[deliverable/linux.git] / include / net / inet_ecn.h

#ifndef _INET_ECN_H_
#define _INET_ECN_H_

#include <linux/ip.h>
#include <linux/skbuff.h>

#include <net/inet_sock.h>
#include <net/dsfield.h>

enum {
	INET_ECN_NOT_ECT = 0,
	INET_ECN_ECT_1 = 1,
	INET_ECN_ECT_0 = 2,
	INET_ECN_CE = 3,
	INET_ECN_MASK = 3,
};

extern int sysctl_tunnel_ecn_log;

static inline int INET_ECN_is_ce(__u8 dsfield)
{
	return (dsfield & INET_ECN_MASK) == INET_ECN_CE;
}

static inline int INET_ECN_is_not_ect(__u8 dsfield)
{
	return (dsfield & INET_ECN_MASK) == INET_ECN_NOT_ECT;
}

static inline int INET_ECN_is_capable(__u8 dsfield)
{
	return dsfield & INET_ECN_ECT_0;
}

/*
 * RFC 3168 9.1.1
 *  The full-functionality option for ECN encapsulation is to copy the
 *  ECN codepoint of the inside header to the outside header on
 *  encapsulation if the inside header is not-ECT or ECT, and to set the
 *  ECN codepoint of the outside header to ECT(0) if the ECN codepoint of
 *  the inside header is CE.
 */
static inline __u8 INET_ECN_encapsulate(__u8 outer, __u8 inner)
{
	outer &= ~INET_ECN_MASK;
	outer |= !INET_ECN_is_ce(inner) ? (inner & INET_ECN_MASK) :
					  INET_ECN_ECT_0;
	return outer;
}

static inline void INET_ECN_xmit(struct sock *sk)
{
	inet_sk(sk)->tos |= INET_ECN_ECT_0;
	if (inet6_sk(sk) != NULL)
		inet6_sk(sk)->tclass |= INET_ECN_ECT_0;
}

static inline void INET_ECN_dontxmit(struct sock *sk)
{
	inet_sk(sk)->tos &= ~INET_ECN_MASK;
	if (inet6_sk(sk) != NULL)
		inet6_sk(sk)->tclass &= ~INET_ECN_MASK;
}

#define IP6_ECN_flow_init(label) do {		\
      (label) &= ~htonl(INET_ECN_MASK << 20);	\
    } while (0)

#define	IP6_ECN_flow_xmit(sk, label) do {				\
	if (INET_ECN_is_capable(inet6_sk(sk)->tclass))			\
		(label) |= htonl(INET_ECN_ECT_0 << 20);			\
    } while (0)

static inline int IP_ECN_set_ce(struct iphdr *iph)
{
	u32 check = (__force u32)iph->check;
	u32 ecn = (iph->tos + 1) & INET_ECN_MASK;

	/*
	 * After the last operation we have (in binary):
	 * INET_ECN_NOT_ECT => 01
	 * INET_ECN_ECT_1   => 10
	 * INET_ECN_ECT_0   => 11
	 * INET_ECN_CE      => 00
	 */
	if (!(ecn & 2))
		return !ecn;

	/*
	 * The following gives us:
	 * INET_ECN_ECT_1 => check += htons(0xFFFD)
	 * INET_ECN_ECT_0 => check += htons(0xFFFE)
	 */
	check += (__force u16)htons(0xFFFB) + (__force u16)htons(ecn);

	iph->check = (__force __sum16)(check + (check>=0xFFFF));
	iph->tos |= INET_ECN_CE;
	return 1;
}

static inline void IP_ECN_clear(struct iphdr *iph)
{
	iph->tos &= ~INET_ECN_MASK;
}

static inline void ipv4_copy_dscp(unsigned int dscp, struct iphdr *inner)
{
	dscp &= ~INET_ECN_MASK;
	ipv4_change_dsfield(inner, INET_ECN_MASK, dscp);
}

struct ipv6hdr;

