[net-next-2.6.git] / include / linux / skbuff.h

/*
 *	Definitions for the 'struct sk_buff' memory handlers.
 *
 *	Authors:
 *		Alan Cox, <gw4pts@gw4pts.ampr.org>
 *		Florian La Roche, <rzsfl@rz.uni-sb.de>
 *
 *	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 _LINUX_SKBUFF_H
#define _LINUX_SKBUFF_H

#include <linux/kernel.h>
#include <linux/compiler.h>
#include <linux/time.h>
#include <linux/cache.h>

#include <asm/atomic.h>
#include <asm/types.h>
#include <linux/spinlock.h>
#include <linux/net.h>
#include <linux/textsearch.h>
#include <net/checksum.h>
#include <linux/rcupdate.h>
#include <linux/dmaengine.h>
#include <linux/hrtimer.h>

#define HAVE_ALLOC_SKB		/* For the drivers to know */
#define HAVE_ALIGNABLE_SKB	/* Ditto 8)		   */

#define CHECKSUM_NONE 0
#define CHECKSUM_PARTIAL 1
#define CHECKSUM_UNNECESSARY 2
#define CHECKSUM_COMPLETE 3

#define SKB_DATA_ALIGN(X)	(((X) + (SMP_CACHE_BYTES - 1)) & \
				 ~(SMP_CACHE_BYTES - 1))
#define SKB_WITH_OVERHEAD(X)	\
	(((X) - sizeof(struct skb_shared_info)) & \
	 ~(SMP_CACHE_BYTES - 1))
#define SKB_MAX_ORDER(X, ORDER) \
	SKB_WITH_OVERHEAD((PAGE_SIZE << (ORDER)) - (X))
#define SKB_MAX_HEAD(X)		(SKB_MAX_ORDER((X), 0))
#define SKB_MAX_ALLOC		(SKB_MAX_ORDER(0, 2))

/* A. Checksumming of received packets by device.
 *
 *	NONE: device failed to checksum this packet.
 *		skb->csum is undefined.
 *
 *	UNNECESSARY: device parsed packet and wouldbe verified checksum.
 *		skb->csum is undefined.
 *	      It is bad option, but, unfortunately, many of vendors do this.
 *	      Apparently with secret goal to sell you new device, when you
 *	      will add new protocol to your host. F.e. IPv6. 8)
 *
 *	COMPLETE: the most generic way. Device supplied checksum of _all_
 *	    the packet as seen by netif_rx in skb->csum.
 *	    NOTE: Even if device supports only some protocols, but
 *	    is able to produce some skb->csum, it MUST use COMPLETE,
 *	    not UNNECESSARY.
 *
 * B. Checksumming on output.
 *
 *	NONE: skb is checksummed by protocol or csum is not required.
 *
 *	PARTIAL: device is required to csum packet as seen by hard_start_xmit
 *	from skb->h.raw to the end and to record the checksum
 *	at skb->h.raw+skb->csum.
 *
 *	Device must show its capabilities in dev->features, set
 *	at device setup time.
 *	NETIF_F_HW_CSUM	- it is clever device, it is able to checksum
 *			  everything.
 *	NETIF_F_NO_CSUM - loopback or reliable single hop media.
 *	NETIF_F_IP_CSUM - device is dumb. It is able to csum only
 *			  TCP/UDP over IPv4. Sigh. Vendors like this
 *			  way by an unknown reason. Though, see comment above
 *			  about CHECKSUM_UNNECESSARY. 8)
 *
 *	Any questions? No questions, good. 		--ANK
 */

struct net_device;

#ifdef CONFIG_NETFILTER
struct nf_conntrack {
	atomic_t use;
	void (*destroy)(struct nf_conntrack *);
};

#ifdef CONFIG_BRIDGE_NETFILTER
struct nf_bridge_info {
	atomic_t use;
	struct net_device *physindev;
	struct net_device *physoutdev;
#if defined(CONFIG_VLAN_8021Q) || defined(CONFIG_VLAN_8021Q_MODULE)
	struct net_device *netoutdev;
#endif
	unsigned int mask;
	unsigned long data[32 / sizeof(unsigned long)];
};
#endif

