[linux-block.git] / net / xfrm / xfrm_input.c

// SPDX-License-Identifier: GPL-2.0
/*
 * xfrm_input.c
 *
 * Changes:
 * 	YOSHIFUJI Hideaki @USAGI
 * 		Split up af-specific portion
 *
 */

#include <linux/bottom_half.h>
#include <linux/cache.h>
#include <linux/interrupt.h>
#include <linux/slab.h>
#include <linux/module.h>
#include <linux/netdevice.h>
#include <linux/percpu.h>
#include <net/dst.h>
#include <net/ip.h>
#include <net/xfrm.h>
#include <net/ip_tunnels.h>
#include <net/ip6_tunnel.h>

struct xfrm_trans_tasklet {
	struct tasklet_struct tasklet;
	struct sk_buff_head queue;
};

struct xfrm_trans_cb {
	union {
		struct inet_skb_parm	h4;
#if IS_ENABLED(CONFIG_IPV6)
		struct inet6_skb_parm	h6;
#endif
	} header;
	int (*finish)(struct net *net, struct sock *sk, struct sk_buff *skb);
};

#define XFRM_TRANS_SKB_CB(__skb) ((struct xfrm_trans_cb *)&((__skb)->cb[0]))

static struct kmem_cache *secpath_cachep __ro_after_init;

static DEFINE_SPINLOCK(xfrm_input_afinfo_lock);
static struct xfrm_input_afinfo const __rcu *xfrm_input_afinfo[AF_INET6 + 1];

static struct gro_cells gro_cells;
static struct net_device xfrm_napi_dev;

static DEFINE_PER_CPU(struct xfrm_trans_tasklet, xfrm_trans_tasklet);

int xfrm_input_register_afinfo(const struct xfrm_input_afinfo *afinfo)
{
	int err = 0;

	if (WARN_ON(afinfo->family >= ARRAY_SIZE(xfrm_input_afinfo)))
		return -EAFNOSUPPORT;

	spin_lock_bh(&xfrm_input_afinfo_lock);
	if (unlikely(xfrm_input_afinfo[afinfo->family] != NULL))
		err = -EEXIST;
	else
		rcu_assign_pointer(xfrm_input_afinfo[afinfo->family], afinfo);
	spin_unlock_bh(&xfrm_input_afinfo_lock);
	return err;
}
EXPORT_SYMBOL(xfrm_input_register_afinfo);

int xfrm_input_unregister_afinfo(const struct xfrm_input_afinfo *afinfo)
{
	int err = 0;

	spin_lock_bh(&xfrm_input_afinfo_lock);
	if (likely(xfrm_input_afinfo[afinfo->family] != NULL)) {
		if (unlikely(xfrm_input_afinfo[afinfo->family] != afinfo))
			err = -EINVAL;
		else
			RCU_INIT_POINTER(xfrm_input_afinfo[afinfo->family], NULL);
	}
	spin_unlock_bh(&xfrm_input_afinfo_lock);
	synchronize_rcu();
	return err;
}
EXPORT_SYMBOL(xfrm_input_unregister_afinfo);

static const struct xfrm_input_afinfo *xfrm_input_get_afinfo(unsigned int family)
{
	const struct xfrm_input_afinfo *afinfo;

	if (WARN_ON_ONCE(family >= ARRAY_SIZE(xfrm_input_afinfo)))
		return NULL;

	rcu_read_lock();
	afinfo = rcu_dereference(xfrm_input_afinfo[family]);
	if (unlikely(!afinfo))
		rcu_read_unlock();
	return afinfo;
}

static int xfrm_rcv_cb(struct sk_buff *skb, unsigned int family, u8 protocol,
		       int err)
{
	int ret;
	const struct xfrm_input_afinfo *afinfo = xfrm_input_get_afinfo(family);

	if (!afinfo)
		return -EAFNOSUPPORT;

	ret = afinfo->callback(skb, protocol, err);
	rcu_read_unlock();

	return ret;
}

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;
	sp->olen = 0;

	memset(sp->ovec, 0, sizeof(sp->ovec[XFRM_MAX_OFFLOAD_DEPTH]));

	if (src) {
		int i;

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

int secpath_set(struct sk_buff *skb)
{
	struct sec_path *sp;

	/* Allocate new secpath or COW existing one. */
	if (!skb->sp || refcount_read(&skb->sp->refcnt) != 1) {
		sp = secpath_dup(skb->sp);
		if (!sp)
			return -ENOMEM;

		if (skb->sp)
			secpath_put(skb->sp);
		skb->sp = sp;
	}
	return 0;
}
EXPORT_SYMBOL(secpath_set);