/* Note:
 * IP_ECN_set_ce() has to tweak IPV4 checksum when setting CE,
 * meaning both changes have no effect on skb->csum if/when CHECKSUM_COMPLETE
 * In IPv6 case, no checksum compensates the change in IPv6 header,
 * so we have to update skb->csum.
 */
static inline int IP6_ECN_set_ce(struct sk_buff *skb, struct ipv6hdr *iph)
{
	__be32 from, to;

	if (INET_ECN_is_not_ect(ipv6_get_dsfield(iph)))
		return 0;

	from = *(__be32 *)iph;
	to = from | htonl(INET_ECN_CE << 20);
	*(__be32 *)iph = to;
	if (skb->ip_summed == CHECKSUM_COMPLETE)
		skb->csum = csum_add(csum_sub(skb->csum, from), to);
	return 1;
}

static inline void IP6_ECN_clear(struct ipv6hdr *iph)
{
	*(__be32*)iph &= ~htonl(INET_ECN_MASK << 20);
}

static inline void ipv6_copy_dscp(unsigned int dscp, struct ipv6hdr *inner)
{
	dscp &= ~INET_ECN_MASK;
	ipv6_change_dsfield(inner, INET_ECN_MASK, dscp);
}

static inline int INET_ECN_set_ce(struct sk_buff *skb)
{
	switch (skb->protocol) {
	case cpu_to_be16(ETH_P_IP):
		if (skb_network_header(skb) + sizeof(struct iphdr) <=
		    skb_tail_pointer(skb))
			return IP_ECN_set_ce(ip_hdr(skb));
		break;

	case cpu_to_be16(ETH_P_IPV6):
		if (skb_network_header(skb) + sizeof(struct ipv6hdr) <=
		    skb_tail_pointer(skb))
			return IP6_ECN_set_ce(skb, ipv6_hdr(skb));
		break;
	}

	return 0;
}

/*
 * RFC 6040 4.2
 *  To decapsulate the inner header at the tunnel egress, a compliant
 *  tunnel egress MUST set the outgoing ECN field to the codepoint at the
 *  intersection of the appropriate arriving inner header (row) and outer
 *  header (column) in Figure 4
 *
 *      +---------+------------------------------------------------+
 *      |Arriving |            Arriving Outer Header               |
 *      |   Inner +---------+------------+------------+------------+
 *      |  Header | Not-ECT | ECT(0)     | ECT(1)     |     CE     |
 *      +---------+---------+------------+------------+------------+
 *      | Not-ECT | Not-ECT |Not-ECT(!!!)|Not-ECT(!!!)| <drop>(!!!)|
 *      |  ECT(0) |  ECT(0) | ECT(0)     | ECT(1)     |     CE     |
 *      |  ECT(1) |  ECT(1) | ECT(1) (!) | ECT(1)     |     CE     |
 *      |    CE   |      CE |     CE     |     CE(!!!)|     CE     |
 *      +---------+---------+------------+------------+------------+
 *
 *             Figure 4: New IP in IP Decapsulation Behaviour
 *
 *  returns 0 on success
 *          1 if something is broken and should be logged (!!! above)
 *          2 if packet should be dropped
 */
static inline int INET_ECN_decapsulate(struct sk_buff *skb,
				       __u8 outer, __u8 inner)
{
	if (INET_ECN_is_not_ect(inner)) {
		switch (outer & INET_ECN_MASK) {
		case INET_ECN_NOT_ECT:
			return 0;
		case INET_ECN_ECT_0:
		case INET_ECN_ECT_1:
			return 1;
		case INET_ECN_CE:
			return 2;
		}
	}

	if (INET_ECN_is_ce(outer))
		INET_ECN_set_ce(skb);

	return 0;
}

static inline int IP_ECN_decapsulate(const struct iphdr *oiph,
				     struct sk_buff *skb)
{
	__u8 inner;