#endif

struct sk_buff_head {
	/* These two members must be first. */
	struct sk_buff	*next;
	struct sk_buff	*prev;

	__u32		qlen;
	spinlock_t	lock;
};

struct sk_buff;

/* To allow 64K frame to be packed as single skb without frag_list */
#define MAX_SKB_FRAGS (65536/PAGE_SIZE + 2)

typedef struct skb_frag_struct skb_frag_t;

struct skb_frag_struct {
	struct page *page;
	__u16 page_offset;
	__u16 size;
};

/* This data is invariant across clones and lives at
 * the end of the header data, ie. at skb->end.
 */
struct skb_shared_info {
	atomic_t	dataref;
	unsigned short	nr_frags;
	unsigned short	gso_size;
	/* Warning: this field is not always filled in (UFO)! */
	unsigned short	gso_segs;
	unsigned short  gso_type;
	__be32          ip6_frag_id;
	struct sk_buff	*frag_list;
	skb_frag_t	frags[MAX_SKB_FRAGS];
};

/* We divide dataref into two halves.  The higher 16 bits hold references
 * to the payload part of skb->data.  The lower 16 bits hold references to
 * the entire skb->data.  It is up to the users of the skb to agree on
 * where the payload starts.
 *
 * All users must obey the rule that the skb->data reference count must be
 * greater than or equal to the payload reference count.
 *
 * Holding a reference to the payload part means that the user does not
 * care about modifications to the header part of skb->data.
 */
#define SKB_DATAREF_SHIFT 16
#define SKB_DATAREF_MASK ((1 << SKB_DATAREF_SHIFT) - 1)


enum {
	SKB_FCLONE_UNAVAILABLE,
	SKB_FCLONE_ORIG,
	SKB_FCLONE_CLONE,
};

enum {
	SKB_GSO_TCPV4 = 1 << 0,
	SKB_GSO_UDP = 1 << 1,

	/* This indicates the skb is from an untrusted source. */
	SKB_GSO_DODGY = 1 << 2,

	/* This indicates the tcp segment has CWR set. */
	SKB_GSO_TCP_ECN = 1 << 3,

	SKB_GSO_TCPV6 = 1 << 4,
};

/** 
 *	struct sk_buff - socket buffer
 *	@next: Next buffer in list
 *	@prev: Previous buffer in list
 *	@sk: Socket we are owned by
 *	@tstamp: Time we arrived
 *	@dev: Device we arrived on/are leaving by
 *	@iif: ifindex of device we arrived on
 *	@h: Transport layer header
 *	@nh: Network layer header
 *	@mac: Link layer header
 *	@dst: destination entry
 *	@sp: the security path, used for xfrm
 *	@cb: Control buffer. Free for use by every layer. Put private vars here
 *	@len: Length of actual data
 *	@data_len: Data length
 *	@mac_len: Length of link layer header
 *	@csum: Checksum
 *	@local_df: allow local fragmentation
 *	@cloned: Head may be cloned (check refcnt to be sure)
 *	@nohdr: Payload reference only, must not modify header
 *	@pkt_type: Packet class
 *	@fclone: skbuff clone status
 *	@ip_summed: Driver fed us an IP checksum
 *	@priority: Packet queueing priority
 *	@users: User count - see {datagram,tcp}.c
 *	@protocol: Packet protocol from driver
 *	@truesize: Buffer size 
 *	@head: Head of buffer
 *	@data: Data head pointer
 *	@tail: Tail pointer
 *	@end: End pointer
 *	@destructor: Destruct function
 *	@mark: Generic packet mark
 *	@nfct: Associated connection, if any
 *	@ipvs_property: skbuff is owned by ipvs
 *	@nfctinfo: Relationship of this skb to the connection
 *	@nfct_reasm: netfilter conntrack re-assembly pointer
 *	@nf_bridge: Saved data about a bridged frame - see br_netfilter.c
 *	@tc_index: Traffic control index
 *	@tc_verd: traffic control verdict
 *	@dma_cookie: a cookie to one of several possible DMA operations
 *		done by skb DMA functions
 *	@secmark: security marking
 */

struct sk_buff {
	/* These two members must be first. */
	struct sk_buff		*next;
	struct sk_buff		*prev;

	struct sock		*sk;
	ktime_t			tstamp;
	struct net_device	*dev;
	int			iif;
	/* 4 byte hole on 64 bit*/

	union {
		unsigned char	*raw;
	} h;

	union {
		unsigned char	*raw;
	} nh;

	union {
	  	unsigned char 	*raw;
	} mac;

	struct  dst_entry	*dst;
	struct	sec_path	*sp;