/* 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;
}
EXPORT_SYMBOL(xfrm_parse_spi);

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;
	u32 mark = skb->mark;
	unsigned int family = AF_UNSPEC;
	int decaps = 0;
	int async = 0;
	bool xfrm_gro = false;
	bool crypto_done = false;
	struct xfrm_offload *xo = xfrm_offload(skb);

	if (encap_type < 0) {
		x = xfrm_input_state(skb);

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

		family = x->outer_mode->afinfo->family;

		/* An encap_type of -1 indicates async resumption. */
		if (encap_type == -1) {
			async = 1;
			seq = XFRM_SKB_CB(skb)->seq.input.low;
			goto resume;
		}

		/* encap_type < -1 indicates a GRO call. */
		encap_type = 0;
		seq = XFRM_SPI_SKB_CB(skb)->seq;

		if (xo && (xo->flags & CRYPTO_DONE)) {
			crypto_done = true;
			family = XFRM_SPI_SKB_CB(skb)->family;

			if (!(xo->status & CRYPTO_SUCCESS)) {
				if (xo->status &
				    (CRYPTO_TRANSPORT_AH_AUTH_FAILED |
				     CRYPTO_TRANSPORT_ESP_AUTH_FAILED |
				     CRYPTO_TUNNEL_AH_AUTH_FAILED |
				     CRYPTO_TUNNEL_ESP_AUTH_FAILED)) {

					xfrm_audit_state_icvfail(x, skb,
								 x->type->proto);
					x->stats.integrity_failed++;
					XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEPROTOERROR);
					goto drop;
				}

				if (xo->status & CRYPTO_INVALID_PROTOCOL) {
					XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEPROTOERROR);
					goto drop;
				}

				XFRM_INC_STATS(net, LINUX_MIB_XFRMINBUFFERERROR);
				goto drop;
			}

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

		goto lock;
	}

	family = XFRM_SPI_SKB_CB(skb)->family;

	/* if tunnel is present override skb->mark value with tunnel i_key */
	switch (family) {
	case AF_INET:
		if (XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip4)
			mark = be32_to_cpu(XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip4->parms.i_key);
		break;
	case AF_INET6:
		if (XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip6)
			mark = be32_to_cpu(XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip6->parms.i_key);
		break;
	}

	err = secpath_set(skb);
	if (err) {
		XFRM_INC_STATS(net, LINUX_MIB_XFRMINERROR);
		goto drop;
	}

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

	daddr = (xfrm_address_t *)(skb_network_header(skb) +
				   XFRM_SPI_SKB_CB(skb)->daddroff);
	do {
		if (skb->sp->len == XFRM_MAX_DEPTH) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINBUFFERERROR);
			goto drop;
		}

		x = xfrm_state_lookup(net, 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;

lock:
		spin_lock(&x->lock);

		if (unlikely(x->km.state != XFRM_STATE_VALID)) {
			if (x->km.state == XFRM_STATE_ACQ)
				XFRM_INC_STATS(net, LINUX_MIB_XFRMACQUIREERROR);
			else
				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);

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

		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);
		dev_hold(skb->dev);

		if (crypto_done)
			nexthdr = x->type_offload->input_tail(x, skb);
		else
			nexthdr = x->type->input(x, skb);

		if (nexthdr == -EINPROGRESS)
			return 0;
resume:
		dev_put(skb->dev);

		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) {
				XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMODEERROR);
				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) {
		if (skb->sp)
			skb->sp->olen = 0;
		skb_dst_drop(skb);
		gro_cells_receive(&gro_cells, skb);
		return 0;
	} else {
		xo = xfrm_offload(skb);
		if (xo)
			xfrm_gro = xo->flags & XFRM_GRO;

		err = x->inner_mode->afinfo->transport_finish(skb, xfrm_gro || async);
		if (xfrm_gro) {
			if (skb->sp)
				skb->sp->olen = 0;
			skb_dst_drop(skb);
			gro_cells_receive(&gro_cells, skb);
			return err;
		}

		return err;
	}

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);

static void xfrm_trans_reinject(unsigned long data)
{
	struct xfrm_trans_tasklet *trans = (void *)data;
	struct sk_buff_head queue;
	struct sk_buff *skb;