	if (skb->protocol == htons(ETH_P_IP))
		inner = ip_hdr(skb)->tos;
	else if (skb->protocol == htons(ETH_P_IPV6))
		inner = ipv6_get_dsfield(ipv6_hdr(skb));
	else
		return 0;

	return INET_ECN_decapsulate(skb, oiph->tos, inner);
}

static inline int IP6_ECN_decapsulate(const struct ipv6hdr *oipv6h,
				      struct sk_buff *skb)
{
	__u8 inner;

	if (skb->protocol == htons(ETH_P_IP))
		inner = ip_hdr(skb)->tos;
	else if (skb->protocol == htons(ETH_P_IPV6))
		inner = ipv6_get_dsfield(ipv6_hdr(skb));
	else
		return 0;

	return INET_ECN_decapsulate(skb, ipv6_get_dsfield(oipv6h), inner);
}
#endif
Commit	Line	Data
1da177e4 LT	1	#ifndef _INET_ECN_H_
	2	#define _INET_ECN_H_
	3
	4	#include <linux/ip.h>
2566a509	5	#include <linux/skbuff.h>
14c85021 ACM	6
14c85021 ACM	7	#include <net/inet_sock.h>
1da177e4 LT	8	#include <net/dsfield.h>
	9
	10	enum {
	11	INET_ECN_NOT_ECT = 0,
	12	INET_ECN_ECT_1 = 1,
	13	INET_ECN_ECT_0 = 2,
	14	INET_ECN_CE = 3,
	15	INET_ECN_MASK = 3,
	16	};
	17
eccc1bb8	18	extern int sysctl_tunnel_ecn_log;
eccc1bb8	19
1da177e4 LT	20	static inline int INET_ECN_is_ce(__u8 dsfield)
	21	{
	22	return (dsfield & INET_ECN_MASK) == INET_ECN_CE;
	23	}
	24
	25	static inline int INET_ECN_is_not_ect(__u8 dsfield)
	26	{
	27	return (dsfield & INET_ECN_MASK) == INET_ECN_NOT_ECT;
	28	}
	29
	30	static inline int INET_ECN_is_capable(__u8 dsfield)
	31	{
a02cec21	32	return dsfield & INET_ECN_ECT_0;
1da177e4 LT	33	}
1da177e4 LT	34
b5d9c9c2 ED	35	/*
	36	* RFC 3168 9.1.1
	37	* The full-functionality option for ECN encapsulation is to copy the
	38	* ECN codepoint of the inside header to the outside header on
	39	* encapsulation if the inside header is not-ECT or ECT, and to set the
	40	* ECN codepoint of the outside header to ECT(0) if the ECN codepoint of
	41	* the inside header is CE.
	42	*/
1da177e4 LT	43	static inline __u8 INET_ECN_encapsulate(__u8 outer, __u8 inner)
	44	{
	45	outer &= ~INET_ECN_MASK;
	46	outer \|= !INET_ECN_is_ce(inner) ? (inner & INET_ECN_MASK) :
	47	INET_ECN_ECT_0;
	48	return outer;
	49	}
	50
ca067070 SG	51	static inline void INET_ECN_xmit(struct sock *sk)
	52	{
	53	inet_sk(sk)->tos \|= INET_ECN_ECT_0;
	54	if (inet6_sk(sk) != NULL)
	55	inet6_sk(sk)->tclass \|= INET_ECN_ECT_0;
	56	}
	57
	58	static inline void INET_ECN_dontxmit(struct sock *sk)
	59	{
	60	inet_sk(sk)->tos &= ~INET_ECN_MASK;
	61	if (inet6_sk(sk) != NULL)
	62	inet6_sk(sk)->tclass &= ~INET_ECN_MASK;
	63	}
1da177e4 LT	64
	65	#define IP6_ECN_flow_init(label) do { \
	66	(label) &= ~htonl(INET_ECN_MASK << 20); \
	67	} while (0)
	68
	69	#define IP6_ECN_flow_xmit(sk, label) do { \
e9df2e8f	70	if (INET_ECN_is_capable(inet6_sk(sk)->tclass)) \
95026cd2	71	(label) \|= htonl(INET_ECN_ECT_0 << 20); \