	/*
	 * This is the control buffer. It is free to use for every
	 * layer. Please put your private variables there. If you
	 * want to keep them across layers you have to do a skb_clone()
	 * first. This is owned by whoever has the skb queued ATM.
	 */
	char			cb[48];

	unsigned int		len,
				data_len,
				mac_len;
	union {
		__wsum		csum;
		__u32		csum_offset;
	};
	__u32			priority;
	__u8			local_df:1,
				cloned:1,
				ip_summed:2,
				nohdr:1,
				nfctinfo:3;
	__u8			pkt_type:3,
				fclone:2,
				ipvs_property:1;
	__be16			protocol;

	void			(*destructor)(struct sk_buff *skb);
#ifdef CONFIG_NETFILTER
	struct nf_conntrack	*nfct;
#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
	struct sk_buff		*nfct_reasm;
#endif
#ifdef CONFIG_BRIDGE_NETFILTER
	struct nf_bridge_info	*nf_bridge;
#endif
#endif /* CONFIG_NETFILTER */
#ifdef CONFIG_NET_SCHED
	__u16			tc_index;	/* traffic control index */
#ifdef CONFIG_NET_CLS_ACT
	__u16			tc_verd;	/* traffic control verdict */
#endif
#endif
#ifdef CONFIG_NET_DMA
	dma_cookie_t		dma_cookie;
#endif
#ifdef CONFIG_NETWORK_SECMARK
	__u32			secmark;
#endif

	__u32			mark;

	/* These elements must be at the end, see alloc_skb() for details.  */
	unsigned int		truesize;
	atomic_t		users;
	unsigned char		*head,
				*data,
				*tail,
				*end;
};

#ifdef __KERNEL__
/*
 *	Handling routines are only of interest to the kernel
 */
#include <linux/slab.h>

