[mirror_ubuntu-jammy-kernel.git] / net / netfilter / nf_flow_table_core.c

#include <linux/kernel.h>
#include <linux/init.h>
#include <linux/module.h>
#include <linux/netfilter.h>
#include <linux/rhashtable.h>
#include <linux/netdevice.h>
#include <net/ip.h>
#include <net/ip6_route.h>
#include <net/netfilter/nf_tables.h>
#include <net/netfilter/nf_flow_table.h>
#include <net/netfilter/nf_conntrack.h>
#include <net/netfilter/nf_conntrack_core.h>
#include <net/netfilter/nf_conntrack_tuple.h>

struct flow_offload_entry {
	struct flow_offload	flow;
	struct nf_conn		*ct;
	struct rcu_head		rcu_head;
};

static DEFINE_MUTEX(flowtable_lock);
static LIST_HEAD(flowtables);

static void
flow_offload_fill_dir(struct flow_offload *flow, struct nf_conn *ct,
		      struct nf_flow_route *route,
		      enum flow_offload_tuple_dir dir)
{
	struct flow_offload_tuple *ft = &flow->tuplehash[dir].tuple;
	struct nf_conntrack_tuple *ctt = &ct->tuplehash[dir].tuple;
	struct dst_entry *dst = route->tuple[dir].dst;

	ft->dir = dir;

	switch (ctt->src.l3num) {
	case NFPROTO_IPV4:
		ft->src_v4 = ctt->src.u3.in;
		ft->dst_v4 = ctt->dst.u3.in;
		ft->mtu = ip_dst_mtu_maybe_forward(dst, true);
		break;
	case NFPROTO_IPV6:
		ft->src_v6 = ctt->src.u3.in6;
		ft->dst_v6 = ctt->dst.u3.in6;
		ft->mtu = ip6_dst_mtu_forward(dst);
		break;
	}

	ft->l3proto = ctt->src.l3num;
	ft->l4proto = ctt->dst.protonum;
	ft->src_port = ctt->src.u.tcp.port;
	ft->dst_port = ctt->dst.u.tcp.port;

	ft->iifidx = route->tuple[dir].ifindex;
	ft->oifidx = route->tuple[!dir].ifindex;
	ft->dst_cache = dst;
}

struct flow_offload *
flow_offload_alloc(struct nf_conn *ct, struct nf_flow_route *route)
{
	struct flow_offload_entry *entry;
	struct flow_offload *flow;

	if (unlikely(nf_ct_is_dying(ct) ||
	    !atomic_inc_not_zero(&ct->ct_general.use)))
		return NULL;

	entry = kzalloc(sizeof(*entry), GFP_ATOMIC);
	if (!entry)
		goto err_ct_refcnt;

	flow = &entry->flow;

	if (!dst_hold_safe(route->tuple[FLOW_OFFLOAD_DIR_ORIGINAL].dst))
		goto err_dst_cache_original;

	if (!dst_hold_safe(route->tuple[FLOW_OFFLOAD_DIR_REPLY].dst))
		goto err_dst_cache_reply;

	entry->ct = ct;

	flow_offload_fill_dir(flow, ct, route, FLOW_OFFLOAD_DIR_ORIGINAL);
	flow_offload_fill_dir(flow, ct, route, FLOW_OFFLOAD_DIR_REPLY);

	if (ct->status & IPS_SRC_NAT)
		flow->flags |= FLOW_OFFLOAD_SNAT;
	else if (ct->status & IPS_DST_NAT)
		flow->flags |= FLOW_OFFLOAD_DNAT;

	return flow;

err_dst_cache_reply:
	dst_release(route->tuple[FLOW_OFFLOAD_DIR_ORIGINAL].dst);
err_dst_cache_original:
	kfree(entry);
err_ct_refcnt:
	nf_ct_put(ct);

	return NULL;
}
EXPORT_SYMBOL_GPL(flow_offload_alloc);

void flow_offload_free(struct flow_offload *flow)
{
	struct flow_offload_entry *e;

	dst_release(flow->tuplehash[FLOW_OFFLOAD_DIR_ORIGINAL].tuple.dst_cache);
	dst_release(flow->tuplehash[FLOW_OFFLOAD_DIR_REPLY].tuple.dst_cache);
	e = container_of(flow, struct flow_offload_entry, flow);
	nf_ct_delete(e->ct, 0, 0);
	nf_ct_put(e->ct);
	kfree_rcu(e, rcu_head);
}
EXPORT_SYMBOL_GPL(flow_offload_free);

static u32 flow_offload_hash(const void *data, u32 len, u32 seed)
{
	const struct flow_offload_tuple *tuple = data;

