[net-next-2.6.git] / include / net / dst.h

/*
 * net/dst.h	Protocol independent destination cache definitions.
 *
 * Authors:	Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru>
 *
 */

#ifndef _NET_DST_H
#define _NET_DST_H

#include <net/dst_ops.h>
#include <linux/netdevice.h>
#include <linux/rtnetlink.h>
#include <linux/rcupdate.h>
#include <linux/jiffies.h>
#include <net/neighbour.h>
#include <asm/processor.h>

/*
 * 0 - no debugging messages
 * 1 - rare events and bugs (default)
 * 2 - trace mode.
 */
#define RT_CACHE_DEBUG		0

#define DST_GC_MIN	(HZ/10)
#define DST_GC_INC	(HZ/2)
#define DST_GC_MAX	(120*HZ)

/* Each dst_entry has reference count and sits in some parent list(s).
 * When it is removed from parent list, it is "freed" (dst_free).
 * After this it enters dead state (dst->obsolete > 0) and if its refcnt
 * is zero, it can be destroyed immediately, otherwise it is added
 * to gc list and garbage collector periodically checks the refcnt.
 */

struct sk_buff;

struct dst_entry {
	struct rcu_head		rcu_head;
	struct dst_entry	*child;
	struct net_device       *dev;
	short			error;
	short			obsolete;
	int			flags;
#define DST_HOST		1
#define DST_NOXFRM		2
#define DST_NOPOLICY		4
#define DST_NOHASH		8
	unsigned long		expires;

	unsigned short		header_len;	/* more space at head required */
	unsigned short		trailer_len;	/* space to reserve at tail */

	unsigned int		rate_tokens;
	unsigned long		rate_last;	/* rate limiting for ICMP */

	struct dst_entry	*path;

	struct neighbour	*neighbour;
	struct hh_cache		*hh;
#ifdef CONFIG_XFRM
	struct xfrm_state	*xfrm;
#else
	void			*__pad1;
#endif
	int			(*input)(struct sk_buff*);
	int			(*output)(struct sk_buff*);

	struct  dst_ops	        *ops;

	u32			metrics[RTAX_MAX];

#ifdef CONFIG_NET_CLS_ROUTE
	__u32			tclassid;
#else
	__u32			__pad2;
#endif


	/*
	 * Align __refcnt to a 64 bytes alignment
	 * (L1_CACHE_SIZE would be too much)
	 */
#ifdef CONFIG_64BIT
	long			__pad_to_align_refcnt[1];
#endif
	/*
	 * __refcnt wants to be on a different cache line from
	 * input/output/ops or performance tanks badly
	 */
	atomic_t		__refcnt;	/* client references	*/
	int			__use;
	unsigned long		lastuse;
	union {
		struct dst_entry *next;
		struct rtable    *rt_next;
		struct rt6_info   *rt6_next;
		struct dn_route  *dn_next;
	};
};

#ifdef __KERNEL__

static inline u32
dst_metric(const struct dst_entry *dst, int metric)
{
	return dst->metrics[metric-1];
}

static inline u32
dst_feature(const struct dst_entry *dst, u32 feature)
{
	return dst_metric(dst, RTAX_FEATURES) & feature;
}

static inline u32 dst_mtu(const struct dst_entry *dst)
{
	u32 mtu = dst_metric(dst, RTAX_MTU);
	/*
	 * Alexey put it here, so ask him about it :)
	 */
	barrier();
	return mtu;
}

