[deliverable/linux.git] / net / xfrm / xfrm_input.c

/*
 * xfrm_input.c
 *
 * Changes:
 * 	YOSHIFUJI Hideaki @USAGI
 * 		Split up af-specific portion
 *
 */

#include <linux/slab.h>
#include <linux/module.h>
#include <linux/netdevice.h>
#include <net/dst.h>
#include <net/ip.h>
#include <net/xfrm.h>

static struct kmem_cache *secpath_cachep __read_mostly;

void __secpath_destroy(struct sec_path *sp)
{
	int i;
	for (i = 0; i < sp->len; i++)
		xfrm_state_put(sp->xvec[i]);
	kmem_cache_free(secpath_cachep, sp);
}
EXPORT_SYMBOL(__secpath_destroy);

struct sec_path *secpath_dup(struct sec_path *src)
{
	struct sec_path *sp;

	sp = kmem_cache_alloc(secpath_cachep, GFP_ATOMIC);
	if (!sp)
		return NULL;

	sp->len = 0;
	if (src) {
		int i;

		memcpy(sp, src, sizeof(*sp));
		for (i = 0; i < sp->len; i++)
			xfrm_state_hold(sp->xvec[i]);
	}
	atomic_set(&sp->refcnt, 1);
	return sp;
}
EXPORT_SYMBOL(secpath_dup);

/* Fetch spi and seq from ipsec header */

int xfrm_parse_spi(struct sk_buff *skb, u8 nexthdr, __be32 *spi, __be32 *seq)
{
	int offset, offset_seq;
	int hlen;

	switch (nexthdr) {
	case IPPROTO_AH:
		hlen = sizeof(struct ip_auth_hdr);
		offset = offsetof(struct ip_auth_hdr, spi);
		offset_seq = offsetof(struct ip_auth_hdr, seq_no);
		break;
	case IPPROTO_ESP:
		hlen = sizeof(struct ip_esp_hdr);
		offset = offsetof(struct ip_esp_hdr, spi);
		offset_seq = offsetof(struct ip_esp_hdr, seq_no);
		break;
	case IPPROTO_COMP:
		if (!pskb_may_pull(skb, sizeof(struct ip_comp_hdr)))
			return -EINVAL;
		*spi = htonl(ntohs(*(__be16 *)(skb_transport_header(skb) + 2)));
		*seq = 0;
		return 0;
	default:
		return 1;
	}

	if (!pskb_may_pull(skb, hlen))
		return -EINVAL;

	*spi = *(__be32 *)(skb_transport_header(skb) + offset);
	*seq = *(__be32 *)(skb_transport_header(skb) + offset_seq);
	return 0;
}

int xfrm_prepare_input(struct xfrm_state *x, struct sk_buff *skb)
{
	struct xfrm_mode *inner_mode = x->inner_mode;
	int err;

	err = x->outer_mode->afinfo->extract_input(x, skb);
	if (err)
		return err;

	if (x->sel.family == AF_UNSPEC) {
		inner_mode = xfrm_ip2inner_mode(x, XFRM_MODE_SKB_CB(skb)->protocol);
		if (inner_mode == NULL)
			return -EAFNOSUPPORT;
	}

	skb->protocol = inner_mode->afinfo->eth_proto;
	return inner_mode->input2(x, skb);
}
EXPORT_SYMBOL(xfrm_prepare_input);

int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
{
	struct net *net = dev_net(skb->dev);
	int err;
	__be32 seq;
	__be32 seq_hi;
	struct xfrm_state *x = NULL;
	xfrm_address_t *daddr;
	struct xfrm_mode *inner_mode;
	unsigned int family;
	int decaps = 0;
	int async = 0;

	/* A negative encap_type indicates async resumption. */
	if (encap_type < 0) {
		async = 1;
		x = xfrm_input_state(skb);
		seq = XFRM_SKB_CB(skb)->seq.input.low;
		family = x->outer_mode->afinfo->family;
		goto resume;
	}

	daddr = (xfrm_address_t *)(skb_network_header(skb) +
				   XFRM_SPI_SKB_CB(skb)->daddroff);
	family = XFRM_SPI_SKB_CB(skb)->family;