	return jhash(tuple, offsetof(struct flow_offload_tuple, dir), seed);
}

static u32 flow_offload_hash_obj(const void *data, u32 len, u32 seed)
{
	const struct flow_offload_tuple_rhash *tuplehash = data;

	return jhash(&tuplehash->tuple, offsetof(struct flow_offload_tuple, dir), seed);
}

static int flow_offload_hash_cmp(struct rhashtable_compare_arg *arg,
					const void *ptr)
{
	const struct flow_offload_tuple *tuple = arg->key;
	const struct flow_offload_tuple_rhash *x = ptr;

	if (memcmp(&x->tuple, tuple, offsetof(struct flow_offload_tuple, dir)))
		return 1;

	return 0;
}

static const struct rhashtable_params nf_flow_offload_rhash_params = {
	.head_offset		= offsetof(struct flow_offload_tuple_rhash, node),
	.hashfn			= flow_offload_hash,
	.obj_hashfn		= flow_offload_hash_obj,
	.obj_cmpfn		= flow_offload_hash_cmp,
	.automatic_shrinking	= true,
};

int flow_offload_add(struct nf_flowtable *flow_table, struct flow_offload *flow)
{
	flow->timeout = (u32)jiffies;

	rhashtable_insert_fast(&flow_table->rhashtable,
			       &flow->tuplehash[FLOW_OFFLOAD_DIR_ORIGINAL].node,
			       nf_flow_offload_rhash_params);
	rhashtable_insert_fast(&flow_table->rhashtable,
			       &flow->tuplehash[FLOW_OFFLOAD_DIR_REPLY].node,
			       nf_flow_offload_rhash_params);
	return 0;
}
EXPORT_SYMBOL_GPL(flow_offload_add);

static void flow_offload_del(struct nf_flowtable *flow_table,
			     struct flow_offload *flow)
{
	rhashtable_remove_fast(&flow_table->rhashtable,
			       &flow->tuplehash[FLOW_OFFLOAD_DIR_ORIGINAL].node,
			       nf_flow_offload_rhash_params);
	rhashtable_remove_fast(&flow_table->rhashtable,
			       &flow->tuplehash[FLOW_OFFLOAD_DIR_REPLY].node,
			       nf_flow_offload_rhash_params);

	flow_offload_free(flow);
}

void flow_offload_teardown(struct flow_offload *flow)
{
	flow->flags |= FLOW_OFFLOAD_TEARDOWN;
}
EXPORT_SYMBOL_GPL(flow_offload_teardown);

struct flow_offload_tuple_rhash *
flow_offload_lookup(struct nf_flowtable *flow_table,
		    struct flow_offload_tuple *tuple)
{
	return rhashtable_lookup_fast(&flow_table->rhashtable, tuple,
				      nf_flow_offload_rhash_params);
}
EXPORT_SYMBOL_GPL(flow_offload_lookup);

int nf_flow_table_iterate(struct nf_flowtable *flow_table,
			  void (*iter)(struct flow_offload *flow, void *data),
			  void *data)
{
	struct flow_offload_tuple_rhash *tuplehash;
	struct rhashtable_iter hti;
	struct flow_offload *flow;
	int err;

	err = rhashtable_walk_init(&flow_table->rhashtable, &hti, GFP_KERNEL);
	if (err)
		return err;

	rhashtable_walk_start(&hti);

	while ((tuplehash = rhashtable_walk_next(&hti))) {
		if (IS_ERR(tuplehash)) {
			err = PTR_ERR(tuplehash);
			if (err != -EAGAIN)
				goto out;

			continue;
		}
		if (tuplehash->tuple.dir)
			continue;

		flow = container_of(tuplehash, struct flow_offload, tuplehash[0]);

		iter(flow, data);
	}
out:
	rhashtable_walk_stop(&hti);
	rhashtable_walk_exit(&hti);

	return err;
}
EXPORT_SYMBOL_GPL(nf_flow_table_iterate);

static inline bool nf_flow_has_expired(const struct flow_offload *flow)
{
	return (__s32)(flow->timeout - (u32)jiffies) <= 0;
}

static int nf_flow_offload_gc_step(struct nf_flowtable *flow_table)
{
	struct flow_offload_tuple_rhash *tuplehash;
	struct rhashtable_iter hti;
	struct flow_offload *flow;
	int err;

	err = rhashtable_walk_init(&flow_table->rhashtable, &hti, GFP_KERNEL);
	if (err)
		return 0;

	rhashtable_walk_start(&hti);

	while ((tuplehash = rhashtable_walk_next(&hti))) {
		if (IS_ERR(tuplehash)) {
			err = PTR_ERR(tuplehash);
			if (err != -EAGAIN)
				goto out;

			continue;
		}
		if (tuplehash->tuple.dir)
			continue;

		flow = container_of(tuplehash, struct flow_offload, tuplehash[0]);