/* RTT metrics are stored in milliseconds for user ABI, but used as jiffies */
static inline unsigned long dst_metric_rtt(const struct dst_entry *dst, int metric)
{
	return msecs_to_jiffies(dst_metric(dst, metric));
}

static inline void set_dst_metric_rtt(struct dst_entry *dst, int metric,
				      unsigned long rtt)
{
	dst->metrics[metric-1] = jiffies_to_msecs(rtt);
}

static inline u32
dst_allfrag(const struct dst_entry *dst)
{
	int ret = dst_feature(dst,  RTAX_FEATURE_ALLFRAG);
	/* Yes, _exactly_. This is paranoia. */
	barrier();
	return ret;
}

static inline int
dst_metric_locked(struct dst_entry *dst, int metric)
{
	return dst_metric(dst, RTAX_LOCK) & (1<<metric);
}

static inline void dst_hold(struct dst_entry * dst)
{
	/*
	 * If your kernel compilation stops here, please check
	 * __pad_to_align_refcnt declaration in struct dst_entry
	 */
	BUILD_BUG_ON(offsetof(struct dst_entry, __refcnt) & 63);
	atomic_inc(&dst->__refcnt);
}

static inline void dst_use(struct dst_entry *dst, unsigned long time)
{
	dst_hold(dst);
	dst->__use++;
	dst->lastuse = time;
}

static inline void dst_use_noref(struct dst_entry *dst, unsigned long time)
{
	dst->__use++;
	dst->lastuse = time;
}

static inline
struct dst_entry * dst_clone(struct dst_entry * dst)
{
	if (dst)
		atomic_inc(&dst->__refcnt);
	return dst;
}

extern void dst_release(struct dst_entry *dst);

static inline void refdst_drop(unsigned long refdst)
{
	if (!(refdst & SKB_DST_NOREF))
		dst_release((struct dst_entry *)(refdst & SKB_DST_PTRMASK));
}

/**
 * skb_dst_drop - drops skb dst
 * @skb: buffer
 *
 * Drops dst reference count if a reference was taken.
 */
static inline void skb_dst_drop(struct sk_buff *skb)
{
	if (skb->_skb_refdst) {
		refdst_drop(skb->_skb_refdst);
		skb->_skb_refdst = 0UL;
	}
}

static inline void skb_dst_copy(struct sk_buff *nskb, const struct sk_buff *oskb)
{
	nskb->_skb_refdst = oskb->_skb_refdst;
	if (!(nskb->_skb_refdst & SKB_DST_NOREF))
		dst_clone(skb_dst(nskb));
}

/**
 * skb_dst_force - makes sure skb dst is refcounted
 * @skb: buffer
 *
 * If dst is not yet refcounted, let's do it
 */
static inline void skb_dst_force(struct sk_buff *skb)
{
	if (skb_dst_is_noref(skb)) {
		WARN_ON(!rcu_read_lock_held());
		skb->_skb_refdst &= ~SKB_DST_NOREF;
		dst_clone(skb_dst(skb));
	}
}


/**
 *	skb_tunnel_rx - prepare skb for rx reinsert
 *	@skb: buffer
 *	@dev: tunnel device
 *
 *	After decapsulation, packet is going to re-enter (netif_rx()) our stack,
 *	so make some cleanups, and perform accounting.
 */
static inline void skb_tunnel_rx(struct sk_buff *skb, struct net_device *dev)
{
	skb->dev = dev;
	/* TODO : stats should be SMP safe */
	dev->stats.rx_packets++;
	dev->stats.rx_bytes += skb->len;
	skb->rxhash = 0;
	skb_dst_drop(skb);
	nf_reset(skb);
}

/* Children define the path of the packet through the
 * Linux networking.  Thus, destinations are stackable.
 */

static inline struct dst_entry *skb_dst_pop(struct sk_buff *skb)
{
	struct dst_entry *child = skb_dst(skb)->child;

	skb_dst_drop(skb);
	return child;
}

extern int dst_discard(struct sk_buff *skb);
extern void * dst_alloc(struct dst_ops * ops);
extern void __dst_free(struct dst_entry * dst);
extern struct dst_entry *dst_destroy(struct dst_entry * dst);

static inline void dst_free(struct dst_entry * dst)
{
	if (dst->obsolete > 1)
		return;
	if (!atomic_read(&dst->__refcnt)) {
		dst = dst_destroy(dst);
		if (!dst)
			return;
	}
	__dst_free(dst);
}

static inline void dst_rcu_free(struct rcu_head *head)
{
	struct dst_entry *dst = container_of(head, struct dst_entry, rcu_head);
	dst_free(dst);
}