	/* Allocate new secpath or COW existing one. */
	if (!skb->sp || atomic_read(&skb->sp->refcnt) != 1) {
		struct sec_path *sp;

		sp = secpath_dup(skb->sp);
		if (!sp) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINERROR);
			goto drop;
		}
		if (skb->sp)
			secpath_put(skb->sp);
		skb->sp = sp;
	}

	seq = 0;
	if (!spi && (err = xfrm_parse_spi(skb, nexthdr, &spi, &seq)) != 0) {
		XFRM_INC_STATS(net, LINUX_MIB_XFRMINHDRERROR);
		goto drop;
	}

	do {
		if (skb->sp->len == XFRM_MAX_DEPTH) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINBUFFERERROR);
			goto drop;
		}

		x = xfrm_state_lookup(net, skb->mark, daddr, spi, nexthdr, family);
		if (x == NULL) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINNOSTATES);
			xfrm_audit_state_notfound(skb, family, spi, seq);
			goto drop;
		}

		skb->sp->xvec[skb->sp->len++] = x;

		if (xfrm_tunnel_check(skb, x, family)) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMODEERROR);
			goto drop;
		}

		spin_lock(&x->lock);
		if (unlikely(x->km.state == XFRM_STATE_ACQ)) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMACQUIREERROR);
			goto drop_unlock;
		}

		if (unlikely(x->km.state != XFRM_STATE_VALID)) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEINVALID);
			goto drop_unlock;
		}

		if ((x->encap ? x->encap->encap_type : 0) != encap_type) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMISMATCH);
			goto drop_unlock;
		}

		if (x->repl->check(x, skb, seq)) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATESEQERROR);
			goto drop_unlock;
		}

		if (xfrm_state_check_expire(x)) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEEXPIRED);
			goto drop_unlock;
		}

		spin_unlock(&x->lock);

		seq_hi = htonl(xfrm_replay_seqhi(x, seq));

		XFRM_SKB_CB(skb)->seq.input.low = seq;
		XFRM_SKB_CB(skb)->seq.input.hi = seq_hi;

		skb_dst_force(skb);

		nexthdr = x->type->input(x, skb);

		if (nexthdr == -EINPROGRESS)
			return 0;
resume:
		spin_lock(&x->lock);
		if (nexthdr <= 0) {
			if (nexthdr == -EBADMSG) {
				xfrm_audit_state_icvfail(x, skb,
							 x->type->proto);
				x->stats.integrity_failed++;
			}
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEPROTOERROR);
			goto drop_unlock;
		}

		/* only the first xfrm gets the encap type */
		encap_type = 0;

		if (async && x->repl->recheck(x, skb, seq)) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATESEQERROR);
			goto drop_unlock;
		}

		x->repl->advance(x, seq);

		x->curlft.bytes += skb->len;
		x->curlft.packets++;

		spin_unlock(&x->lock);

		XFRM_MODE_SKB_CB(skb)->protocol = nexthdr;

		inner_mode = x->inner_mode;

		if (x->sel.family == AF_UNSPEC) {
			inner_mode = xfrm_ip2inner_mode(x, XFRM_MODE_SKB_CB(skb)->protocol);
			if (inner_mode == NULL)
				goto drop;
		}

		if (inner_mode->input(x, skb)) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMODEERROR);
			goto drop;
		}

		if (x->outer_mode->flags & XFRM_MODE_FLAG_TUNNEL) {
			decaps = 1;
			break;
		}

		/*
		 * We need the inner address.  However, we only get here for
		 * transport mode so the outer address is identical.
		 */
		daddr = &x->id.daddr;
		family = x->outer_mode->afinfo->family;

		err = xfrm_parse_spi(skb, nexthdr, &spi, &seq);
		if (err < 0) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINHDRERROR);
			goto drop;
		}
	} while (!err);

	err = xfrm_rcv_cb(skb, family, x->type->proto, 0);
	if (err)
		goto drop;

	nf_reset(skb);

	if (decaps) {
		skb_dst_drop(skb);
		netif_rx(skb);
		return 0;
	} else {
		return x->inner_mode->afinfo->transport_finish(skb, async);
	}

drop_unlock:
	spin_unlock(&x->lock);
drop:
	xfrm_rcv_cb(skb, family, x && x->type ? x->type->proto : nexthdr, -1);
	kfree_skb(skb);
	return 0;
}
EXPORT_SYMBOL(xfrm_input);

int xfrm_input_resume(struct sk_buff *skb, int nexthdr)
{
	return xfrm_input(skb, nexthdr, 0, -1);
}
EXPORT_SYMBOL(xfrm_input_resume);