		if (nf_flow_has_expired(flow) ||
		    (flow->flags & (FLOW_OFFLOAD_DYING |
				    FLOW_OFFLOAD_TEARDOWN)))
			flow_offload_del(flow_table, flow);
	}
out:
	rhashtable_walk_stop(&hti);
	rhashtable_walk_exit(&hti);

	return 1;
}

static void nf_flow_offload_work_gc(struct work_struct *work)
{
	struct nf_flowtable *flow_table;

	flow_table = container_of(work, struct nf_flowtable, gc_work.work);
	nf_flow_offload_gc_step(flow_table);
	queue_delayed_work(system_power_efficient_wq, &flow_table->gc_work, HZ);
}

static int nf_flow_nat_port_tcp(struct sk_buff *skb, unsigned int thoff,
				__be16 port, __be16 new_port)
{
	struct tcphdr *tcph;

	if (!pskb_may_pull(skb, thoff + sizeof(*tcph)) ||
	    skb_try_make_writable(skb, thoff + sizeof(*tcph)))
		return -1;

	tcph = (void *)(skb_network_header(skb) + thoff);
	inet_proto_csum_replace2(&tcph->check, skb, port, new_port, true);

	return 0;
}

static int nf_flow_nat_port_udp(struct sk_buff *skb, unsigned int thoff,
				__be16 port, __be16 new_port)
{
	struct udphdr *udph;

	if (!pskb_may_pull(skb, thoff + sizeof(*udph)) ||
	    skb_try_make_writable(skb, thoff + sizeof(*udph)))
		return -1;

	udph = (void *)(skb_network_header(skb) + thoff);
	if (udph->check || skb->ip_summed == CHECKSUM_PARTIAL) {
		inet_proto_csum_replace2(&udph->check, skb, port,
					 new_port, true);
		if (!udph->check)
			udph->check = CSUM_MANGLED_0;
	}

	return 0;
}

static int nf_flow_nat_port(struct sk_buff *skb, unsigned int thoff,
			    u8 protocol, __be16 port, __be16 new_port)
{
	switch (protocol) {
	case IPPROTO_TCP:
		if (nf_flow_nat_port_tcp(skb, thoff, port, new_port) < 0)
			return NF_DROP;
		break;
	case IPPROTO_UDP:
		if (nf_flow_nat_port_udp(skb, thoff, port, new_port) < 0)
			return NF_DROP;
		break;
	}

	return 0;
}

int nf_flow_snat_port(const struct flow_offload *flow,
		      struct sk_buff *skb, unsigned int thoff,
		      u8 protocol, enum flow_offload_tuple_dir dir)
{
	struct flow_ports *hdr;
	__be16 port, new_port;

	if (!pskb_may_pull(skb, thoff + sizeof(*hdr)) ||
	    skb_try_make_writable(skb, thoff + sizeof(*hdr)))
		return -1;

	hdr = (void *)(skb_network_header(skb) + thoff);

	switch (dir) {
	case FLOW_OFFLOAD_DIR_ORIGINAL:
		port = hdr->source;
		new_port = flow->tuplehash[FLOW_OFFLOAD_DIR_REPLY].tuple.dst_port;
		hdr->source = new_port;
		break;
	case FLOW_OFFLOAD_DIR_REPLY:
		port = hdr->dest;
		new_port = flow->tuplehash[FLOW_OFFLOAD_DIR_ORIGINAL].tuple.src_port;
		hdr->dest = new_port;
		break;
	default:
		return -1;
	}

	return nf_flow_nat_port(skb, thoff, protocol, port, new_port);
}
EXPORT_SYMBOL_GPL(nf_flow_snat_port);

int nf_flow_dnat_port(const struct flow_offload *flow,
		      struct sk_buff *skb, unsigned int thoff,
		      u8 protocol, enum flow_offload_tuple_dir dir)
{
	struct flow_ports *hdr;
	__be16 port, new_port;

	if (!pskb_may_pull(skb, thoff + sizeof(*hdr)) ||
	    skb_try_make_writable(skb, thoff + sizeof(*hdr)))
		return -1;

	hdr = (void *)(skb_network_header(skb) + thoff);

	switch (dir) {
	case FLOW_OFFLOAD_DIR_ORIGINAL:
		port = hdr->dest;
		new_port = flow->tuplehash[FLOW_OFFLOAD_DIR_REPLY].tuple.src_port;
		hdr->dest = new_port;
		break;
	case FLOW_OFFLOAD_DIR_REPLY:
		port = hdr->source;
		new_port = flow->tuplehash[FLOW_OFFLOAD_DIR_ORIGINAL].tuple.dst_port;
		hdr->source = new_port;
		break;
	default:
		return -1;
	}

	return nf_flow_nat_port(skb, thoff, protocol, port, new_port);
}
EXPORT_SYMBOL_GPL(nf_flow_dnat_port);

int nf_flow_table_init(struct nf_flowtable *flowtable)
{
	int err;