	__skb_queue_head_init(&queue);
	skb_queue_splice_init(&trans->queue, &queue);

	while ((skb = __skb_dequeue(&queue)))
		XFRM_TRANS_SKB_CB(skb)->finish(dev_net(skb->dev), NULL, skb);
}

int xfrm_trans_queue(struct sk_buff *skb,
		     int (*finish)(struct net *, struct sock *,
				   struct sk_buff *))
{
	struct xfrm_trans_tasklet *trans;

	trans = this_cpu_ptr(&xfrm_trans_tasklet);

	if (skb_queue_len(&trans->queue) >= netdev_max_backlog)
		return -ENOBUFS;

	XFRM_TRANS_SKB_CB(skb)->finish = finish;
	__skb_queue_tail(&trans->queue, skb);
	tasklet_schedule(&trans->tasklet);
	return 0;
}
EXPORT_SYMBOL(xfrm_trans_queue);

void __init xfrm_input_init(void)
{
	int err;
	int i;

	init_dummy_netdev(&xfrm_napi_dev);
	err = gro_cells_init(&gro_cells, &xfrm_napi_dev);
	if (err)
		gro_cells.cells = NULL;

	secpath_cachep = kmem_cache_create("secpath_cache",
					   sizeof(struct sec_path),
					   0, SLAB_HWCACHE_ALIGN|SLAB_PANIC,
					   NULL);

	for_each_possible_cpu(i) {
		struct xfrm_trans_tasklet *trans;