#include <asm/system.h>
Commit	Line	Data
1da177e4 LT	1	/*
	2	* Definitions for the 'struct sk_buff' memory handlers.
	3	*
	4	* Authors:
	5	* Alan Cox, <gw4pts@gw4pts.ampr.org>
	6	* Florian La Roche, <rzsfl@rz.uni-sb.de>
	7	*
	8	* This program is free software; you can redistribute it and/or
	9	* modify it under the terms of the GNU General Public License
	10	* as published by the Free Software Foundation; either version
	11	* 2 of the License, or (at your option) any later version.
	12	*/
	13
	14	#ifndef _LINUX_SKBUFF_H
	15	#define _LINUX_SKBUFF_H
	16
1da177e4 LT	17	#include <linux/kernel.h>
	18	#include <linux/compiler.h>
	19	#include <linux/time.h>
	20	#include <linux/cache.h>
	21
	22	#include <asm/atomic.h>
	23	#include <asm/types.h>
	24	#include <linux/spinlock.h>
1da177e4	25	#include <linux/net.h>
3fc7e8a6	26	#include <linux/textsearch.h>
1da177e4	27	#include <net/checksum.h>
a80958f4	28	#include <linux/rcupdate.h>
97fc2f08	29	#include <linux/dmaengine.h>
b7aa0bf7	30	#include <linux/hrtimer.h>
1da177e4 LT	31
	32	#define HAVE_ALLOC_SKB /* For the drivers to know */
	33	#define HAVE_ALIGNABLE_SKB /* Ditto 8) */
1da177e4 LT	34
1da177e4 LT	35	#define CHECKSUM_NONE 0
84fa7933	36	#define CHECKSUM_PARTIAL 1
1da177e4	37	#define CHECKSUM_UNNECESSARY 2
84fa7933	38	#define CHECKSUM_COMPLETE 3
1da177e4 LT	39
	40	#define SKB_DATA_ALIGN(X) (((X) + (SMP_CACHE_BYTES - 1)) & \
	41	~(SMP_CACHE_BYTES - 1))
fc910a27 DM	42	#define SKB_WITH_OVERHEAD(X) \
	43	(((X) - sizeof(struct skb_shared_info)) & \
	44	~(SMP_CACHE_BYTES - 1))
	45	#define SKB_MAX_ORDER(X, ORDER) \
	46	SKB_WITH_OVERHEAD((PAGE_SIZE << (ORDER)) - (X))
1da177e4 LT	47	#define SKB_MAX_HEAD(X) (SKB_MAX_ORDER((X), 0))
	48	#define SKB_MAX_ALLOC (SKB_MAX_ORDER(0, 2))
	49
	50	/* A. Checksumming of received packets by device.
	51	*
	52	* NONE: device failed to checksum this packet.
	53	* skb->csum is undefined.
	54	*
	55	* UNNECESSARY: device parsed packet and wouldbe verified checksum.
	56	* skb->csum is undefined.
	57	* It is bad option, but, unfortunately, many of vendors do this.
	58	* Apparently with secret goal to sell you new device, when you
	59	* will add new protocol to your host. F.e. IPv6. 8)
	60	*
84fa7933	61	* COMPLETE: the most generic way. Device supplied checksum of _all_
1da177e4 LT	62	* the packet as seen by netif_rx in skb->csum.
1da177e4 LT	63	* NOTE: Even if device supports only some protocols, but
84fa7933	64	* is able to produce some skb->csum, it MUST use COMPLETE,
1da177e4 LT	65	* not UNNECESSARY.
	66	*
	67	* B. Checksumming on output.
	68	*
	69	* NONE: skb is checksummed by protocol or csum is not required.
	70	*
84fa7933	71	* PARTIAL: device is required to csum packet as seen by hard_start_xmit
1da177e4 LT	72	* from skb->h.raw to the end and to record the checksum
	73	* at skb->h.raw+skb->csum.
	74	*
	75	* Device must show its capabilities in dev->features, set
	76	* at device setup time.
	77	* NETIF_F_HW_CSUM - it is clever device, it is able to checksum
	78	* everything.
	79	* NETIF_F_NO_CSUM - loopback or reliable single hop media.
	80	* NETIF_F_IP_CSUM - device is dumb. It is able to csum only
	81	* TCP/UDP over IPv4. Sigh. Vendors like this
	82	* way by an unknown reason. Though, see comment above
	83	* about CHECKSUM_UNNECESSARY. 8)
	84	*
	85	* Any questions? No questions, good. --ANK
	86	*/
	87
1da177e4 LT	88	struct net_device;
	89
	90	#ifdef CONFIG_NETFILTER
	91	struct nf_conntrack {
	92	atomic_t use;
	93	void (destroy)(struct nf_conntrack );
	94	};
	95
	96	#ifdef CONFIG_BRIDGE_NETFILTER
