[deliverable/linux.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 <net/sock.h>

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

/* empty to "strongly type" an otherwise void parameter.
 */
struct request_values {
};

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

/* struct request_sock - mini sock to represent a connection request
 */
struct request_sock {
	struct request_sock		*dl_next; /* Must be first member! */
	u16				mss;
	u8				retrans;
	u8				cookie_ts; /* syncookie: encode tcpopts in timestamp */
	/* The following two fields can be easily recomputed I think -AK */
	u32				window_clamp; /* window clamp at creation time */
	u32				rcv_wnd;	  /* rcv_wnd offered first time */
	u32				ts_recent;
	unsigned long			expires;
	const struct request_sock_ops	*rsk_ops;
	struct sock			*sk;
	u32				secid;
	u32				peer_secid;
};

static inline struct request_sock *reqsk_alloc(const struct request_sock_ops *ops)
{
	struct request_sock *req = kmem_cache_alloc(ops->slab, GFP_ATOMIC);

	if (req != NULL)
		req->rsk_ops = ops;

	return req;
}

static inline void __reqsk_free(struct request_sock *req)
{
	kmem_cache_free(req->rsk_ops->slab, req);
}

static inline void reqsk_free(struct request_sock *req)
{
	req->rsk_ops->destructor(req);
	__reqsk_free(req);
}

extern int sysctl_max_syn_backlog;

/** struct listen_sock - listen state
 *
 * @max_qlen_log - log_2 of maximal queued SYNs/REQUESTs
 */
struct listen_sock {
	u8			max_qlen_log;
	u8			synflood_warned;
	/* 2 bytes hole, try to use */
	int			qlen;
	int			qlen_young;
	int			clock_hand;
	u32			hash_rnd;
	u32			nr_table_entries;
	struct request_sock	*syn_table[0];
};

/** 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()
 * @syn_wait_lock - serializer
 *
 * %syn_wait_lock is necessary only to avoid proc interface having to grab the main
 * lock sock while browsing the listening hash (otherwise it's deadlock prone).
 *
 * This lock is acquired in read mode only from listening_get_next() seq_file
 * op and it's acquired in write mode _only_ from code that is actively
 * changing rskq_accept_head. All readers that are holding the master sock lock
 * don't need to grab this lock in read mode too as rskq_accept_head. writes
 * are always protected from the main sock lock.
 */
struct request_sock_queue {
	struct request_sock	*rskq_accept_head;
	struct request_sock	*rskq_accept_tail;
	rwlock_t		syn_wait_lock;
	u8			rskq_defer_accept;
	/* 3 bytes hole, try to pack */
	struct listen_sock	*listen_opt;
};

extern int reqsk_queue_alloc(struct request_sock_queue *queue,
			     unsigned int nr_table_entries);

extern void __reqsk_queue_destroy(struct request_sock_queue *queue);
extern void reqsk_queue_destroy(struct request_sock_queue *queue);

static inline struct request_sock *
	reqsk_queue_yank_acceptq(struct request_sock_queue *queue)
{
	struct request_sock *req = queue->rskq_accept_head;

	queue->rskq_accept_head = NULL;
	return req;
}

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

static inline void reqsk_queue_unlink(struct request_sock_queue *queue,
				      struct request_sock *req,
				      struct request_sock **prev_req)
{
	write_lock(&queue->syn_wait_lock);
	*prev_req = req->dl_next;
	write_unlock(&queue->syn_wait_lock);
}

static inline void reqsk_queue_add(struct request_sock_queue *queue,
				   struct request_sock *req,
				   struct sock *parent,
				   struct sock *child)
{
	req->sk = child;
	sk_acceptq_added(parent);

	if (queue->rskq_accept_head == NULL)
		queue->rskq_accept_head = req;
	else
		queue->rskq_accept_tail->dl_next = req;

	queue->rskq_accept_tail = req;
	req->dl_next = NULL;
}

static inline struct request_sock *reqsk_queue_remove(struct request_sock_queue *queue)
{
	struct request_sock *req = queue->rskq_accept_head;