		trans = &per_cpu(xfrm_trans_tasklet, i);
		__skb_queue_head_init(&trans->queue);
		tasklet_init(&trans->tasklet, xfrm_trans_reinject,
			     (unsigned long)trans);
	}
}
Commit	Line	Data
b2441318	1	// SPDX-License-Identifier: GPL-2.0
1da177e4 LT	2	/*
	3	* xfrm_input.c
	4	*
	5	* Changes:
	6	* YOSHIFUJI Hideaki @USAGI
	7	* Split up af-specific portion
a716c119	8	*
1da177e4 LT	9	*/
1da177e4 LT	10
acf568ee	11	#include <linux/bottom_half.h>
f8c3d0dd	12	#include <linux/cache.h>
acf568ee	13	#include <linux/interrupt.h>
1da177e4 LT	14	#include <linux/slab.h>
1da177e4 LT	15	#include <linux/module.h>
716062fd	16	#include <linux/netdevice.h>
acf568ee	17	#include <linux/percpu.h>
716062fd	18	#include <net/dst.h>
1da177e4 LT	19	#include <net/ip.h>
1da177e4 LT	20	#include <net/xfrm.h>
049f8e2e AD	21	#include <net/ip_tunnels.h>
049f8e2e AD	22	#include <net/ip6_tunnel.h>
1da177e4	23
acf568ee HX	24	struct xfrm_trans_tasklet {
	25	struct tasklet_struct tasklet;
	26	struct sk_buff_head queue;
	27	};
	28
	29	struct xfrm_trans_cb {
9a3fb9fb SK	30	union {
	31	struct inet_skb_parm h4;
	32	#if IS_ENABLED(CONFIG_IPV6)
	33	struct inet6_skb_parm h6;
	34	#endif
	35	} header;
acf568ee HX	36	int (finish)(struct net net, struct sock sk, struct sk_buff skb);
	37	};
	38
	39	#define XFRM_TRANS_SKB_CB(__skb) ((struct xfrm_trans_cb *)&((__skb)->cb[0]))
	40
f8c3d0dd	41	static struct kmem_cache *secpath_cachep __ro_after_init;
1da177e4	42
2f32b51b	43	static DEFINE_SPINLOCK(xfrm_input_afinfo_lock);
960fdfde	44	static struct xfrm_input_afinfo const __rcu *xfrm_input_afinfo[AF_INET6 + 1];
2f32b51b	45
1995876a SK	46	static struct gro_cells gro_cells;
	47	static struct net_device xfrm_napi_dev;
	48
acf568ee HX	49	static DEFINE_PER_CPU(struct xfrm_trans_tasklet, xfrm_trans_tasklet);
acf568ee HX	50
960fdfde	51	int xfrm_input_register_afinfo(const struct xfrm_input_afinfo *afinfo)
2f32b51b SK	52	{
	53	int err = 0;
	54
960fdfde	55	if (WARN_ON(afinfo->family >= ARRAY_SIZE(xfrm_input_afinfo)))
2f32b51b	56	return -EAFNOSUPPORT;
960fdfde	57
2f32b51b SK	58	spin_lock_bh(&xfrm_input_afinfo_lock);
2f32b51b SK	59	if (unlikely(xfrm_input_afinfo[afinfo->family] != NULL))
f31e8d4f	60	err = -EEXIST;
2f32b51b SK	61	else
	62	rcu_assign_pointer(xfrm_input_afinfo[afinfo->family], afinfo);
	63	spin_unlock_bh(&xfrm_input_afinfo_lock);
	64	return err;
	65	}
	66	EXPORT_SYMBOL(xfrm_input_register_afinfo);
	67
960fdfde	68	int xfrm_input_unregister_afinfo(const struct xfrm_input_afinfo *afinfo)
2f32b51b SK	69	{
	70	int err = 0;
	71
2f32b51b SK	72	spin_lock_bh(&xfrm_input_afinfo_lock);
	73	if (likely(xfrm_input_afinfo[afinfo->family] != NULL)) {
	74	if (unlikely(xfrm_input_afinfo[afinfo->family] != afinfo))
	75	err = -EINVAL;
	76	else
	77	RCU_INIT_POINTER(xfrm_input_afinfo[afinfo->family], NULL);
	78	}
	79	spin_unlock_bh(&xfrm_input_afinfo_lock);
	80	synchronize_rcu();
	81	return err;
	82	}
	83	EXPORT_SYMBOL(xfrm_input_unregister_afinfo);
	84
960fdfde	85	static const struct xfrm_input_afinfo *xfrm_input_get_afinfo(unsigned int family)
2f32b51b	86	{
960fdfde	87	const struct xfrm_input_afinfo *afinfo;
2f32b51b	88
960fdfde	89	if (WARN_ON_ONCE(family >= ARRAY_SIZE(xfrm_input_afinfo)))
2f32b51b	90	return NULL;
960fdfde	91
2f32b51b SK	92	rcu_read_lock();
	93	afinfo = rcu_dereference(xfrm_input_afinfo[family]);
	94	if (unlikely(!afinfo))
	95	rcu_read_unlock();
	96	return afinfo;
	97	}
	98
2f32b51b SK	99	static int xfrm_rcv_cb(struct sk_buff *skb, unsigned int family, u8 protocol,
	100	int err)
	101	{
	102	int ret;