	INIT_DEFERRABLE_WORK(&flowtable->gc_work, nf_flow_offload_work_gc);

	err = rhashtable_init(&flowtable->rhashtable,
			      &nf_flow_offload_rhash_params);
	if (err < 0)
		return err;

	queue_delayed_work(system_power_efficient_wq,
			   &flowtable->gc_work, HZ);

	mutex_lock(&flowtable_lock);
	list_add(&flowtable->list, &flowtables);
	mutex_unlock(&flowtable_lock);

	return 0;
}
EXPORT_SYMBOL_GPL(nf_flow_table_init);

static void nf_flow_table_do_cleanup(struct flow_offload *flow, void *data)
{
	struct net_device *dev = data;

	if (!dev) {
		flow_offload_teardown(flow);
		return;
	}

	if (flow->tuplehash[0].tuple.iifidx == dev->ifindex ||
	    flow->tuplehash[1].tuple.iifidx == dev->ifindex)
		flow_offload_dead(flow);
}

static void nf_flow_table_iterate_cleanup(struct nf_flowtable *flowtable,
					  struct net_device *dev)
{
	nf_flow_table_iterate(flowtable, nf_flow_table_do_cleanup, dev);
	flush_delayed_work(&flowtable->gc_work);
}

void nf_flow_table_cleanup(struct net *net, struct net_device *dev)
{
	struct nf_flowtable *flowtable;

	mutex_lock(&flowtable_lock);
	list_for_each_entry(flowtable, &flowtables, list)
		nf_flow_table_iterate_cleanup(flowtable, dev);
	mutex_unlock(&flowtable_lock);
}
EXPORT_SYMBOL_GPL(nf_flow_table_cleanup);

void nf_flow_table_free(struct nf_flowtable *flow_table)
{
	mutex_lock(&flowtable_lock);
	list_del(&flow_table->list);
	mutex_unlock(&flowtable_lock);
	cancel_delayed_work_sync(&flow_table->gc_work);
	nf_flow_table_iterate(flow_table, nf_flow_table_do_cleanup, NULL);
	WARN_ON(!nf_flow_offload_gc_step(flow_table));
	rhashtable_destroy(&flow_table->rhashtable);
}
EXPORT_SYMBOL_GPL(nf_flow_table_free);