static inline void dst_confirm(struct dst_entry *dst)
{
	if (dst)
		neigh_confirm(dst->neighbour);
}

static inline void dst_link_failure(struct sk_buff *skb)
{
	struct dst_entry *dst = skb_dst(skb);
	if (dst && dst->ops && dst->ops->link_failure)
		dst->ops->link_failure(skb);
}

static inline void dst_set_expires(struct dst_entry *dst, int timeout)
{
	unsigned long expires = jiffies + timeout;

	if (expires == 0)
		expires = 1;

	if (dst->expires == 0 || time_before(expires, dst->expires))
		dst->expires = expires;
}

/* Output packet to network from transport.  */
static inline int dst_output(struct sk_buff *skb)
{
	return skb_dst(skb)->output(skb);
}

/* Input packet from network to transport.  */
static inline int dst_input(struct sk_buff *skb)
{
	return skb_dst(skb)->input(skb);
}

static inline struct dst_entry *dst_check(struct dst_entry *dst, u32 cookie)
{
	if (dst->obsolete)
		dst = dst->ops->check(dst, cookie);
	return dst;
}

extern void		dst_init(void);

/* Flags for xfrm_lookup flags argument. */
enum {
	XFRM_LOOKUP_WAIT = 1 << 0,
	XFRM_LOOKUP_ICMP = 1 << 1,
};

struct flowi;
#ifndef CONFIG_XFRM
static inline int xfrm_lookup(struct net *net, struct dst_entry **dst_p,
			      struct flowi *fl, struct sock *sk, int flags)
{
	return 0;
} 
static inline int __xfrm_lookup(struct net *net, struct dst_entry **dst_p,
				struct flowi *fl, struct sock *sk, int flags)
{
	return 0;
}
#else
extern int xfrm_lookup(struct net *net, struct dst_entry **dst_p,
		       struct flowi *fl, struct sock *sk, int flags);
extern int __xfrm_lookup(struct net *net, struct dst_entry **dst_p,
			 struct flowi *fl, struct sock *sk, int flags);
#endif
#endif