960fdfde	103	const struct xfrm_input_afinfo *afinfo = xfrm_input_get_afinfo(family);
2f32b51b SK	104
	105	if (!afinfo)
	106	return -EAFNOSUPPORT;
	107
	108	ret = afinfo->callback(skb, protocol, err);
960fdfde	109	rcu_read_unlock();
2f32b51b SK	110
	111	return ret;
	112	}
	113
1da177e4 LT	114	void __secpath_destroy(struct sec_path *sp)
	115	{
	116	int i;
	117	for (i = 0; i < sp->len; i++)
dbe5b4aa	118	xfrm_state_put(sp->xvec[i]);
1da177e4 LT	119	kmem_cache_free(secpath_cachep, sp);
	120	}
	121	EXPORT_SYMBOL(__secpath_destroy);
	122
	123	struct sec_path secpath_dup(struct sec_path src)
	124	{
	125	struct sec_path *sp;
	126
54e6ecb2	127	sp = kmem_cache_alloc(secpath_cachep, GFP_ATOMIC);
1da177e4 LT	128	if (!sp)
	129	return NULL;
	130
	131	sp->len = 0;
54ef207a SK	132	sp->olen = 0;
54ef207a SK	133
d77e38e6 SK	134	memset(sp->ovec, 0, sizeof(sp->ovec[XFRM_MAX_OFFLOAD_DEPTH]));
d77e38e6 SK	135
1da177e4 LT	136	if (src) {
	137	int i;
	138
	139	memcpy(sp, src, sizeof(*sp));
	140	for (i = 0; i < sp->len; i++)
dbe5b4aa	141	xfrm_state_hold(sp->xvec[i]);
1da177e4	142	}
55eabed6	143	refcount_set(&sp->refcnt, 1);
1da177e4 LT	144	return sp;
	145	}
	146	EXPORT_SYMBOL(secpath_dup);
	147
b0fcee82 SK	148	int secpath_set(struct sk_buff *skb)
	149	{
	150	struct sec_path *sp;
	151
	152	/* Allocate new secpath or COW existing one. */
55eabed6	153	if (!skb->sp \|\| refcount_read(&skb->sp->refcnt) != 1) {
b0fcee82 SK	154	sp = secpath_dup(skb->sp);
	155	if (!sp)
	156	return -ENOMEM;
	157
	158	if (skb->sp)
	159	secpath_put(skb->sp);
	160	skb->sp = sp;
	161	}
	162	return 0;
	163	}
	164	EXPORT_SYMBOL(secpath_set);
	165
1da177e4 LT	166	/* Fetch spi and seq from ipsec header */
1da177e4 LT	167
6067b2ba	168	int xfrm_parse_spi(struct sk_buff skb, u8 nexthdr, __be32 spi, __be32 *seq)
1da177e4 LT	169	{
1da177e4 LT	170	int offset, offset_seq;
44072500	171	int hlen;
1da177e4 LT	172
	173	switch (nexthdr) {
	174	case IPPROTO_AH:
44072500	175	hlen = sizeof(struct ip_auth_hdr);
1da177e4 LT	176	offset = offsetof(struct ip_auth_hdr, spi);
	177	offset_seq = offsetof(struct ip_auth_hdr, seq_no);
	178	break;
	179	case IPPROTO_ESP:
44072500	180	hlen = sizeof(struct ip_esp_hdr);
1da177e4 LT	181	offset = offsetof(struct ip_esp_hdr, spi);
	182	offset_seq = offsetof(struct ip_esp_hdr, seq_no);
	183	break;
	184	case IPPROTO_COMP:
	185	if (!pskb_may_pull(skb, sizeof(struct ip_comp_hdr)))
	186	return -EINVAL;
3e94c2dc	187	spi = htonl(ntohs((__be16 *)(skb_transport_header(skb) + 2)));
1da177e4 LT	188	*seq = 0;
	189	return 0;
	190	default:
	191	return 1;
	192	}
	193
44072500	194	if (!pskb_may_pull(skb, hlen))
1da177e4 LT	195	return -EINVAL;
1da177e4 LT	196
3e94c2dc WC	197	spi = (__be32 *)(skb_transport_header(skb) + offset);
3e94c2dc WC	198	seq = (__be32 *)(skb_transport_header(skb) + offset_seq);
1da177e4 LT	199	return 0;
1da177e4 LT	200	}
1e295370	201	EXPORT_SYMBOL(xfrm_parse_spi);
1da177e4	202
227620e2 HX	203	int xfrm_prepare_input(struct xfrm_state x, struct sk_buff skb)
227620e2 HX	204	{
df9dcb45	205	struct xfrm_mode *inner_mode = x->inner_mode;
227620e2 HX	206	int err;
	207
	208	err = x->outer_mode->afinfo->extract_input(x, skb);
	209	if (err)
	210	return err;
	211
df9dcb45 KM	212	if (x->sel.family == AF_UNSPEC) {
	213	inner_mode = xfrm_ip2inner_mode(x, XFRM_MODE_SKB_CB(skb)->protocol);
	214	if (inner_mode == NULL)
	215	return -EAFNOSUPPORT;
	216	}
	217
	218	skb->protocol = inner_mode->afinfo->eth_proto;
	219	return inner_mode->input2(x, skb);
227620e2 HX	220	}
	221	EXPORT_SYMBOL(xfrm_prepare_input);
	222