void __init xfrm_input_init(void)
{
	secpath_cachep = kmem_cache_create("secpath_cache",
					   sizeof(struct sec_path),
					   0, SLAB_HWCACHE_ALIGN|SLAB_PANIC,
					   NULL);
}
Commit	Line	Data
1da177e4 LT	1	/*
	2	* xfrm_input.c
	3	*
	4	* Changes:
	5	* YOSHIFUJI Hideaki @USAGI
	6	* Split up af-specific portion
a716c119	7	*
1da177e4 LT	8	*/
	9
	10	#include <linux/slab.h>
	11	#include <linux/module.h>
716062fd HX	12	#include <linux/netdevice.h>
716062fd HX	13	#include <net/dst.h>
1da177e4 LT	14	#include <net/ip.h>
	15	#include <net/xfrm.h>
	16
e18b890b	17	static struct kmem_cache *secpath_cachep __read_mostly;
1da177e4 LT	18
	19	void __secpath_destroy(struct sec_path *sp)
	20	{
	21	int i;
	22	for (i = 0; i < sp->len; i++)
dbe5b4aa	23	xfrm_state_put(sp->xvec[i]);
1da177e4 LT	24	kmem_cache_free(secpath_cachep, sp);
	25	}
	26	EXPORT_SYMBOL(__secpath_destroy);
	27
	28	struct sec_path secpath_dup(struct sec_path src)
	29	{
	30	struct sec_path *sp;
	31
54e6ecb2	32	sp = kmem_cache_alloc(secpath_cachep, GFP_ATOMIC);
1da177e4 LT	33	if (!sp)
	34	return NULL;
	35
	36	sp->len = 0;
	37	if (src) {
	38	int i;
	39
	40	memcpy(sp, src, sizeof(*sp));
	41	for (i = 0; i < sp->len; i++)
dbe5b4aa	42	xfrm_state_hold(sp->xvec[i]);
1da177e4 LT	43	}
	44	atomic_set(&sp->refcnt, 1);
	45	return sp;
	46	}
	47	EXPORT_SYMBOL(secpath_dup);
	48
	49	/* Fetch spi and seq from ipsec header */
	50
6067b2ba	51	int xfrm_parse_spi(struct sk_buff skb, u8 nexthdr, __be32 spi, __be32 *seq)
1da177e4 LT	52	{
1da177e4 LT	53	int offset, offset_seq;
44072500	54	int hlen;
1da177e4 LT	55
	56	switch (nexthdr) {
	57	case IPPROTO_AH:
44072500	58	hlen = sizeof(struct ip_auth_hdr);
1da177e4 LT	59	offset = offsetof(struct ip_auth_hdr, spi);
	60	offset_seq = offsetof(struct ip_auth_hdr, seq_no);
	61	break;
	62	case IPPROTO_ESP:
44072500	63	hlen = sizeof(struct ip_esp_hdr);
1da177e4 LT	64	offset = offsetof(struct ip_esp_hdr, spi);
	65	offset_seq = offsetof(struct ip_esp_hdr, seq_no);
	66	break;
	67	case IPPROTO_COMP:
	68	if (!pskb_may_pull(skb, sizeof(struct ip_comp_hdr)))
	69	return -EINVAL;
3e94c2dc	70	spi = htonl(ntohs((__be16 *)(skb_transport_header(skb) + 2)));
1da177e4 LT	71	*seq = 0;
	72	return 0;
	73	default:
	74	return 1;
	75	}
	76
44072500	77	if (!pskb_may_pull(skb, hlen))
1da177e4 LT	78	return -EINVAL;
1da177e4 LT	79
3e94c2dc WC	80	spi = (__be32 *)(skb_transport_header(skb) + offset);
3e94c2dc WC	81	seq = (__be32 *)(skb_transport_header(skb) + offset_seq);
1da177e4 LT	82	return 0;
1da177e4 LT	83	}
1da177e4	84
227620e2 HX	85	int xfrm_prepare_input(struct xfrm_state x, struct sk_buff skb)
227620e2 HX	86	{
df9dcb45	87	struct xfrm_mode *inner_mode = x->inner_mode;
227620e2 HX	88	int err;
	89
	90	err = x->outer_mode->afinfo->extract_input(x, skb);
	91	if (err)
	92	return err;
	93
df9dcb45 KM	94	if (x->sel.family == AF_UNSPEC) {
	95	inner_mode = xfrm_ip2inner_mode(x, XFRM_MODE_SKB_CB(skb)->protocol);
	96	if (inner_mode == NULL)
	97	return -EAFNOSUPPORT;
	98	}
	99
	100	skb->protocol = inner_mode->afinfo->eth_proto;
	101	return inner_mode->input2(x, skb);
227620e2 HX	102	}
	103	EXPORT_SYMBOL(xfrm_prepare_input);
	104
716062fd HX	105	int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