#endif /* _NET_DST_H */
Commit	Line	Data
1da177e4 LT	1	/*
	2	* net/dst.h Protocol independent destination cache definitions.
	3	*
	4	* Authors: Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru>
	5	*
	6	*/
	7
	8	#ifndef _NET_DST_H
	9	#define _NET_DST_H
	10
86393e52	11	#include <net/dst_ops.h>
14c85021	12	#include <linux/netdevice.h>
1da177e4 LT	13	#include <linux/rtnetlink.h>
	14	#include <linux/rcupdate.h>
	15	#include <linux/jiffies.h>
	16	#include <net/neighbour.h>
	17	#include <asm/processor.h>
	18
	19	/*
	20	* 0 - no debugging messages
	21	* 1 - rare events and bugs (default)
	22	* 2 - trace mode.
	23	*/
	24	#define RT_CACHE_DEBUG 0
	25
	26	#define DST_GC_MIN (HZ/10)
	27	#define DST_GC_INC (HZ/2)
	28	#define DST_GC_MAX (120*HZ)
	29
	30	/* Each dst_entry has reference count and sits in some parent list(s).
	31	* When it is removed from parent list, it is "freed" (dst_free).
	32	* After this it enters dead state (dst->obsolete > 0) and if its refcnt
	33	* is zero, it can be destroyed immediately, otherwise it is added
	34	* to gc list and garbage collector periodically checks the refcnt.
	35	*/
	36
	37	struct sk_buff;
	38
fd2c3ef7	39	struct dst_entry {
1e19e02c	40	struct rcu_head rcu_head;
1da177e4 LT	41	struct dst_entry *child;
1da177e4 LT	42	struct net_device *dev;
c4d54110 HX	43	short error;
c4d54110 HX	44	short obsolete;
1da177e4 LT	45	int flags;
	46	#define DST_HOST 1
	47	#define DST_NOXFRM 2
	48	#define DST_NOPOLICY 4
	49	#define DST_NOHASH 8
1da177e4 LT	50	unsigned long expires;
	51
	52	unsigned short header_len; /* more space at head required */
	53	unsigned short trailer_len; /* space to reserve at tail */
	54
69a73829	55	unsigned int rate_tokens;
f1dd9c37	56	unsigned long rate_last; /* rate limiting for ICMP */
69a73829	57
f1dd9c37	58	struct dst_entry *path;
1da177e4	59
1da177e4 LT	60	struct neighbour *neighbour;
1da177e4 LT	61	struct hh_cache *hh;
def8b4fa	62	#ifdef CONFIG_XFRM
1da177e4	63	struct xfrm_state *xfrm;
5635c10d ED	64	#else
5635c10d ED	65	void *__pad1;
def8b4fa	66	#endif
1da177e4 LT	67	int (input)(struct sk_buff);
	68	int (output)(struct sk_buff);
	69
1da177e4	70	struct dst_ops *ops;
f1dd9c37 ZY	71
	72	u32 metrics[RTAX_MAX];
	73
	74	#ifdef CONFIG_NET_CLS_ROUTE
	75	__u32 tclassid;
5635c10d ED	76	#else
5635c10d ED	77	__u32 __pad2;
f1dd9c37 ZY	78	#endif
f1dd9c37 ZY	79
5635c10d ED	80
	81	/*
	82	* Align __refcnt to a 64 bytes alignment
	83	* (L1_CACHE_SIZE would be too much)
	84	*/
	85	#ifdef CONFIG_64BIT
5635c10d ED	86	long __pad_to_align_refcnt[1];
5635c10d ED	87	#endif
f1dd9c37 ZY	88	/*
	89	* __refcnt wants to be on a different cache line from
	90	* input/output/ops or performance tanks badly
	91	*/
1e19e02c ED	92	atomic_t __refcnt; /* client references */
1e19e02c ED	93	int __use;
f1dd9c37	94	unsigned long lastuse;
1e19e02c ED	95	union {
	96	struct dst_entry *next;
	97	struct rtable *rt_next;
	98	struct rt6_info *rt6_next;
	99	struct dn_route *dn_next;
	100	};
1da177e4 LT	101	};
1da177e4 LT	102
1da177e4 LT	103	#ifdef __KERNEL__
	104
	105	static inline u32
	106	dst_metric(const struct dst_entry *dst, int metric)
	107	{
	108	return dst->metrics[metric-1];
	109	}
	110
0c3adfb8 GBY	111	static inline u32
	112	dst_feature(const struct dst_entry *dst, u32 feature)
	113	{
bb5b7c11	114	return dst_metric(dst, RTAX_FEATURES) & feature;
0c3adfb8 GBY	115	}
0c3adfb8 GBY	116
1da177e4 LT	117	static inline u32 dst_mtu(const struct dst_entry *dst)
	118	{