716062fd HX	223	int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
716062fd HX	224	{
bd235e3c	225	struct net *net = dev_net(skb->dev);
716062fd HX	226	int err;
716062fd HX	227	__be32 seq;
2cd08467	228	__be32 seq_hi;
3328715e	229	struct xfrm_state *x = NULL;
1bf06cd2	230	xfrm_address_t *daddr;
df9dcb45	231	struct xfrm_mode *inner_mode;
049f8e2e	232	u32 mark = skb->mark;
4ce3dbe3	233	unsigned int family = AF_UNSPEC;
716062fd	234	int decaps = 0;
1bf06cd2	235	int async = 0;
7785bba2	236	bool xfrm_gro = false;
d77e38e6 SK	237	bool crypto_done = false;
d77e38e6 SK	238	struct xfrm_offload *xo = xfrm_offload(skb);
1bf06cd2	239
1bf06cd2	240	if (encap_type < 0) {
00501121	241	x = xfrm_input_state(skb);
4ce3dbe3 AH	242
	243	if (unlikely(x->km.state != XFRM_STATE_VALID)) {
	244	if (x->km.state == XFRM_STATE_ACQ)
	245	XFRM_INC_STATS(net, LINUX_MIB_XFRMACQUIREERROR);
	246	else
	247	XFRM_INC_STATS(net,
	248	LINUX_MIB_XFRMINSTATEINVALID);
	249	goto drop;
	250	}
	251
3328715e	252	family = x->outer_mode->afinfo->family;
7785bba2 SK	253
	254	/* An encap_type of -1 indicates async resumption. */
	255	if (encap_type == -1) {
	256	async = 1;
	257	seq = XFRM_SKB_CB(skb)->seq.input.low;
	258	goto resume;
	259	}
bcd1f8a4	260
7785bba2 SK	261	/* encap_type < -1 indicates a GRO call. */
	262	encap_type = 0;
	263	seq = XFRM_SPI_SKB_CB(skb)->seq;
716062fd	264
bcd1f8a4 SK	265	if (xo && (xo->flags & CRYPTO_DONE)) {
bcd1f8a4 SK	266	crypto_done = true;
bcd1f8a4 SK	267	family = XFRM_SPI_SKB_CB(skb)->family;
	268
	269	if (!(xo->status & CRYPTO_SUCCESS)) {
	270	if (xo->status &
	271	(CRYPTO_TRANSPORT_AH_AUTH_FAILED \|
	272	CRYPTO_TRANSPORT_ESP_AUTH_FAILED \|
	273	CRYPTO_TUNNEL_AH_AUTH_FAILED \|
	274	CRYPTO_TUNNEL_ESP_AUTH_FAILED)) {
	275
	276	xfrm_audit_state_icvfail(x, skb,
	277	x->type->proto);
	278	x->stats.integrity_failed++;
	279	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEPROTOERROR);
	280	goto drop;
	281	}
	282
47ebcc0b YK	283	if (xo->status & CRYPTO_INVALID_PROTOCOL) {
	284	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEPROTOERROR);
	285	goto drop;
	286	}
	287
bcd1f8a4	288	XFRM_INC_STATS(net, LINUX_MIB_XFRMINBUFFERERROR);
d77e38e6 SK	289	goto drop;
	290	}
	291
bcd1f8a4 SK	292	if ((err = xfrm_parse_spi(skb, nexthdr, &spi, &seq)) != 0) {
	293	XFRM_INC_STATS(net, LINUX_MIB_XFRMINHDRERROR);
	294	goto drop;
	295	}
d77e38e6 SK	296	}
d77e38e6 SK	297
7785bba2	298	goto lock;
1bf06cd2	299	}
716062fd	300
3328715e SK	301	family = XFRM_SPI_SKB_CB(skb)->family;
3328715e SK	302
049f8e2e	303	/* if tunnel is present override skb->mark value with tunnel i_key */
1625f452 AK	304	switch (family) {
	305	case AF_INET:
	306	if (XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip4)
049f8e2e	307	mark = be32_to_cpu(XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip4->parms.i_key);
1625f452 AK	308	break;
	309	case AF_INET6:
	310	if (XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip6)
049f8e2e	311	mark = be32_to_cpu(XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip6->parms.i_key);
1625f452	312	break;
049f8e2e AD	313	}
049f8e2e AD	314
b0fcee82 SK	315	err = secpath_set(skb);
	316	if (err) {
	317	XFRM_INC_STATS(net, LINUX_MIB_XFRMINERROR);
	318	goto drop;
b2aa5e9d HX	319	}
b2aa5e9d HX	320
716062fd	321	seq = 0;
0aa64774	322	if (!spi && (err = xfrm_parse_spi(skb, nexthdr, &spi, &seq)) != 0) {
59c9940e	323	XFRM_INC_STATS(net, LINUX_MIB_XFRMINHDRERROR);
716062fd	324	goto drop;
0aa64774	325	}
716062fd	326
cb79a180 FW	327	daddr = (xfrm_address_t *)(skb_network_header(skb) +
