)req;
	80	}
	81
	82	static inline struct request_sock *
	83	reqsk_alloc(const struct request_sock_ops ops, struct sock sk_listener,
	84	bool attach_listener)
	85	{
	86	struct request_sock *req;
	87
	88	req = kmem_cache_alloc(ops->slab, GFP_ATOMIC \| __GFP_NOWARN);
	89	if (!req)
	90	return NULL;
	91	req->rsk_listener = NULL;
	92	if (attach_listener) {
	93	if (unlikely(!refcount_inc_not_zero(&sk_listener->sk_refcnt))) {
	94	kmem_cache_free(ops->slab, req);
	95	return NULL;
	96	}
	97	req->rsk_listener = sk_listener;
	98	}
	99	req->rsk_ops = ops;
	100	req_to_sk(req)->sk_prot = sk_listener->sk_prot;
	101	sk_node_init(&req_to_sk(req)->sk_node);
	102	sk_tx_queue_clear(req_to_sk(req));
	103	req->saved_syn = NULL;
	104	refcount_set(&req->rsk_refcnt, 0);
	105
	106	return req;
	107	}
	108
	109	static inline void reqsk_free(struct request_sock *req)
	110	{
	111	/* temporary debugging */
	112	WARN_ON_ONCE(refcount_read(&req->rsk_refcnt) != 0);
	113
	114	req->rsk_ops->destructor(req);
	115	if (req->rsk_listener)
	116	sock_put(req->rsk_listener);
	117	kfree(req->saved_syn);
	118	kmem_cache_free(req->rsk_ops->slab, req);
	119	}
	120
	121	static inline void reqsk_put(struct request_sock *req)
	122	{
	123	if (refcount_dec_and_test(&req->rsk_refcnt))
	124	reqsk_free(req);
	125	}
	126
	127	/*
	128	* For a TCP Fast Open listener -
	129	* lock - protects the access to all the reqsk, which is co-owned by
	130	* the listener and the child socket.
	131	* qlen - pending TFO requests (still in TCP_SYN_RECV).
	132	* max_qlen - max TFO reqs allowed before TFO is disabled.
	133	*
	134	* XXX (TFO) - ideally these fields can be made as part of "listen_sock"
	135	* structure above. But there is some implementation difficulty due to
	136	* listen_sock being part of request_sock_queue hence will be freed when
	137	* a listener is stopped. But TFO related fields may continue to be
	138	* accessed even after a listener is closed, until its sk_refcnt drops
	139	* to 0 implying no more outstanding TFO reqs. One solution is to keep
	140	* listen_opt around until sk_refcnt drops to 0. But there is some other
	141	* complexity that needs to be resolved. E.g., a listener can be disabled
	142	* temporarily through shutdown()->tcp_disconnect(), and re-enabled later.
	143	*/
	144	struct fastopen_queue {
	145	struct request_sock rskq_rst_head; / Keep track of past TFO */
	146	struct request_sock rskq_rst_tail; / requests that caused RST.
	147	* This is part of the defense
	148	* against spoofing attack.
	149	*/
	150	spinlock_t lock;
	151	int qlen; /* # of pending (TCP_SYN_RECV) reqs */
	152	int max_qlen; /* != 0 iff TFO is currently enabled */
	153
	154	struct tcp_fastopen_context __rcu ctx; / cipher context for cookie */
	155	};
	156
	157	/** struct request_sock_queue - queue of request_socks
	158	*
	159	* @rskq_accept_head - FIFO head of established children
	160	* @rskq_accept_tail - FIFO tail of established children
	161	* @rskq_defer_accept - User waits for some data after accept()
	162	*
	163	*/
	164	struct request_sock_queue {
	165	spinlock_t rskq_lock;
	166	u8 rskq_defer_accept;
	167
	168	u32 synflood_warned;
	169	atomic_t qlen;
	170	atomic_t young;
	171
	172	struct request_sock *rskq_accept_head;
	173	struct request_sock *rskq_accept_tail;
	174	struct fastopen_queue fastopenq; /* Check max_qlen != 0 to determine
	175	* if TFO is enabled.
	176	*/
	177	};
	178
	179	void reqsk_queue_alloc(struct request_sock_queue *queue);
	180
	181	void reqsk_fastopen_remove(struct sock sk, struct request_sock req,
	182	bool reset);
	183
	184	static inline bool reqsk_queue_empty(const struct request_sock_queue *queue)
	185	{
	186	return queue->rskq_accept_head == NULL;
	187	}
	188
	189	static inline struct request_sock reqsk_queue_remove(struct request_sock_queue queue,
	190	struct sock *parent)
	191	{
	192	struct request_sock *req;
	193
	194	spin_lock_bh(&queue->rskq_lock);
	195	req = queue->rskq_accept_head;
	196	if (req) {
	197	sk_acceptq_removed(parent);
	198	queue->rskq_accept_head = req->dl_next;
	199	if (queue->rskq_accept_head == NULL)
	200	queue->rskq_accept_tail = NULL;
	201	}
	202	spin_unlock_bh(&queue->rskq_lock);
	203	return req;
	204	}
	205
	206	static inline void reqsk_queue_removed(struct request_sock_queue *queue,
	207	const struct request_sock *req)
	208	{
	209	if (req->num_timeout == 0)
	210	atomic_dec(&queue->young);
	211	atomic_dec(&queue->qlen);
	212	}
	213
	214	static inline void reqsk_queue_added(struct request_sock_queue *queue)
	215	{
	216	atomic_inc(&queue->young);
	217	atomic_inc(&queue->qlen);
	218	}
	219
	220	static inline int reqsk_queue_len(const struct request_sock_queue *queue)
	221	{
	222	return atomic_read(&queue->qlen);
	223	}
	224
	225	static inline int reqsk_queue_len_young(const struct request_sock_queue *queue)
	226	{
	227	return atomic_read(&queue->young);
	228	}
	229
	230	#endif /* _REQUEST_SOCK_H */