	119	u32 mtu = dst_metric(dst, RTAX_MTU);
	120	/*
	121	* Alexey put it here, so ask him about it :)
	122	*/
	123	barrier();
	124	return mtu;
	125	}
	126
c1e20f7c SH	127	/* RTT metrics are stored in milliseconds for user ABI, but used as jiffies */
	128	static inline unsigned long dst_metric_rtt(const struct dst_entry *dst, int metric)
	129	{
	130	return msecs_to_jiffies(dst_metric(dst, metric));
	131	}
	132
	133	static inline void set_dst_metric_rtt(struct dst_entry *dst, int metric,
	134	unsigned long rtt)
	135	{
	136	dst->metrics[metric-1] = jiffies_to_msecs(rtt);
	137	}
	138
1da177e4 LT	139	static inline u32
	140	dst_allfrag(const struct dst_entry *dst)
	141	{
0c3adfb8	142	int ret = dst_feature(dst, RTAX_FEATURE_ALLFRAG);
1da177e4 LT	143	/* Yes, _exactly_. This is paranoia. */
	144	barrier();
	145	return ret;
	146	}
	147
	148	static inline int
	149	dst_metric_locked(struct dst_entry *dst, int metric)
	150	{
	151	return dst_metric(dst, RTAX_LOCK) & (1<<metric);
	152	}
	153
	154	static inline void dst_hold(struct dst_entry * dst)
	155	{
5635c10d ED	156	/*
	157	* If your kernel compilation stops here, please check
	158	* __pad_to_align_refcnt declaration in struct dst_entry
	159	*/
	160	BUILD_BUG_ON(offsetof(struct dst_entry, __refcnt) & 63);
1da177e4 LT	161	atomic_inc(&dst->__refcnt);
	162	}
	163
03f49f34 PE	164	static inline void dst_use(struct dst_entry *dst, unsigned long time)
	165	{
	166	dst_hold(dst);
	167	dst->__use++;
	168	dst->lastuse = time;
	169	}
	170
7fee226a ED	171	static inline void dst_use_noref(struct dst_entry *dst, unsigned long time)
	172	{
	173	dst->__use++;
	174	dst->lastuse = time;
	175	}
	176
1da177e4 LT	177	static inline
	178	struct dst_entry * dst_clone(struct dst_entry * dst)
	179	{
	180	if (dst)
	181	atomic_inc(&dst->__refcnt);
	182	return dst;
	183	}
	184
8d330868	185	extern void dst_release(struct dst_entry *dst);
7fee226a ED	186
	187	static inline void refdst_drop(unsigned long refdst)
	188	{
	189	if (!(refdst & SKB_DST_NOREF))
	190	dst_release((struct dst_entry *)(refdst & SKB_DST_PTRMASK));
	191	}
	192
	193	/**
	194	* skb_dst_drop - drops skb dst
	195	* @skb: buffer
	196	*
	197	* Drops dst reference count if a reference was taken.
	198	*/
adf30907 ED	199	static inline void skb_dst_drop(struct sk_buff *skb)
adf30907 ED	200	{
7fee226a ED	201	if (skb->_skb_refdst) {
	202	refdst_drop(skb->_skb_refdst);
	203	skb->_skb_refdst = 0UL;
	204	}
	205	}
	206
	207	static inline void skb_dst_copy(struct sk_buff nskb, const struct sk_buff oskb)
	208	{
	209	nskb->_skb_refdst = oskb->_skb_refdst;
	210	if (!(nskb->_skb_refdst & SKB_DST_NOREF))
	211	dst_clone(skb_dst(nskb));
	212	}
	213
	214	/**
	215	* skb_dst_force - makes sure skb dst is refcounted
	216	* @skb: buffer
	217	*
	218	* If dst is not yet refcounted, let's do it
	219	*/
	220	static inline void skb_dst_force(struct sk_buff *skb)
	221	{
	222	if (skb_dst_is_noref(skb)) {
	223	WARN_ON(!rcu_read_lock_held());
	224	skb->_skb_refdst &= ~SKB_DST_NOREF;
	225	dst_clone(skb_dst(skb));
	226	}
adf30907	227	}
1da177e4	228
d19d56dd ED	229
	230	/**
	231	* skb_tunnel_rx - prepare skb for rx reinsert
	232	* @skb: buffer
	233	* @dev: tunnel device
	234	*
	235	* After decapsulation, packet is going to re-enter (netif_rx()) our stack,
	236	* so make some cleanups, and perform accounting.
	237	*/
	238	static inline void skb_tunnel_rx(struct sk_buff skb, struct net_device dev)
	239	{
	240	skb->dev = dev;