cb79a180 FW	328	XFRM_SPI_SKB_CB(skb)->daddroff);
716062fd	329	do {
0aa64774	330	if (skb->sp->len == XFRM_MAX_DEPTH) {
59c9940e	331	XFRM_INC_STATS(net, LINUX_MIB_XFRMINBUFFERERROR);
716062fd	332	goto drop;
0aa64774	333	}
716062fd	334
049f8e2e	335	x = xfrm_state_lookup(net, mark, daddr, spi, nexthdr, family);
0aa64774	336	if (x == NULL) {
59c9940e	337	XFRM_INC_STATS(net, LINUX_MIB_XFRMINNOSTATES);
afeb14b4	338	xfrm_audit_state_notfound(skb, family, spi, seq);
716062fd	339	goto drop;
0aa64774	340	}
716062fd	341
b2aa5e9d HX	342	skb->sp->xvec[skb->sp->len++] = x;
b2aa5e9d HX	343
7785bba2	344	lock:
716062fd	345	spin_lock(&x->lock);
4c4d41f2	346
0aa64774	347	if (unlikely(x->km.state != XFRM_STATE_VALID)) {
dc0565ce LR	348	if (x->km.state == XFRM_STATE_ACQ)
	349	XFRM_INC_STATS(net, LINUX_MIB_XFRMACQUIREERROR);
	350	else
	351	XFRM_INC_STATS(net,
	352	LINUX_MIB_XFRMINSTATEINVALID);
716062fd	353	goto drop_unlock;
0aa64774	354	}
716062fd	355
3de77cf2 DM	356	if ((x->encap ? x->encap->encap_type : 0) != encap_type) {
	357	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMISMATCH);
	358	goto drop_unlock;
	359	}
	360
36ae0148	361	if (x->repl->check(x, skb, seq)) {
59c9940e	362	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATESEQERROR);
716062fd	363	goto drop_unlock;
0aa64774	364	}
716062fd	365
0aa64774	366	if (xfrm_state_check_expire(x)) {
59c9940e	367	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEEXPIRED);
716062fd	368	goto drop_unlock;
0aa64774	369	}
716062fd	370
0ebea8ef HX	371	spin_unlock(&x->lock);
0ebea8ef HX	372
68c11e98 AD	373	if (xfrm_tunnel_check(skb, x, family)) {
	374	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMODEERROR);
	375	goto drop;
	376	}
	377
2cd08467 SK	378	seq_hi = htonl(xfrm_replay_seqhi(x, seq));
2cd08467 SK	379
1ce3644a	380	XFRM_SKB_CB(skb)->seq.input.low = seq;
2cd08467	381	XFRM_SKB_CB(skb)->seq.input.hi = seq_hi;
1bf06cd2	382
3bc07321	383	skb_dst_force(skb);
071d36bf	384	dev_hold(skb->dev);
3bc07321	385
d77e38e6 SK	386	if (crypto_done)
	387	nexthdr = x->type_offload->input_tail(x, skb);
	388	else
	389	nexthdr = x->type->input(x, skb);
0ebea8ef	390
1bf06cd2 HX	391	if (nexthdr == -EINPROGRESS)
1bf06cd2 HX	392	return 0;
1bf06cd2	393	resume:
071d36bf	394	dev_put(skb->dev);
071d36bf	395
0ebea8ef	396	spin_lock(&x->lock);
668dc8af	397	if (nexthdr <= 0) {
9dd3245a HX	398	if (nexthdr == -EBADMSG) {
	399	xfrm_audit_state_icvfail(x, skb,
	400	x->type->proto);
668dc8af	401	x->stats.integrity_failed++;
9dd3245a	402	}
59c9940e	403	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEPROTOERROR);
716062fd	404	goto drop_unlock;
668dc8af	405	}
716062fd	406
716062fd HX	407	/* only the first xfrm gets the encap type */
	408	encap_type = 0;
	409
3b59df46	410	if (async && x->repl->recheck(x, skb, seq)) {
bcf66bf5 SK	411	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATESEQERROR);
	412	goto drop_unlock;
	413	}
	414
9fdc4883	415	x->repl->advance(x, seq);
716062fd HX	416
	417	x->curlft.bytes += skb->len;
	418	x->curlft.packets++;
	419
	420	spin_unlock(&x->lock);
	421
60d5fcfb HX	422	XFRM_MODE_SKB_CB(skb)->protocol = nexthdr;
60d5fcfb HX	423
df9dcb45 KM	424	inner_mode = x->inner_mode;
	425
	426	if (x->sel.family == AF_UNSPEC) {
	427	inner_mode = xfrm_ip2inner_mode(x, XFRM_MODE_SKB_CB(skb)->protocol);
cb866e32 SK	428	if (inner_mode == NULL) {
cb866e32 SK	429	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMODEERROR);
df9dcb45	430	goto drop;
cb866e32	431	}
df9dcb45 KM	432	}
	433
	434	if (inner_mode->input(x, skb)) {