1da177e4 LT	72	} while (0)
1da177e4 LT	73
2566a509	74	static inline int IP_ECN_set_ce(struct iphdr *iph)
1da177e4	75	{
5c78f275	76	u32 check = (__force u32)iph->check;
1da177e4 LT	77	u32 ecn = (iph->tos + 1) & INET_ECN_MASK;
	78
	79	/*
	80	* After the last operation we have (in binary):
	81	* INET_ECN_NOT_ECT => 01
	82	* INET_ECN_ECT_1 => 10
	83	* INET_ECN_ECT_0 => 11
	84	* INET_ECN_CE => 00
	85	*/
	86	if (!(ecn & 2))
2566a509	87	return !ecn;
1da177e4 LT	88
	89	/*
	90	* The following gives us:
	91	* INET_ECN_ECT_1 => check += htons(0xFFFD)
	92	* INET_ECN_ECT_0 => check += htons(0xFFFE)
	93	*/
5c78f275	94	check += (__force u16)htons(0xFFFB) + (__force u16)htons(ecn);
1da177e4	95
5c78f275	96	iph->check = (__force __sum16)(check + (check>=0xFFFF));
1da177e4	97	iph->tos \|= INET_ECN_CE;
2566a509	98	return 1;
1da177e4 LT	99	}
	100
	101	static inline void IP_ECN_clear(struct iphdr *iph)
	102	{
	103	iph->tos &= ~INET_ECN_MASK;
	104	}
	105
29bb43b4	106	static inline void ipv4_copy_dscp(unsigned int dscp, struct iphdr *inner)
1da177e4	107	{
29bb43b4	108	dscp &= ~INET_ECN_MASK;
1da177e4 LT	109	ipv4_change_dsfield(inner, INET_ECN_MASK, dscp);
	110	}
	111
	112	struct ipv6hdr;
	113
34ae6a1a ED	114	/* Note:
	115	* IP_ECN_set_ce() has to tweak IPV4 checksum when setting CE,
	116	* meaning both changes have no effect on skb->csum if/when CHECKSUM_COMPLETE
	117	* In IPv6 case, no checksum compensates the change in IPv6 header,
	118	* so we have to update skb->csum.
	119	*/
	120	static inline int IP6_ECN_set_ce(struct sk_buff skb, struct ipv6hdr iph)
1da177e4	121	{
34ae6a1a ED	122	__be32 from, to;
34ae6a1a ED	123
1da177e4	124	if (INET_ECN_is_not_ect(ipv6_get_dsfield(iph)))
2566a509	125	return 0;
34ae6a1a ED	126
	127	from = (__be32 )iph;
	128	to = from \| htonl(INET_ECN_CE << 20);
	129	(__be32 )iph = to;
	130	if (skb->ip_summed == CHECKSUM_COMPLETE)
	131	skb->csum = csum_add(csum_sub(skb->csum, from), to);
2566a509	132	return 1;
1da177e4 LT	133	}
	134
	135	static inline void IP6_ECN_clear(struct ipv6hdr *iph)
	136	{
92d9ece7	137	(__be32)iph &= ~htonl(INET_ECN_MASK << 20);
1da177e4 LT	138	}
1da177e4 LT	139
29bb43b4	140	static inline void ipv6_copy_dscp(unsigned int dscp, struct ipv6hdr *inner)
1da177e4	141	{
29bb43b4	142	dscp &= ~INET_ECN_MASK;
1da177e4 LT	143	ipv6_change_dsfield(inner, INET_ECN_MASK, dscp);
	144	}
	145
2566a509 TG	146	static inline int INET_ECN_set_ce(struct sk_buff *skb)
	147	{
	148	switch (skb->protocol) {
f3a7c66b	149	case cpu_to_be16(ETH_P_IP):
ced14f68 SH	150	if (skb_network_header(skb) + sizeof(struct iphdr) <=
ced14f68 SH	151	skb_tail_pointer(skb))
eddc9ec5	152	return IP_ECN_set_ce(ip_hdr(skb));
2566a509 TG	153	break;
