[net-next-2.6.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;

struct request_sock_ops {
	int		family;
	int		obj_size;
	struct kmem_cache	*slab;
	int		(*rtx_syn_ack)(struct sock *sk,
				       struct request_sock *req);
	void		(*send_ack)(struct sk_buff *skb,
				    struct request_sock *req);
	void		(*send_reset)(struct sock *sk,
				      struct sk_buff *skb);
	void		(*destructor)(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;
	/* 3 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
60236fdd	30	struct request_sock_ops {
2e6599cb	31	int family;
2e6599cb	32	int obj_size;
e18b890b	33	struct kmem_cache *slab;
2e6599cb	34	int (rtx_syn_ack)(struct sock sk,
fd80eb94	35	struct request_sock *req);
2e6599cb	36	void (send_ack)(struct sk_buff skb,
60236fdd	37	struct request_sock *req);
cfb6eeb4 YH	38	void (send_reset)(struct sock sk,
cfb6eeb4 YH	39	struct sk_buff *skb);
60236fdd	40	void (destructor)(struct request_sock req);
2e6599cb ACM	41	};
2e6599cb ACM	42
60236fdd	43	/* struct request_sock - mini sock to represent a connection request
2e6599cb	44	*/
60236fdd ACM	45	struct request_sock {
60236fdd ACM	46	struct request_sock dl_next; / Must be first member! */
2e6599cb ACM	47	u16 mss;
2e6599cb ACM	48	u8 retrans;
4dfc2817	49	u8 cookie_ts; /* syncookie: encode tcpopts in timestamp */
2e6599cb ACM	50	/* The following two fields can be easily recomputed I think -AK */
	51	u32 window_clamp; /* window clamp at creation time */
	52	u32 rcv_wnd; /* rcv_wnd offered first time */
	53	u32 ts_recent;
	54	unsigned long expires;
72a3effa	55	const struct request_sock_ops *rsk_ops;
2e6599cb	56	struct sock *sk;
4237c75c	57	u32 secid;
6b877699	58	u32 peer_secid;
2e6599cb ACM	59	};
2e6599cb ACM	60
72a3effa	61	static inline struct request_sock reqsk_alloc(const struct request_sock_ops ops)
2e6599cb	62	{
54e6ecb2	63	struct request_sock *req = kmem_cache_alloc(ops->slab, GFP_ATOMIC);
2e6599cb ACM	64
2e6599cb ACM	65	if (req != NULL)
60236fdd	66	req->rsk_ops = ops;
2e6599cb ACM	67
	68	return req;
	69	}
	70
60236fdd	71	static inline void __reqsk_free(struct request_sock *req)
2e6599cb	72	{
60236fdd	73	kmem_cache_free(req->rsk_ops->slab, req);
2e6599cb ACM	74	}
2e6599cb ACM	75
60236fdd	76	static inline void reqsk_free(struct request_sock *req)
2e6599cb	77	{
60236fdd ACM	78	req->rsk_ops->destructor(req);
60236fdd ACM	79	__reqsk_free(req);
2e6599cb ACM	80	}
2e6599cb ACM	81
0e87506f ACM	82	extern int sysctl_max_syn_backlog;
0e87506f ACM	83
2ad69c55	84	/** struct listen_sock - listen state
0e87506f ACM	85	*
	86	* @max_qlen_log - log_2 of maximal queued SYNs/REQUESTs
	87	*/
2ad69c55	88	struct listen_sock {
0e87506f ACM	89	u8 max_qlen_log;
	90	/* 3 bytes hole, try to use */
	91	int qlen;
	92	int qlen_young;
	93	int clock_hand;
	94	u32 hash_rnd;
83e3609e	95	u32 nr_table_entries;
0e87506f ACM	96	struct request_sock *syn_table[0];
	97	};
	98
	99	/** struct request_sock_queue - queue of request_socks
	100	*
	101	* @rskq_accept_head - FIFO head of established children
	102	* @rskq_accept_tail - FIFO tail of established children
295f7324	103	* @rskq_defer_accept - User waits for some data after accept()
0e87506f ACM	104	* @syn_wait_lock - serializer
	105	*
	106	* %syn_wait_lock is necessary only to avoid proc interface having to grab the main
	107	* lock sock while browsing the listening hash (otherwise it's deadlock prone).
	108	*
	109	* This lock is acquired in read mode only from listening_get_next() seq_file
	110	* op and it's acquired in write mode _only_ from code that is actively
	111	* changing rskq_accept_head. All readers that are holding the master sock lock
	112	* don't need to grab this lock in read mode too as rskq_accept_head. writes
	113	* are always protected from the main sock lock.