MODULE_LICENSE("GPL");
MODULE_AUTHOR("Pablo Neira Ayuso <pablo@netfilter.org>");
Commit	Line	Data
ac2a6666 PNA	1	#include <linux/kernel.h>
	2	#include <linux/init.h>
	3	#include <linux/module.h>
	4	#include <linux/netfilter.h>
	5	#include <linux/rhashtable.h>
	6	#include <linux/netdevice.h>
4f3780c0 FF	7	#include <net/ip.h>
4f3780c0 FF	8	#include <net/ip6_route.h>
c0ea1bcb	9	#include <net/netfilter/nf_tables.h>
ac2a6666 PNA	10	#include <net/netfilter/nf_flow_table.h>
	11	#include <net/netfilter/nf_conntrack.h>
	12	#include <net/netfilter/nf_conntrack_core.h>
	13	#include <net/netfilter/nf_conntrack_tuple.h>
	14
	15	struct flow_offload_entry {
	16	struct flow_offload flow;
	17	struct nf_conn *ct;
	18	struct rcu_head rcu_head;
	19	};
	20
84453a90 FF	21	static DEFINE_MUTEX(flowtable_lock);
	22	static LIST_HEAD(flowtables);
	23
047b300e FF	24	static void
	25	flow_offload_fill_dir(struct flow_offload flow, struct nf_conn ct,
	26	struct nf_flow_route *route,
	27	enum flow_offload_tuple_dir dir)
	28	{
	29	struct flow_offload_tuple *ft = &flow->tuplehash[dir].tuple;
	30	struct nf_conntrack_tuple *ctt = &ct->tuplehash[dir].tuple;
4f3780c0	31	struct dst_entry *dst = route->tuple[dir].dst;
047b300e FF	32
	33	ft->dir = dir;
	34
	35	switch (ctt->src.l3num) {
	36	case NFPROTO_IPV4:
	37	ft->src_v4 = ctt->src.u3.in;
	38	ft->dst_v4 = ctt->dst.u3.in;
4f3780c0	39	ft->mtu = ip_dst_mtu_maybe_forward(dst, true);
047b300e FF	40	break;
	41	case NFPROTO_IPV6:
	42	ft->src_v6 = ctt->src.u3.in6;
	43	ft->dst_v6 = ctt->dst.u3.in6;
4f3780c0	44	ft->mtu = ip6_dst_mtu_forward(dst);
047b300e FF	45	break;
	46	}
	47
	48	ft->l3proto = ctt->src.l3num;
	49	ft->l4proto = ctt->dst.protonum;
	50	ft->src_port = ctt->src.u.tcp.port;
	51	ft->dst_port = ctt->dst.u.tcp.port;
	52
	53	ft->iifidx = route->tuple[dir].ifindex;
	54	ft->oifidx = route->tuple[!dir].ifindex;
4f3780c0	55	ft->dst_cache = dst;
047b300e FF	56	}
047b300e FF	57
ac2a6666 PNA	58	struct flow_offload *
	59	flow_offload_alloc(struct nf_conn ct, struct nf_flow_route route)
	60	{
	61	struct flow_offload_entry *entry;
	62	struct flow_offload *flow;
	63
	64	if (unlikely(nf_ct_is_dying(ct) \|\|
	65	!atomic_inc_not_zero(&ct->ct_general.use)))
	66	return NULL;
	67
	68	entry = kzalloc(sizeof(*entry), GFP_ATOMIC);
	69	if (!entry)
	70	goto err_ct_refcnt;
	71
	72	flow = &entry->flow;
	73
	74	if (!dst_hold_safe(route->tuple[FLOW_OFFLOAD_DIR_ORIGINAL].dst))
	75	goto err_dst_cache_original;
	76
	77	if (!dst_hold_safe(route->tuple[FLOW_OFFLOAD_DIR_REPLY].dst))
	78	goto err_dst_cache_reply;
	79
	80	entry->ct = ct;
	81
047b300e FF	82	flow_offload_fill_dir(flow, ct, route, FLOW_OFFLOAD_DIR_ORIGINAL);
047b300e FF	83	flow_offload_fill_dir(flow, ct, route, FLOW_OFFLOAD_DIR_REPLY);
ac2a6666 PNA	84
	85	if (ct->status & IPS_SRC_NAT)
	86	flow->flags \|= FLOW_OFFLOAD_SNAT;
	87	else if (ct->status & IPS_DST_NAT)
	88	flow->flags \|= FLOW_OFFLOAD_DNAT;
	89
	90	return flow;
	91
	92	err_dst_cache_reply:
	93	dst_release(route->tuple[FLOW_OFFLOAD_DIR_ORIGINAL].dst);
	94	err_dst_cache_original:
	95	kfree(entry);
	96	err_ct_refcnt:
	97	nf_ct_put(ct);
	98
	99	return NULL;
	100	}
	101	EXPORT_SYMBOL_GPL(flow_offload_alloc);
	102
	103	void flow_offload_free(struct flow_offload *flow)
	104	{
	105	struct flow_offload_entry *e;
	106
	107	dst_release(flow->tuplehash[FLOW_OFFLOAD_DIR_ORIGINAL].tuple.dst_cache);
	108	dst_release(flow->tuplehash[FLOW_OFFLOAD_DIR_REPLY].tuple.dst_cache);
	109	e = container_of(flow, struct flow_offload_entry, flow);
0ff90b6c FF	110	nf_ct_delete(e->ct, 0, 0);
	111	nf_ct_put(e->ct);
	112	kfree_rcu(e, rcu_head);
ac2a6666 PNA	113	}
	114	EXPORT_SYMBOL_GPL(flow_offload_free);
	115