	WARN_ON(req == NULL);

	queue->rskq_accept_head = req->dl_next;
	if (queue->rskq_accept_head == NULL)
		queue->rskq_accept_tail = NULL;

	return req;
}

static inline struct sock *reqsk_queue_get_child(struct request_sock_queue *queue,
						 struct sock *parent)
{
	struct request_sock *req = reqsk_queue_remove(queue);
	struct sock *child = req->sk;

	WARN_ON(child == NULL);

	sk_acceptq_removed(parent);
	__reqsk_free(req);
	return child;
}

static inline int reqsk_queue_removed(struct request_sock_queue *queue,
				      struct request_sock *req)
{
	struct listen_sock *lopt = queue->listen_opt;

	if (req->retrans == 0)
		--lopt->qlen_young;

	return --lopt->qlen;
}

static inline int reqsk_queue_added(struct request_sock_queue *queue)
{
	struct listen_sock *lopt = queue->listen_opt;
	const int prev_qlen = lopt->qlen;

	lopt->qlen_young++;
	lopt->qlen++;
	return prev_qlen;
}

static inline int reqsk_queue_len(const struct request_sock_queue *queue)
{
	return queue->listen_opt != NULL ? queue->listen_opt->qlen : 0;
}

static inline int reqsk_queue_len_young(const struct request_sock_queue *queue)
{
	return queue->listen_opt->qlen_young;
}

static inline int reqsk_queue_is_full(const struct request_sock_queue *queue)
{
	return queue->listen_opt->qlen >> queue->listen_opt->max_qlen_log;
}

static inline void reqsk_queue_hash_req(struct request_sock_queue *queue,
					u32 hash, struct request_sock *req,
					unsigned long timeout)
{
	struct listen_sock *lopt = queue->listen_opt;

	req->expires = jiffies + timeout;
	req->retrans = 0;
	req->sk = NULL;
	req->dl_next = lopt->syn_table[hash];

	write_lock(&queue->syn_wait_lock);
	lopt->syn_table[hash] = req;
	write_unlock(&queue->syn_wait_lock);
}