	114	*/
	115	struct request_sock_queue {
	116	struct request_sock *rskq_accept_head;
	117	struct request_sock *rskq_accept_tail;
	118	rwlock_t syn_wait_lock;
ec0a1966 DM	119	u8 rskq_defer_accept;
ec0a1966 DM	120	/* 3 bytes hole, try to pack */
2ad69c55	121	struct listen_sock *listen_opt;
0e87506f ACM	122	};
	123
	124	extern int reqsk_queue_alloc(struct request_sock_queue *queue,
72a3effa	125	unsigned int nr_table_entries);
0e87506f	126
dab6ba36	127	extern void __reqsk_queue_destroy(struct request_sock_queue *queue);
83e3609e ACM	128	extern void reqsk_queue_destroy(struct request_sock_queue *queue);
83e3609e ACM	129
0e87506f ACM	130	static inline struct request_sock *
	131	reqsk_queue_yank_acceptq(struct request_sock_queue *queue)
	132	{
	133	struct request_sock *req = queue->rskq_accept_head;
	134
3eb4801d	135	queue->rskq_accept_head = NULL;
0e87506f ACM	136	return req;
	137	}
	138
	139	static inline int reqsk_queue_empty(struct request_sock_queue *queue)
	140	{
	141	return queue->rskq_accept_head == NULL;
	142	}
	143
	144	static inline void reqsk_queue_unlink(struct request_sock_queue *queue,
	145	struct request_sock *req,
	146	struct request_sock **prev_req)
	147	{
	148	write_lock(&queue->syn_wait_lock);
	149	*prev_req = req->dl_next;
	150	write_unlock(&queue->syn_wait_lock);
	151	}
	152
	153	static inline void reqsk_queue_add(struct request_sock_queue *queue,
	154	struct request_sock *req,
	155	struct sock *parent,
	156	struct sock *child)
	157	{
	158	req->sk = child;
	159	sk_acceptq_added(parent);
	160
	161	if (queue->rskq_accept_head == NULL)
	162	queue->rskq_accept_head = req;
	163	else
	164	queue->rskq_accept_tail->dl_next = req;
	165
	166	queue->rskq_accept_tail = req;
	167	req->dl_next = NULL;
	168	}
	169
	170	static inline struct request_sock reqsk_queue_remove(struct request_sock_queue queue)
	171	{
	172	struct request_sock *req = queue->rskq_accept_head;
	173
547b792c	174	WARN_ON(req == NULL);
0e87506f ACM	175
	176	queue->rskq_accept_head = req->dl_next;
	177	if (queue->rskq_accept_head == NULL)
	178	queue->rskq_accept_tail = NULL;
	179
	180	return req;
	181	}
	182
	183	static inline struct sock reqsk_queue_get_child(struct request_sock_queue queue,
	184	struct sock *parent)
	185	{
	186	struct request_sock *req = reqsk_queue_remove(queue);
	187	struct sock *child = req->sk;
	188
547b792c	189	WARN_ON(child == NULL);
0e87506f ACM	190
	191	sk_acceptq_removed(parent);
	192	__reqsk_free(req);
	193	return child;
	194	}
	195
	196	static inline int reqsk_queue_removed(struct request_sock_queue *queue,
	197	struct request_sock *req)
	198	{
2ad69c55	199	struct listen_sock *lopt = queue->listen_opt;
0e87506f ACM	200
	201	if (req->retrans == 0)
	202	--lopt->qlen_young;
	203
	204	return --lopt->qlen;
	205	}
	206
	207	static inline int reqsk_queue_added(struct request_sock_queue *queue)
	208	{
2ad69c55	209	struct listen_sock *lopt = queue->listen_opt;
0e87506f ACM	210	const int prev_qlen = lopt->qlen;
	211
	212	lopt->qlen_young++;
	213	lopt->qlen++;
	214	return prev_qlen;
	215	}
	216
463c84b9	217	static inline int reqsk_queue_len(const struct request_sock_queue *queue)
0e87506f ACM	218	{
	219	return queue->listen_opt != NULL ? queue->listen_opt->qlen : 0;
	220	}
	221
463c84b9	222	static inline int reqsk_queue_len_young(const struct request_sock_queue *queue)
0e87506f ACM	223	{
	224	return queue->listen_opt->qlen_young;
	225	}
	226
463c84b9	227	static inline int reqsk_queue_is_full(const struct request_sock_queue *queue)
0e87506f ACM	228	{
	229	return queue->listen_opt->qlen >> queue->listen_opt->max_qlen_log;
	230	}
	231
	232	static inline void reqsk_queue_hash_req(struct request_sock_queue *queue,
	233	u32 hash, struct request_sock *req,
8129765a	234	unsigned long timeout)
0e87506f	235	{
2ad69c55	236	struct listen_sock *lopt = queue->listen_opt;
0e87506f ACM	237
	238	req->expires = jiffies + timeout;
	239	req->retrans = 0;
	240	req->sk = NULL;
	241	req->dl_next = lopt->syn_table[hash];
	242
	243	write_lock(&queue->syn_wait_lock);
	244	lopt->syn_table[hash] = req;
	245	write_unlock(&queue->syn_wait_lock);
	246	}
	247
2e6599cb	248	#endif /* _REQUEST_SOCK_H */