a268de77 FF	116	static u32 flow_offload_hash(const void *data, u32 len, u32 seed)
	117	{
	118	const struct flow_offload_tuple *tuple = data;
	119
	120	return jhash(tuple, offsetof(struct flow_offload_tuple, dir), seed);
	121	}
	122
	123	static u32 flow_offload_hash_obj(const void *data, u32 len, u32 seed)
	124	{
	125	const struct flow_offload_tuple_rhash *tuplehash = data;
	126
	127	return jhash(&tuplehash->tuple, offsetof(struct flow_offload_tuple, dir), seed);
	128	}
	129
	130	static int flow_offload_hash_cmp(struct rhashtable_compare_arg *arg,
	131	const void *ptr)
	132	{
	133	const struct flow_offload_tuple *tuple = arg->key;
	134	const struct flow_offload_tuple_rhash *x = ptr;
	135
	136	if (memcmp(&x->tuple, tuple, offsetof(struct flow_offload_tuple, dir)))
	137	return 1;
	138
	139	return 0;
	140	}
	141
	142	static const struct rhashtable_params nf_flow_offload_rhash_params = {
	143	.head_offset = offsetof(struct flow_offload_tuple_rhash, node),
	144	.hashfn = flow_offload_hash,
	145	.obj_hashfn = flow_offload_hash_obj,
	146	.obj_cmpfn = flow_offload_hash_cmp,
	147	.automatic_shrinking = true,
	148	};
	149
ac2a6666 PNA	150	int flow_offload_add(struct nf_flowtable flow_table, struct flow_offload flow)
	151	{
	152	flow->timeout = (u32)jiffies;
	153
	154	rhashtable_insert_fast(&flow_table->rhashtable,
	155	&flow->tuplehash[FLOW_OFFLOAD_DIR_ORIGINAL].node,
a268de77	156	nf_flow_offload_rhash_params);
ac2a6666 PNA	157	rhashtable_insert_fast(&flow_table->rhashtable,
ac2a6666 PNA	158	&flow->tuplehash[FLOW_OFFLOAD_DIR_REPLY].node,
a268de77	159	nf_flow_offload_rhash_params);
ac2a6666 PNA	160	return 0;
	161	}
	162	EXPORT_SYMBOL_GPL(flow_offload_add);
	163
0ff90b6c FF	164	static void flow_offload_del(struct nf_flowtable *flow_table,
0ff90b6c FF	165	struct flow_offload *flow)
ac2a6666	166	{
ac2a6666 PNA	167	rhashtable_remove_fast(&flow_table->rhashtable,
ac2a6666 PNA	168	&flow->tuplehash[FLOW_OFFLOAD_DIR_ORIGINAL].node,
a268de77	169	nf_flow_offload_rhash_params);
ac2a6666 PNA	170	rhashtable_remove_fast(&flow_table->rhashtable,
ac2a6666 PNA	171	&flow->tuplehash[FLOW_OFFLOAD_DIR_REPLY].node,
a268de77	172	nf_flow_offload_rhash_params);
ac2a6666	173
0ff90b6c	174	flow_offload_free(flow);
ac2a6666	175	}
ac2a6666	176
59c466dd FF	177	void flow_offload_teardown(struct flow_offload *flow)
	178	{
	179	flow->flags \|= FLOW_OFFLOAD_TEARDOWN;
	180	}
	181	EXPORT_SYMBOL_GPL(flow_offload_teardown);
	182
ac2a6666 PNA	183	struct flow_offload_tuple_rhash *
	184	flow_offload_lookup(struct nf_flowtable *flow_table,
	185	struct flow_offload_tuple *tuple)
	186	{
	187	return rhashtable_lookup_fast(&flow_table->rhashtable, tuple,
a268de77	188	nf_flow_offload_rhash_params);
ac2a6666 PNA	189	}
	190	EXPORT_SYMBOL_GPL(flow_offload_lookup);
	191
ac2a6666 PNA	192	int nf_flow_table_iterate(struct nf_flowtable *flow_table,
	193	void (iter)(struct flow_offload flow, void *data),
	194	void *data)
	195	{
	196	struct flow_offload_tuple_rhash *tuplehash;
	197	struct rhashtable_iter hti;
	198	struct flow_offload *flow;
	199	int err;
	200
	201	err = rhashtable_walk_init(&flow_table->rhashtable, &hti, GFP_KERNEL);
	202	if (err)
	203	return err;
	204
	205	rhashtable_walk_start(&hti);
	206
	207	while ((tuplehash = rhashtable_walk_next(&hti))) {
	208	if (IS_ERR(tuplehash)) {
	209	err = PTR_ERR(tuplehash);
	210	if (err != -EAGAIN)
	211	goto out;
	212
	213	continue;
	214	}
	215	if (tuplehash->tuple.dir)
	216	continue;
	217
	218	flow = container_of(tuplehash, struct flow_offload, tuplehash[0]);
	219
	220	iter(flow, data);
	221	}
	222	out:
	223	rhashtable_walk_stop(&hti);
	224	rhashtable_walk_exit(&hti);
	225
	226	return err;
	227	}