	97	struct nf_bridge_info {
	98	atomic_t use;
	99	struct net_device *physindev;
	100	struct net_device *physoutdev;
	101	#if defined(CONFIG_VLAN_8021Q) \|\| defined(CONFIG_VLAN_8021Q_MODULE)
	102	struct net_device *netoutdev;
	103	#endif
	104	unsigned int mask;
	105	unsigned long data[32 / sizeof(unsigned long)];
	106	};
	107	#endif
	108
	109	#endif
	110
	111	struct sk_buff_head {
	112	/* These two members must be first. */
	113	struct sk_buff *next;
	114	struct sk_buff *prev;
	115
	116	__u32 qlen;
	117	spinlock_t lock;
	118	};
	119
	120	struct sk_buff;
	121
	122	/* To allow 64K frame to be packed as single skb without frag_list */
	123	#define MAX_SKB_FRAGS (65536/PAGE_SIZE + 2)
	124
	125	typedef struct skb_frag_struct skb_frag_t;
	126
	127	struct skb_frag_struct {
	128	struct page *page;
	129	__u16 page_offset;
	130	__u16 size;
	131	};
	132
	133	/* This data is invariant across clones and lives at
	134	* the end of the header data, ie. at skb->end.
	135	*/
	136	struct skb_shared_info {
	137	atomic_t dataref;
4947d3ef	138	unsigned short nr_frags;
7967168c HX	139	unsigned short gso_size;
	140	/* Warning: this field is not always filled in (UFO)! */
	141	unsigned short gso_segs;
	142	unsigned short gso_type;
ae08e1f0	143	__be32 ip6_frag_id;
1da177e4 LT	144	struct sk_buff *frag_list;
	145	skb_frag_t frags[MAX_SKB_FRAGS];
	146	};
	147
	148	/* We divide dataref into two halves. The higher 16 bits hold references
	149	* to the payload part of skb->data. The lower 16 bits hold references to
	150	* the entire skb->data. It is up to the users of the skb to agree on
	151	* where the payload starts.
	152	*
	153	* All users must obey the rule that the skb->data reference count must be
	154	* greater than or equal to the payload reference count.
	155	*
	156	* Holding a reference to the payload part means that the user does not
	157	* care about modifications to the header part of skb->data.
	158	*/
	159	#define SKB_DATAREF_SHIFT 16
	160	#define SKB_DATAREF_MASK ((1 << SKB_DATAREF_SHIFT) - 1)
	161
d179cd12 DM	162
	163	enum {
	164	SKB_FCLONE_UNAVAILABLE,
	165	SKB_FCLONE_ORIG,
	166	SKB_FCLONE_CLONE,
	167	};
	168
7967168c HX	169	enum {
7967168c HX	170	SKB_GSO_TCPV4 = 1 << 0,
f83ef8c0	171	SKB_GSO_UDP = 1 << 1,
576a30eb HX	172
	173	/* This indicates the skb is from an untrusted source. */
	174	SKB_GSO_DODGY = 1 << 2,
b0da8537 MC	175
b0da8537 MC	176	/* This indicates the tcp segment has CWR set. */
f83ef8c0 HX	177	SKB_GSO_TCP_ECN = 1 << 3,
	178
	179	SKB_GSO_TCPV6 = 1 << 4,
7967168c HX	180	};
7967168c HX	181
1da177e4 LT	182	/**
	183	* struct sk_buff - socket buffer
	184	* @next: Next buffer in list
	185	* @prev: Previous buffer in list
1da177e4	186	* @sk: Socket we are owned by
325ed823	187	* @tstamp: Time we arrived
1da177e4	188	* @dev: Device we arrived on/are leaving by
c01003c2	189	* @iif: ifindex of device we arrived on
1da177e4 LT	190	* @h: Transport layer header
	191	* @nh: Network layer header
	192	* @mac: Link layer header
67be2dd1 MW	193	* @dst: destination entry
67be2dd1 MW	194	* @sp: the security path, used for xfrm
1da177e4 LT	195	* @cb: Control buffer. Free for use by every layer. Put private vars here
	196	* @len: Length of actual data
	197	* @data_len: Data length
	198	* @mac_len: Length of link layer header
	199	* @csum: Checksum
67be2dd1	200	* @local_df: allow local fragmentation
1da177e4 LT	201	* @cloned: Head may be cloned (check refcnt to be sure)
	202	* @nohdr: Payload reference only, must not modify header
	203	* @pkt_type: Packet class
c83c2486	204	* @fclone: skbuff clone status