2566a509 TG	154
f3a7c66b	155	case cpu_to_be16(ETH_P_IPV6):
ced14f68 SH	156	if (skb_network_header(skb) + sizeof(struct ipv6hdr) <=
ced14f68 SH	157	skb_tail_pointer(skb))
34ae6a1a	158	return IP6_ECN_set_ce(skb, ipv6_hdr(skb));
2566a509 TG	159	break;
	160	}
	161
	162	return 0;
	163	}
	164
eccc1bb8	165	/*
d28071d1	166	* RFC 6040 4.2
eccc1bb8	167	* To decapsulate the inner header at the tunnel egress, a compliant
	168	* tunnel egress MUST set the outgoing ECN field to the codepoint at the
	169	* intersection of the appropriate arriving inner header (row) and outer
	170	* header (column) in Figure 4
	171	*
	172	* +---------+------------------------------------------------+
	173	* \|Arriving \| Arriving Outer Header \|
	174	* \| Inner +---------+------------+------------+------------+
	175	* \| Header \| Not-ECT \| ECT(0) \| ECT(1) \| CE \|
	176	* +---------+---------+------------+------------+------------+
	177	* \| Not-ECT \| Not-ECT \|Not-ECT(!!!)\|Not-ECT(!!!)\| <drop>(!!!)\|
	178	* \| ECT(0) \| ECT(0) \| ECT(0) \| ECT(1) \| CE \|
	179	* \| ECT(1) \| ECT(1) \| ECT(1) (!) \| ECT(1) \| CE \|
	180	* \| CE \| CE \| CE \| CE(!!!)\| CE \|
	181	* +---------+---------+------------+------------+------------+
	182	*
	183	* Figure 4: New IP in IP Decapsulation Behaviour
	184	*
	185	* returns 0 on success
	186	* 1 if something is broken and should be logged (!!! above)
	187	* 2 if packet should be dropped
	188	*/
	189	static inline int INET_ECN_decapsulate(struct sk_buff *skb,
	190	__u8 outer, __u8 inner)
	191	{
	192	if (INET_ECN_is_not_ect(inner)) {
	193	switch (outer & INET_ECN_MASK) {
	194	case INET_ECN_NOT_ECT:
	195	return 0;
	196	case INET_ECN_ECT_0:
	197	case INET_ECN_ECT_1:
	198	return 1;
	199	case INET_ECN_CE:
	200	return 2;
	201	}
	202	}
	203
	204	if (INET_ECN_is_ce(outer))
	205	INET_ECN_set_ce(skb);
	206
	207	return 0;
	208	}
	209
	210	static inline int IP_ECN_decapsulate(const struct iphdr *oiph,
	211	struct sk_buff *skb)
	212	{
	213	__u8 inner;
	214
	215	if (skb->protocol == htons(ETH_P_IP))
	216	inner = ip_hdr(skb)->tos;
	217	else if (skb->protocol == htons(ETH_P_IPV6))
	218	inner = ipv6_get_dsfield(ipv6_hdr(skb));
	219	else
	220	return 0;
	221
	222	return INET_ECN_decapsulate(skb, oiph->tos, inner);
	223	}
	224
	225	static inline int IP6_ECN_decapsulate(const struct ipv6hdr *oipv6h,
	226	struct sk_buff *skb)
	227	{
	228	__u8 inner;
	229
	230	if (skb->protocol == htons(ETH_P_IP))
231	inner = ip_hdr(skb)->tos;
232	else if (skb->protocol == htons(ETH_P_IPV6))
233	inner = ipv6_get_dsfield(ipv6_hdr(skb));
234	else
235	return 0;
236
237	return INET_ECN_decapsulate(skb, ipv6_get_dsfield(oipv6h), inner);
238	}
1da177e4	239	#endif