	241	/* TODO : stats should be SMP safe */
	242	dev->stats.rx_packets++;
	243	dev->stats.rx_bytes += skb->len;
	244	skb->rxhash = 0;
	245	skb_dst_drop(skb);
	246	nf_reset(skb);
	247	}
	248
1da177e4 LT	249	/* Children define the path of the packet through the
	250	* Linux networking. Thus, destinations are stackable.
	251	*/
	252
8764ab2c	253	static inline struct dst_entry skb_dst_pop(struct sk_buff skb)
1da177e4	254	{
8764ab2c	255	struct dst_entry *child = skb_dst(skb)->child;
1da177e4	256
8764ab2c	257	skb_dst_drop(skb);
1da177e4 LT	258	return child;
	259	}
	260
352e512c	261	extern int dst_discard(struct sk_buff *skb);
1da177e4 LT	262	extern void * dst_alloc(struct dst_ops * ops);
	263	extern void __dst_free(struct dst_entry * dst);
	264	extern struct dst_entry dst_destroy(struct dst_entry dst);
	265
	266	static inline void dst_free(struct dst_entry * dst)
	267	{
	268	if (dst->obsolete > 1)
	269	return;
	270	if (!atomic_read(&dst->__refcnt)) {
	271	dst = dst_destroy(dst);
	272	if (!dst)
	273	return;
	274	}
	275	__dst_free(dst);
	276	}
	277
	278	static inline void dst_rcu_free(struct rcu_head *head)
	279	{
	280	struct dst_entry *dst = container_of(head, struct dst_entry, rcu_head);
	281	dst_free(dst);
	282	}
	283
	284	static inline void dst_confirm(struct dst_entry *dst)
	285	{
	286	if (dst)
	287	neigh_confirm(dst->neighbour);
	288	}
	289
1da177e4 LT	290	static inline void dst_link_failure(struct sk_buff *skb)
1da177e4 LT	291	{
adf30907	292	struct dst_entry *dst = skb_dst(skb);
1da177e4 LT	293	if (dst && dst->ops && dst->ops->link_failure)
	294	dst->ops->link_failure(skb);
	295	}
	296
	297	static inline void dst_set_expires(struct dst_entry *dst, int timeout)
	298	{
	299	unsigned long expires = jiffies + timeout;
	300
	301	if (expires == 0)
	302	expires = 1;
	303
	304	if (dst->expires == 0 \|\| time_before(expires, dst->expires))
	305	dst->expires = expires;
	306	}
	307
	308	/* Output packet to network from transport. */
	309	static inline int dst_output(struct sk_buff *skb)
	310	{
adf30907	311	return skb_dst(skb)->output(skb);
1da177e4 LT	312	}
	313
	314	/* Input packet from network to transport. */
	315	static inline int dst_input(struct sk_buff *skb)
	316	{
adf30907	317	return skb_dst(skb)->input(skb);
1da177e4 LT	318	}
	319
	320	static inline struct dst_entry dst_check(struct dst_entry dst, u32 cookie)
	321	{
	322	if (dst->obsolete)
	323	dst = dst->ops->check(dst, cookie);
	324	return dst;
	325	}
	326
	327	extern void dst_init(void);
	328
815f4e57 HX	329	/* Flags for xfrm_lookup flags argument. */
	330	enum {
	331	XFRM_LOOKUP_WAIT = 1 << 0,
8b7817f3	332	XFRM_LOOKUP_ICMP = 1 << 1,
815f4e57 HX	333	};
815f4e57 HX	334
1da177e4 LT	335	struct flowi;
1da177e4 LT	336	#ifndef CONFIG_XFRM
52479b62 AD	337	static inline int xfrm_lookup(struct net net, struct dst_entry *dst_p,
52479b62 AD	338	struct flowi fl, struct sock sk, int flags)
1da177e4 LT	339	{
	340	return 0;
	341	}
52479b62 AD	342	static inline int __xfrm_lookup(struct net net, struct dst_entry *dst_p,
52479b62 AD	343	struct flowi fl, struct sock sk, int flags)
14e50e57 DM	344	{
	345	return 0;
	346	}
1da177e4	347	#else
52479b62 AD	348	extern int xfrm_lookup(struct net net, struct dst_entry *dst_p,
	349	struct flowi fl, struct sock sk, int flags);
	350	extern int __xfrm_lookup(struct net net, struct dst_entry *dst_p,
	351	struct flowi fl, struct sock sk, int flags);
1da177e4 LT	352	#endif
	353	#endif
	354
	355	#endif /* _NET_DST_H */