	228	EXPORT_SYMBOL_GPL(nf_flow_table_iterate);
	229
	230	static inline bool nf_flow_has_expired(const struct flow_offload *flow)
	231	{
	232	return (__s32)(flow->timeout - (u32)jiffies) <= 0;
	233	}
	234
b408c5b0	235	static int nf_flow_offload_gc_step(struct nf_flowtable *flow_table)
ac2a6666 PNA	236	{
ac2a6666 PNA	237	struct flow_offload_tuple_rhash *tuplehash;
ac2a6666 PNA	238	struct rhashtable_iter hti;
	239	struct flow_offload *flow;
	240	int err;
	241
ac2a6666 PNA	242	err = rhashtable_walk_init(&flow_table->rhashtable, &hti, GFP_KERNEL);
ac2a6666 PNA	243	if (err)
b408c5b0	244	return 0;
ac2a6666 PNA	245
	246	rhashtable_walk_start(&hti);
	247
	248	while ((tuplehash = rhashtable_walk_next(&hti))) {
	249	if (IS_ERR(tuplehash)) {
	250	err = PTR_ERR(tuplehash);
	251	if (err != -EAGAIN)
	252	goto out;
	253
	254	continue;
	255	}
	256	if (tuplehash->tuple.dir)
	257	continue;
	258
	259	flow = container_of(tuplehash, struct flow_offload, tuplehash[0]);
	260
	261	if (nf_flow_has_expired(flow) \|\|
59c466dd FF	262	(flow->flags & (FLOW_OFFLOAD_DYING \|
59c466dd FF	263	FLOW_OFFLOAD_TEARDOWN)))
ac2a6666	264	flow_offload_del(flow_table, flow);
ac2a6666 PNA	265	}
	266	out:
	267	rhashtable_walk_stop(&hti);
	268	rhashtable_walk_exit(&hti);
b408c5b0 PNA	269
	270	return 1;
	271	}
	272
a268de77	273	static void nf_flow_offload_work_gc(struct work_struct *work)
b408c5b0 PNA	274	{
	275	struct nf_flowtable *flow_table;
	276
	277	flow_table = container_of(work, struct nf_flowtable, gc_work.work);
	278	nf_flow_offload_gc_step(flow_table);
ac2a6666 PNA	279	queue_delayed_work(system_power_efficient_wq, &flow_table->gc_work, HZ);
ac2a6666 PNA	280	}
ac2a6666 PNA	281
	282	static int nf_flow_nat_port_tcp(struct sk_buff *skb, unsigned int thoff,
	283	__be16 port, __be16 new_port)
	284	{
	285	struct tcphdr *tcph;
	286
	287	if (!pskb_may_pull(skb, thoff + sizeof(*tcph)) \|\|
	288	skb_try_make_writable(skb, thoff + sizeof(*tcph)))
	289	return -1;
	290
	291	tcph = (void *)(skb_network_header(skb) + thoff);
	292	inet_proto_csum_replace2(&tcph->check, skb, port, new_port, true);
	293
	294	return 0;
	295	}
	296
	297	static int nf_flow_nat_port_udp(struct sk_buff *skb, unsigned int thoff,
	298	__be16 port, __be16 new_port)
	299	{
	300	struct udphdr *udph;
	301
	302	if (!pskb_may_pull(skb, thoff + sizeof(*udph)) \|\|
	303	skb_try_make_writable(skb, thoff + sizeof(*udph)))
	304	return -1;
	305
	306	udph = (void *)(skb_network_header(skb) + thoff);
	307	if (udph->check \|\| skb->ip_summed == CHECKSUM_PARTIAL) {
	308	inet_proto_csum_replace2(&udph->check, skb, port,
	309	new_port, true);
	310	if (!udph->check)
	311	udph->check = CSUM_MANGLED_0;
	312	}
	313
	314	return 0;
	315	}
	316
	317	static int nf_flow_nat_port(struct sk_buff *skb, unsigned int thoff,
	318	u8 protocol, __be16 port, __be16 new_port)
	319	{
	320	switch (protocol) {
	321	case IPPROTO_TCP:
	322	if (nf_flow_nat_port_tcp(skb, thoff, port, new_port) < 0)
	323	return NF_DROP;
	324	break;
	325	case IPPROTO_UDP:
	326	if (nf_flow_nat_port_udp(skb, thoff, port, new_port) < 0)
	327	return NF_DROP;
	328	break;
	329	}
	330
	331	return 0;
	332	}
	333
	334	int nf_flow_snat_port(const struct flow_offload *flow,
	335	struct sk_buff *skb, unsigned int thoff,
	336	u8 protocol, enum flow_offload_tuple_dir dir)
	337	{
	338	struct flow_ports *hdr;
	339	__be16 port, new_port;
	340
	341	if (!pskb_may_pull(skb, thoff + sizeof(*hdr)) \|\|
	342	skb_try_make_writable(skb, thoff + sizeof(*hdr)))
	343	return -1;
	344
345	hdr = (void *)(skb_network_header(skb) + thoff);
346
347	switch (dir) {
348	case FLOW_OFFLOAD_DIR_ORIGINAL:
349	port = hdr->source;
350	new_port = flow->tuplehash[FLOW_OFFLOAD_DIR_REPLY].tuple.dst_port;