#endif /* _REQUEST_SOCK_H */
Commit	Line	Data
2e6599cb ACM	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>
0e87506f	19	#include <linux/spinlock.h>
2e6599cb	20	#include <linux/types.h>
547b792c	21	#include <linux/bug.h>
0e87506f	22
2e6599cb ACM	23	#include <net/sock.h>
2e6599cb ACM	24
60236fdd	25	struct request_sock;
2e6599cb ACM	26	struct sk_buff;
	27	struct dst_entry;
	28	struct proto;
	29
e6b4d113 WAS	30	/* empty to "strongly type" an otherwise void parameter.
	31	*/
	32	struct request_values {
	33	};
	34
60236fdd	35	struct request_sock_ops {
2e6599cb	36	int family;
2e6599cb	37	int obj_size;
e18b890b	38	struct kmem_cache *slab;
7e56b5d6	39	char *slab_name;
2e6599cb	40	int (rtx_syn_ack)(struct sock sk,
e6b4d113 WAS	41	struct request_sock *req,
e6b4d113 WAS	42	struct request_values *rvp);
6edafaaf	43	void (send_ack)(struct sock sk, struct sk_buff *skb,
60236fdd	44	struct request_sock *req);
cfb6eeb4 YH	45	void (send_reset)(struct sock sk,
cfb6eeb4 YH	46	struct sk_buff *skb);
60236fdd	47	void (destructor)(struct request_sock req);
72659ecc OP	48	void (syn_ack_timeout)(struct sock sk,
72659ecc OP	49	struct request_sock *req);
2e6599cb ACM	50	};
2e6599cb ACM	51
60236fdd	52	/* struct request_sock - mini sock to represent a connection request
2e6599cb	53	*/
60236fdd ACM	54	struct request_sock {
60236fdd ACM	55	struct request_sock dl_next; / Must be first member! */
2e6599cb ACM	56	u16 mss;
2e6599cb ACM	57	u8 retrans;
4dfc2817	58	u8 cookie_ts; /* syncookie: encode tcpopts in timestamp */
2e6599cb ACM	59	/* The following two fields can be easily recomputed I think -AK */
	60	u32 window_clamp; /* window clamp at creation time */
	61	u32 rcv_wnd; /* rcv_wnd offered first time */
	62	u32 ts_recent;
	63	unsigned long expires;
72a3effa	64	const struct request_sock_ops *rsk_ops;
2e6599cb	65	struct sock *sk;
4237c75c	66	u32 secid;
6b877699	67	u32 peer_secid;
2e6599cb ACM	68	};
2e6599cb ACM	69
72a3effa	70	static inline struct request_sock reqsk_alloc(const struct request_sock_ops ops)
2e6599cb	71	{
54e6ecb2	72	struct request_sock *req = kmem_cache_alloc(ops->slab, GFP_ATOMIC);
2e6599cb ACM	73
2e6599cb ACM	74	if (req != NULL)
60236fdd	75	req->rsk_ops = ops;
2e6599cb ACM	76
	77	return req;
	78	}
	79
60236fdd	80	static inline void __reqsk_free(struct request_sock *req)
2e6599cb	81	{
60236fdd	82	kmem_cache_free(req->rsk_ops->slab, req);
2e6599cb ACM	83	}
2e6599cb ACM	84
60236fdd	85	static inline void reqsk_free(struct request_sock *req)
2e6599cb	86	{
60236fdd ACM	87	req->rsk_ops->destructor(req);
60236fdd ACM	88	__reqsk_free(req);
2e6599cb ACM	89	}
2e6599cb ACM	90
0e87506f ACM	91	extern int sysctl_max_syn_backlog;
0e87506f ACM	92
2ad69c55	93	/** struct listen_sock - listen state
0e87506f ACM	94	*
	95	* @max_qlen_log - log_2 of maximal queued SYNs/REQUESTs
	96	*/
2ad69c55	97	struct listen_sock {
0e87506f	98	u8 max_qlen_log;
946cedcc ED	99	u8 synflood_warned;
946cedcc ED	100	/* 2 bytes hole, try to use */
0e87506f ACM	101	int qlen;
	102	int qlen_young;
	103	int clock_hand;
	104	u32 hash_rnd;
83e3609e	105	u32 nr_table_entries;
0e87506f ACM	106	struct request_sock *syn_table[0];
	107	};
	108
	109	/** struct request_sock_queue - queue of request_socks
	110	*
	111	* @rskq_accept_head - FIFO head of established children
	112	* @rskq_accept_tail - FIFO tail of established children
295f7324	113	* @rskq_defer_accept - User waits for some data after accept()
0e87506f ACM	114	* @syn_wait_lock - serializer
	115	*
	116	* %syn_wait_lock is necessary only to avoid proc interface having to grab the main
	117	* lock sock while browsing the listening hash (otherwise it's deadlock prone).
	118	*
	119	* This lock is acquired in read mode only from listening_get_next() seq_file
	120	* op and it's acquired in write mode _only_ from code that is actively
	121	* changing rskq_accept_head. All readers that are holding the master sock lock
	122	* don't need to grab this lock in read mode too as rskq_accept_head. writes
	123	* are always protected from the main sock lock.
	124	*/
	125	struct request_sock_queue {