716062fd HX	106	{
bd235e3c	107	struct net *net = dev_net(skb->dev);
716062fd HX	108	int err;
716062fd HX	109	__be32 seq;
2cd08467	110	__be32 seq_hi;
3328715e	111	struct xfrm_state *x = NULL;
1bf06cd2	112	xfrm_address_t *daddr;
df9dcb45	113	struct xfrm_mode *inner_mode;
2fcb45b6	114	unsigned int family;
716062fd	115	int decaps = 0;
1bf06cd2 HX	116	int async = 0;
	117
	118	/* A negative encap_type indicates async resumption. */
	119	if (encap_type < 0) {
	120	async = 1;
00501121	121	x = xfrm_input_state(skb);
1ce3644a	122	seq = XFRM_SKB_CB(skb)->seq.input.low;
3328715e	123	family = x->outer_mode->afinfo->family;
1bf06cd2 HX	124	goto resume;
1bf06cd2 HX	125	}
716062fd	126
3328715e SK	127	daddr = (xfrm_address_t *)(skb_network_header(skb) +
	128	XFRM_SPI_SKB_CB(skb)->daddroff);
	129	family = XFRM_SPI_SKB_CB(skb)->family;
	130
b2aa5e9d HX	131	/* Allocate new secpath or COW existing one. */
	132	if (!skb->sp \|\| atomic_read(&skb->sp->refcnt) != 1) {
	133	struct sec_path *sp;
	134
	135	sp = secpath_dup(skb->sp);
0aa64774	136	if (!sp) {
59c9940e	137	XFRM_INC_STATS(net, LINUX_MIB_XFRMINERROR);
b2aa5e9d	138	goto drop;
0aa64774	139	}
b2aa5e9d HX	140	if (skb->sp)
	141	secpath_put(skb->sp);
	142	skb->sp = sp;
	143	}
	144
716062fd	145	seq = 0;
0aa64774	146	if (!spi && (err = xfrm_parse_spi(skb, nexthdr, &spi, &seq)) != 0) {
59c9940e	147	XFRM_INC_STATS(net, LINUX_MIB_XFRMINHDRERROR);
716062fd	148	goto drop;
0aa64774	149	}
716062fd HX	150
716062fd HX	151	do {
0aa64774	152	if (skb->sp->len == XFRM_MAX_DEPTH) {
59c9940e	153	XFRM_INC_STATS(net, LINUX_MIB_XFRMINBUFFERERROR);
716062fd	154	goto drop;
0aa64774	155	}
716062fd	156
bd55775c	157	x = xfrm_state_lookup(net, skb->mark, daddr, spi, nexthdr, family);
0aa64774	158	if (x == NULL) {
59c9940e	159	XFRM_INC_STATS(net, LINUX_MIB_XFRMINNOSTATES);
afeb14b4	160	xfrm_audit_state_notfound(skb, family, spi, seq);
716062fd	161	goto drop;
0aa64774	162	}
716062fd	163
b2aa5e9d HX	164	skb->sp->xvec[skb->sp->len++] = x;
b2aa5e9d HX	165
70be6c91 SK	166	if (xfrm_tunnel_check(skb, x, family)) {
	167	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMODEERROR);
	168	goto drop;
	169	}
	170
716062fd	171	spin_lock(&x->lock);
4c4d41f2 FD	172	if (unlikely(x->km.state == XFRM_STATE_ACQ)) {
	173	XFRM_INC_STATS(net, LINUX_MIB_XFRMACQUIREERROR);
	174	goto drop_unlock;
	175	}
	176
0aa64774	177	if (unlikely(x->km.state != XFRM_STATE_VALID)) {
59c9940e	178	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEINVALID);
716062fd	179	goto drop_unlock;
0aa64774	180	}
716062fd	181
3de77cf2 DM	182	if ((x->encap ? x->encap->encap_type : 0) != encap_type) {
	183	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMISMATCH);
	184	goto drop_unlock;
	185	}
	186
36ae0148	187	if (x->repl->check(x, skb, seq)) {
59c9940e	188	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATESEQERROR);
716062fd	189	goto drop_unlock;
0aa64774	190	}
716062fd	191
0aa64774	192	if (xfrm_state_check_expire(x)) {
59c9940e	193	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEEXPIRED);
716062fd	194	goto drop_unlock;
0aa64774	195	}
716062fd	196
0ebea8ef HX	197	spin_unlock(&x->lock);
0ebea8ef HX	198
2cd08467 SK	199	seq_hi = htonl(xfrm_replay_seqhi(x, seq));