59c9940e	435	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMODEERROR);
716062fd	436	goto drop;
0aa64774	437	}
716062fd HX	438
	439	if (x->outer_mode->flags & XFRM_MODE_FLAG_TUNNEL) {
	440	decaps = 1;
	441	break;
	442	}
	443
1bf06cd2 HX	444	/*
	445	* We need the inner address. However, we only get here for
	446	* transport mode so the outer address is identical.
	447	*/
	448	daddr = &x->id.daddr;
2fcb45b6	449	family = x->outer_mode->afinfo->family;
1bf06cd2	450
716062fd	451	err = xfrm_parse_spi(skb, nexthdr, &spi, &seq);
0aa64774	452	if (err < 0) {
59c9940e	453	XFRM_INC_STATS(net, LINUX_MIB_XFRMINHDRERROR);
716062fd	454	goto drop;
0aa64774	455	}
716062fd HX	456	} while (!err);
716062fd HX	457
3328715e SK	458	err = xfrm_rcv_cb(skb, family, x->type->proto, 0);
	459	if (err)
	460	goto drop;
	461
716062fd HX	462	nf_reset(skb);
	463
	464	if (decaps) {
23e9fcfe AK	465	if (skb->sp)
23e9fcfe AK	466	skb->sp->olen = 0;
adf30907	467	skb_dst_drop(skb);
1995876a	468	gro_cells_receive(&gro_cells, skb);
716062fd HX	469	return 0;
716062fd HX	470	} else {
7785bba2 SK	471	xo = xfrm_offload(skb);
	472	if (xo)
	473	xfrm_gro = xo->flags & XFRM_GRO;
	474
cfcf99f9	475	err = x->inner_mode->afinfo->transport_finish(skb, xfrm_gro \|\| async);
7785bba2	476	if (xfrm_gro) {
23e9fcfe AK	477	if (skb->sp)
23e9fcfe AK	478	skb->sp->olen = 0;
7785bba2 SK	479	skb_dst_drop(skb);
	480	gro_cells_receive(&gro_cells, skb);
	481	return err;
	482	}
	483
	484	return err;
716062fd HX	485	}
	486
	487	drop_unlock:
	488	spin_unlock(&x->lock);
716062fd	489	drop:
3328715e	490	xfrm_rcv_cb(skb, family, x && x->type ? x->type->proto : nexthdr, -1);
716062fd HX	491	kfree_skb(skb);
	492	return 0;
	493	}
	494	EXPORT_SYMBOL(xfrm_input);
	495
1bf06cd2 HX	496	int xfrm_input_resume(struct sk_buff *skb, int nexthdr)
	497	{
	498	return xfrm_input(skb, nexthdr, 0, -1);
	499	}
	500	EXPORT_SYMBOL(xfrm_input_resume);
	501
acf568ee HX	502	static void xfrm_trans_reinject(unsigned long data)
	503	{
	504	struct xfrm_trans_tasklet trans = (void )data;
	505	struct sk_buff_head queue;
	506	struct sk_buff *skb;
	507
	508	__skb_queue_head_init(&queue);
	509	skb_queue_splice_init(&trans->queue, &queue);
	510
	511	while ((skb = __skb_dequeue(&queue)))
	512	XFRM_TRANS_SKB_CB(skb)->finish(dev_net(skb->dev), NULL, skb);
	513	}
	514
	515	int xfrm_trans_queue(struct sk_buff *skb,
	516	int (finish)(struct net , struct sock *,
	517	struct sk_buff *))
	518	{
	519	struct xfrm_trans_tasklet *trans;
	520
	521	trans = this_cpu_ptr(&xfrm_trans_tasklet);
	522
	523	if (skb_queue_len(&trans->queue) >= netdev_max_backlog)
	524	return -ENOBUFS;
	525
	526	XFRM_TRANS_SKB_CB(skb)->finish = finish;
d16b46e4	527	__skb_queue_tail(&trans->queue, skb);
acf568ee HX	528	tasklet_schedule(&trans->tasklet);
	529	return 0;
	530	}
	531	EXPORT_SYMBOL(xfrm_trans_queue);
	532
1da177e4 LT	533	void __init xfrm_input_init(void)
1da177e4 LT	534	{
1995876a	535	int err;
acf568ee	536	int i;
1995876a SK	537
	538	init_dummy_netdev(&xfrm_napi_dev);
	539	err = gro_cells_init(&gro_cells, &xfrm_napi_dev);
	540	if (err)
	541	gro_cells.cells = NULL;
	542
1da177e4 LT	543	secpath_cachep = kmem_cache_create("secpath_cache",
1da177e4 LT	544	sizeof(struct sec_path),
e5d679f3	545	0, SLAB_HWCACHE_ALIGN\|SLAB_PANIC,
20c2df83	546	NULL);
acf568ee HX	547
	548	for_each_possible_cpu(i) {
	549	struct xfrm_trans_tasklet *trans;
	550
	551	trans = &per_cpu(xfrm_trans_tasklet, i);
	552	__skb_queue_head_init(&trans->queue);
	553	tasklet_init(&trans->tasklet, xfrm_trans_reinject,
	554	(unsigned long)trans);
	555	}
1da177e4	556	}