	126	struct request_sock *rskq_accept_head;
	127	struct request_sock *rskq_accept_tail;
	128	rwlock_t syn_wait_lock;
ec0a1966 DM	129	u8 rskq_defer_accept;
ec0a1966 DM	130	/* 3 bytes hole, try to pack */
2ad69c55	131	struct listen_sock *listen_opt;
0e87506f ACM	132	};
	133
	134	extern int reqsk_queue_alloc(struct request_sock_queue *queue,
72a3effa	135	unsigned int nr_table_entries);
0e87506f	136
dab6ba36	137	extern void __reqsk_queue_destroy(struct request_sock_queue *queue);
83e3609e ACM	138	extern void reqsk_queue_destroy(struct request_sock_queue *queue);
83e3609e ACM	139
0e87506f ACM	140	static inline struct request_sock *
	141	reqsk_queue_yank_acceptq(struct request_sock_queue *queue)
	142	{
	143	struct request_sock *req = queue->rskq_accept_head;
	144
3eb4801d	145	queue->rskq_accept_head = NULL;
0e87506f ACM	146	return req;
	147	}
	148
	149	static inline int reqsk_queue_empty(struct request_sock_queue *queue)
	150	{
	151	return queue->rskq_accept_head == NULL;
	152	}
	153
	154	static inline void reqsk_queue_unlink(struct request_sock_queue *queue,
	155	struct request_sock *req,
	156	struct request_sock **prev_req)
	157	{
	158	write_lock(&queue->syn_wait_lock);
	159	*prev_req = req->dl_next;
	160	write_unlock(&queue->syn_wait_lock);
	161	}
	162
	163	static inline void reqsk_queue_add(struct request_sock_queue *queue,
	164	struct request_sock *req,
	165	struct sock *parent,
	166	struct sock *child)
	167	{
	168	req->sk = child;
	169	sk_acceptq_added(parent);
	170
	171	if (queue->rskq_accept_head == NULL)
	172	queue->rskq_accept_head = req;
	173	else
	174	queue->rskq_accept_tail->dl_next = req;
	175
	176	queue->rskq_accept_tail = req;
	177	req->dl_next = NULL;
	178	}
	179
	180	static inline struct request_sock reqsk_queue_remove(struct request_sock_queue queue)
	181	{
	182	struct request_sock *req = queue->rskq_accept_head;
	183
547b792c	184	WARN_ON(req == NULL);
0e87506f ACM	185
	186	queue->rskq_accept_head = req->dl_next;
	187	if (queue->rskq_accept_head == NULL)
	188	queue->rskq_accept_tail = NULL;
	189
	190	return req;
	191	}
	192
	193	static inline struct sock reqsk_queue_get_child(struct request_sock_queue queue,
	194	struct sock *parent)
	195	{
	196	struct request_sock *req = reqsk_queue_remove(queue);
	197	struct sock *child = req->sk;
	198
547b792c	199	WARN_ON(child == NULL);
0e87506f ACM	200
	201	sk_acceptq_removed(parent);
	202	__reqsk_free(req);
	203	return child;
	204	}
	205
	206	static inline int reqsk_queue_removed(struct request_sock_queue *queue,
	207	struct request_sock *req)
	208	{
2ad69c55	209	struct listen_sock *lopt = queue->listen_opt;
0e87506f ACM	210
	211	if (req->retrans == 0)
	212	--lopt->qlen_young;
	213
	214	return --lopt->qlen;
	215	}
	216
	217	static inline int reqsk_queue_added(struct request_sock_queue *queue)
	218	{
2ad69c55	219	struct listen_sock *lopt = queue->listen_opt;
0e87506f ACM	220	const int prev_qlen = lopt->qlen;
	221
	222	lopt->qlen_young++;
	223	lopt->qlen++;
	224	return prev_qlen;
	225	}
	226
463c84b9	227	static inline int reqsk_queue_len(const struct request_sock_queue *queue)
0e87506f ACM	228	{
	229	return queue->listen_opt != NULL ? queue->listen_opt->qlen : 0;
	230	}
	231
463c84b9	232	static inline int reqsk_queue_len_young(const struct request_sock_queue *queue)
0e87506f ACM	233	{
	234	return queue->listen_opt->qlen_young;
	235	}
	236
463c84b9	237	static inline int reqsk_queue_is_full(const struct request_sock_queue *queue)
0e87506f ACM	238	{
	239	return queue->listen_opt->qlen >> queue->listen_opt->max_qlen_log;
	240	}
	241
	242	static inline void reqsk_queue_hash_req(struct request_sock_queue *queue,
	243	u32 hash, struct request_sock *req,
8129765a	244	unsigned long timeout)
0e87506f	245	{
2ad69c55	246	struct listen_sock *lopt = queue->listen_opt;
0e87506f ACM	247
	248	req->expires = jiffies + timeout;
	249	req->retrans = 0;
	250	req->sk = NULL;
	251	req->dl_next = lopt->syn_table[hash];
	252
	253	write_lock(&queue->syn_wait_lock);
	254	lopt->syn_table[hash] = req;
	255	write_unlock(&queue->syn_wait_lock);
	256	}
	257
2e6599cb	258	#endif /* _REQUEST_SOCK_H */