1da177e4 LT	205	* @ip_summed: Driver fed us an IP checksum
	206	* @priority: Packet queueing priority
	207	* @users: User count - see {datagram,tcp}.c
	208	* @protocol: Packet protocol from driver
1da177e4 LT	209	* @truesize: Buffer size
	210	* @head: Head of buffer
	211	* @data: Data head pointer
	212	* @tail: Tail pointer
	213	* @end: End pointer
	214	* @destructor: Destruct function
82e91ffe	215	* @mark: Generic packet mark
1da177e4	216	* @nfct: Associated connection, if any
c83c2486	217	* @ipvs_property: skbuff is owned by ipvs
1da177e4	218	* @nfctinfo: Relationship of this skb to the connection
461ddf3b	219	* @nfct_reasm: netfilter conntrack re-assembly pointer
1da177e4	220	* @nf_bridge: Saved data about a bridged frame - see br_netfilter.c
1da177e4 LT	221	* @tc_index: Traffic control index
1da177e4 LT	222	* @tc_verd: traffic control verdict
f4b8ea78 RD	223	* @dma_cookie: a cookie to one of several possible DMA operations
f4b8ea78 RD	224	* done by skb DMA functions
984bc16c	225	* @secmark: security marking
1da177e4 LT	226	*/
	227
	228	struct sk_buff {
	229	/* These two members must be first. */
	230	struct sk_buff *next;
	231	struct sk_buff *prev;
	232
1da177e4	233	struct sock *sk;
b7aa0bf7	234	ktime_t tstamp;
1da177e4	235	struct net_device *dev;
c01003c2 PM	236	int iif;
c01003c2 PM	237	/* 4 byte hole on 64 bit*/
1da177e4 LT	238
1da177e4 LT	239	union {
1da177e4 LT	240	unsigned char *raw;
	241	} h;
	242
	243	union {
1da177e4 LT	244	unsigned char *raw;
	245	} nh;
	246
	247	union {
	248	unsigned char *raw;
	249	} mac;
	250
	251	struct dst_entry *dst;
	252	struct sec_path *sp;
	253
	254	/*
	255	* This is the control buffer. It is free to use for every
	256	* layer. Please put your private variables there. If you
	257	* want to keep them across layers you have to do a skb_clone()
	258	* first. This is owned by whoever has the skb queued ATM.
	259	*/
3e3850e9	260	char cb[48];
1da177e4 LT	261
	262	unsigned int len,
	263	data_len,
1f61ab5c	264	mac_len;
ff1dcadb AV	265	union {
	266	__wsum csum;
	267	__u32 csum_offset;
	268	};
1da177e4	269	__u32 priority;
1cbb3380 TG	270	__u8 local_df:1,
	271	cloned:1,
	272	ip_summed:2,
6869c4d8 HW	273	nohdr:1,
6869c4d8 HW	274	nfctinfo:3;
d179cd12	275	__u8 pkt_type:3,
b84f4cc9 PM	276	fclone:2,
b84f4cc9 PM	277	ipvs_property:1;
a0d3bea3	278	__be16 protocol;
1da177e4 LT	279
	280	void (destructor)(struct sk_buff skb);
	281	#ifdef CONFIG_NETFILTER
1da177e4	282	struct nf_conntrack *nfct;
9fb9cbb1 YK	283	#if defined(CONFIG_NF_CONNTRACK) \|\| defined(CONFIG_NF_CONNTRACK_MODULE)
	284	struct sk_buff *nfct_reasm;
	285	#endif
1da177e4 LT	286	#ifdef CONFIG_BRIDGE_NETFILTER
	287	struct nf_bridge_info *nf_bridge;
	288	#endif
	289	#endif /* CONFIG_NETFILTER */
1da177e4	290	#ifdef CONFIG_NET_SCHED
b6b99eb5	291	__u16 tc_index; /* traffic control index */
1da177e4	292	#ifdef CONFIG_NET_CLS_ACT
b6b99eb5	293	__u16 tc_verd; /* traffic control verdict */
1da177e4	294	#endif
1da177e4	295	#endif
97fc2f08 CL	296	#ifdef CONFIG_NET_DMA
	297	dma_cookie_t dma_cookie;
	298	#endif
984bc16c JM	299	#ifdef CONFIG_NETWORK_SECMARK
	300	__u32 secmark;
	301	#endif
1da177e4	302
82e91ffe	303	__u32 mark;
1da177e4 LT	304
	305	/* These elements must be at the end, see alloc_skb() for details. */
	306	unsigned int truesize;
	307	atomic_t users;
	308	unsigned char *head,
	309	*data,
	310	*tail,
	311	*end;
	312	};
	313
	314	#ifdef __KERNEL__
	315	/*
	316	* Handling routines are only of interest to the kernel
	317	*/
	318	#include <linux/slab.h>
	319
	320	#include <asm/system.h>
	321