351	hdr->source = new_port;
352	break;
353	case FLOW_OFFLOAD_DIR_REPLY:
354	port = hdr->dest;
355	new_port = flow->tuplehash[FLOW_OFFLOAD_DIR_ORIGINAL].tuple.src_port;
356	hdr->dest = new_port;
357	break;
358	default:
359	return -1;
360	}
361
362	return nf_flow_nat_port(skb, thoff, protocol, port, new_port);
363	}
364	EXPORT_SYMBOL_GPL(nf_flow_snat_port);
365
366	int nf_flow_dnat_port(const struct flow_offload *flow,
367	struct sk_buff *skb, unsigned int thoff,
368	u8 protocol, enum flow_offload_tuple_dir dir)
369	{
370	struct flow_ports *hdr;
371	__be16 port, new_port;
372
373	if (!pskb_may_pull(skb, thoff + sizeof(*hdr)) \|\|
374	skb_try_make_writable(skb, thoff + sizeof(*hdr)))
375	return -1;
376
377	hdr = (void *)(skb_network_header(skb) + thoff);
378
379	switch (dir) {
380	case FLOW_OFFLOAD_DIR_ORIGINAL:
381	port = hdr->dest;
382	new_port = flow->tuplehash[FLOW_OFFLOAD_DIR_REPLY].tuple.src_port;
383	hdr->dest = new_port;
384	break;
385	case FLOW_OFFLOAD_DIR_REPLY:
386	port = hdr->source;
387	new_port = flow->tuplehash[FLOW_OFFLOAD_DIR_ORIGINAL].tuple.dst_port;
388	hdr->source = new_port;
389	break;
390	default:
391	return -1;
392	}
393
394	return nf_flow_nat_port(skb, thoff, protocol, port, new_port);
395	}
396	EXPORT_SYMBOL_GPL(nf_flow_dnat_port);
397
a268de77 FF	398	int nf_flow_table_init(struct nf_flowtable *flowtable)
	399	{
	400	int err;
	401
	402	INIT_DEFERRABLE_WORK(&flowtable->gc_work, nf_flow_offload_work_gc);
	403
	404	err = rhashtable_init(&flowtable->rhashtable,
	405	&nf_flow_offload_rhash_params);
	406	if (err < 0)
	407	return err;
	408
	409	queue_delayed_work(system_power_efficient_wq,
	410	&flowtable->gc_work, HZ);
	411
84453a90 FF	412	mutex_lock(&flowtable_lock);
	413	list_add(&flowtable->list, &flowtables);
	414	mutex_unlock(&flowtable_lock);
	415
a268de77 FF	416	return 0;
	417	}
	418	EXPORT_SYMBOL_GPL(nf_flow_table_init);
	419
c0ea1bcb PNA	420	static void nf_flow_table_do_cleanup(struct flow_offload flow, void data)
	421	{
	422	struct net_device *dev = data;
	423
59c466dd FF	424	if (!dev) {
59c466dd FF	425	flow_offload_teardown(flow);
c0ea1bcb	426	return;
59c466dd	427	}
c0ea1bcb	428
59c466dd FF	429	if (flow->tuplehash[0].tuple.iifidx == dev->ifindex \|\|
	430	flow->tuplehash[1].tuple.iifidx == dev->ifindex)
	431	flow_offload_dead(flow);
c0ea1bcb PNA	432	}
	433
	434	static void nf_flow_table_iterate_cleanup(struct nf_flowtable *flowtable,
84453a90	435	struct net_device *dev)
c0ea1bcb	436	{
84453a90	437	nf_flow_table_iterate(flowtable, nf_flow_table_do_cleanup, dev);
c0ea1bcb PNA	438	flush_delayed_work(&flowtable->gc_work);
	439	}
	440
	441	void nf_flow_table_cleanup(struct net net, struct net_device dev)
	442	{
84453a90 FF	443	struct nf_flowtable *flowtable;
	444
	445	mutex_lock(&flowtable_lock);
	446	list_for_each_entry(flowtable, &flowtables, list)
	447	nf_flow_table_iterate_cleanup(flowtable, dev);
	448	mutex_unlock(&flowtable_lock);
c0ea1bcb PNA	449	}
	450	EXPORT_SYMBOL_GPL(nf_flow_table_cleanup);
	451
b408c5b0 PNA	452	void nf_flow_table_free(struct nf_flowtable *flow_table)
b408c5b0 PNA	453	{
84453a90 FF	454	mutex_lock(&flowtable_lock);
	455	list_del(&flow_table->list);
	456	mutex_unlock(&flowtable_lock);
a268de77	457	cancel_delayed_work_sync(&flow_table->gc_work);
b408c5b0 PNA	458	nf_flow_table_iterate(flow_table, nf_flow_table_do_cleanup, NULL);
b408c5b0 PNA	459	WARN_ON(!nf_flow_offload_gc_step(flow_table));
a268de77	460	rhashtable_destroy(&flow_table->rhashtable);
b408c5b0 PNA	461	}
	462	EXPORT_SYMBOL_GPL(nf_flow_table_free);
	463
ac2a6666 PNA	464	MODULE_LICENSE("GPL");
ac2a6666 PNA	465	MODULE_AUTHOR("Pablo Neira Ayuso <pablo@netfilter.org>");