2cd08467 SK	200
1ce3644a	201	XFRM_SKB_CB(skb)->seq.input.low = seq;
2cd08467	202	XFRM_SKB_CB(skb)->seq.input.hi = seq_hi;
1bf06cd2	203
3bc07321 SK	204	skb_dst_force(skb);
3bc07321 SK	205
716062fd	206	nexthdr = x->type->input(x, skb);
0ebea8ef	207
1bf06cd2 HX	208	if (nexthdr == -EINPROGRESS)
1bf06cd2 HX	209	return 0;
1bf06cd2	210	resume:
0ebea8ef	211	spin_lock(&x->lock);
668dc8af	212	if (nexthdr <= 0) {
9dd3245a HX	213	if (nexthdr == -EBADMSG) {
	214	xfrm_audit_state_icvfail(x, skb,
	215	x->type->proto);
668dc8af	216	x->stats.integrity_failed++;
9dd3245a	217	}
59c9940e	218	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEPROTOERROR);
716062fd	219	goto drop_unlock;
668dc8af	220	}
716062fd	221
716062fd HX	222	/* only the first xfrm gets the encap type */
	223	encap_type = 0;
	224
3b59df46	225	if (async && x->repl->recheck(x, skb, seq)) {
bcf66bf5 SK	226	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATESEQERROR);
	227	goto drop_unlock;
	228	}
	229
9fdc4883	230	x->repl->advance(x, seq);
716062fd HX	231
	232	x->curlft.bytes += skb->len;
	233	x->curlft.packets++;
	234
	235	spin_unlock(&x->lock);
	236
60d5fcfb HX	237	XFRM_MODE_SKB_CB(skb)->protocol = nexthdr;
60d5fcfb HX	238
df9dcb45 KM	239	inner_mode = x->inner_mode;
	240
	241	if (x->sel.family == AF_UNSPEC) {
	242	inner_mode = xfrm_ip2inner_mode(x, XFRM_MODE_SKB_CB(skb)->protocol);
	243	if (inner_mode == NULL)
	244	goto drop;
	245	}
	246
	247	if (inner_mode->input(x, skb)) {
59c9940e	248	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMODEERROR);
716062fd	249	goto drop;
0aa64774	250	}
716062fd HX	251
	252	if (x->outer_mode->flags & XFRM_MODE_FLAG_TUNNEL) {
	253	decaps = 1;
	254	break;
	255	}
	256
1bf06cd2 HX	257	/*
	258	* We need the inner address. However, we only get here for
	259	* transport mode so the outer address is identical.
	260	*/
	261	daddr = &x->id.daddr;
2fcb45b6	262	family = x->outer_mode->afinfo->family;
1bf06cd2	263
716062fd	264	err = xfrm_parse_spi(skb, nexthdr, &spi, &seq);
0aa64774	265	if (err < 0) {
59c9940e	266	XFRM_INC_STATS(net, LINUX_MIB_XFRMINHDRERROR);
716062fd	267	goto drop;
0aa64774	268	}
716062fd HX	269	} while (!err);
716062fd HX	270
3328715e SK	271	err = xfrm_rcv_cb(skb, family, x->type->proto, 0);
	272	if (err)
	273	goto drop;
	274
716062fd HX	275	nf_reset(skb);
	276
	277	if (decaps) {
adf30907	278	skb_dst_drop(skb);
716062fd HX	279	netif_rx(skb);
	280	return 0;
	281	} else {
1bf06cd2	282	return x->inner_mode->afinfo->transport_finish(skb, async);
716062fd HX	283	}
	284
	285	drop_unlock:
	286	spin_unlock(&x->lock);
716062fd	287	drop:
3328715e	288	xfrm_rcv_cb(skb, family, x && x->type ? x->type->proto : nexthdr, -1);
716062fd HX	289	kfree_skb(skb);
	290	return 0;
	291	}
	292	EXPORT_SYMBOL(xfrm_input);
	293
1bf06cd2 HX	294	int xfrm_input_resume(struct sk_buff *skb, int nexthdr)
	295	{
	296	return xfrm_input(skb, nexthdr, 0, -1);
	297	}
	298	EXPORT_SYMBOL(xfrm_input_resume);
	299
1da177e4 LT	300	void __init xfrm_input_init(void)
	301	{
	302	secpath_cachep = kmem_cache_create("secpath_cache",
	303	sizeof(struct sec_path),
e5d679f3	304	0, SLAB_HWCACHE_ALIGN\|SLAB_PANIC,
20c2df83	305	NULL);
1da177e4	306	}