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

// SPDX-License-Identifier: GPL-2.0-only
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/init.h>
#include <linux/netlink.h>
#include <linux/netfilter.h>
#include <linux/workqueue.h>
#include <linux/spinlock.h>
#include <linux/netfilter/nf_conntrack_common.h>
#include <linux/netfilter/nf_tables.h>
#include <net/ip.h> /* for ipv4 options. */
#include <net/netfilter/nf_tables.h>
#include <net/netfilter/nf_tables_core.h>
#include <net/netfilter/nf_conntrack_core.h>
#include <net/netfilter/nf_conntrack_extend.h>
#include <net/netfilter/nf_flow_table.h>

struct nft_flow_offload {
	struct nft_flowtable	*flowtable;
};

static enum flow_offload_xmit_type nft_xmit_type(struct dst_entry *dst)
{
	if (dst_xfrm(dst))
		return FLOW_OFFLOAD_XMIT_XFRM;

	return FLOW_OFFLOAD_XMIT_NEIGH;
}

static void nft_default_forward_path(struct nf_flow_route *route,
				     struct dst_entry *dst_cache,
				     enum ip_conntrack_dir dir)
{
	route->tuple[!dir].in.ifindex	= dst_cache->dev->ifindex;
	route->tuple[dir].dst		= dst_cache;
	route->tuple[dir].xmit_type	= nft_xmit_type(dst_cache);
}

static int nft_dev_fill_forward_path(const struct nf_flow_route *route,
				     const struct dst_entry *dst_cache,
				     const struct nf_conn *ct,
				     enum ip_conntrack_dir dir, u8 *ha,
				     struct net_device_path_stack *stack)
{
	const void *daddr = &ct->tuplehash[!dir].tuple.src.u3;
	struct net_device *dev = dst_cache->dev;
	struct neighbour *n;
	u8 nud_state;

	n = dst_neigh_lookup(dst_cache, daddr);
	if (!n)
		return -1;

	read_lock_bh(&n->lock);
	nud_state = n->nud_state;
	ether_addr_copy(ha, n->ha);
	read_unlock_bh(&n->lock);
	neigh_release(n);

	if (!(nud_state & NUD_VALID))
		return -1;

	return dev_fill_forward_path(dev, ha, stack);
}

struct nft_forward_info {
	const struct net_device *indev;
	const struct net_device *outdev;
	struct id {
		__u16	id;
		__be16	proto;
	} encap[NF_FLOW_TABLE_ENCAP_MAX];
	u8 num_encaps;
	u8 h_source[ETH_ALEN];
	u8 h_dest[ETH_ALEN];
	enum flow_offload_xmit_type xmit_type;
};

static void nft_dev_path_info(const struct net_device_path_stack *stack,
			      struct nft_forward_info *info,
			      unsigned char *ha)
{
	const struct net_device_path *path;
	int i;

	memcpy(info->h_dest, ha, ETH_ALEN);

	for (i = 0; i < stack->num_paths; i++) {
		path = &stack->path[i];
		switch (path->type) {
		case DEV_PATH_ETHERNET:
		case DEV_PATH_VLAN:
			info->indev = path->dev;
			if (is_zero_ether_addr(info->h_source))
				memcpy(info->h_source, path->dev->dev_addr, ETH_ALEN);

			if (path->type == DEV_PATH_ETHERNET)
				break;

			/* DEV_PATH_VLAN */
			if (info->num_encaps >= NF_FLOW_TABLE_ENCAP_MAX) {
				info->indev = NULL;
				break;
			}
			info->outdev = path->dev;
			info->encap[info->num_encaps].id = path->encap.id;
			info->encap[info->num_encaps].proto = path->encap.proto;
			info->num_encaps++;
			break;
		case DEV_PATH_BRIDGE:
			if (is_zero_ether_addr(info->h_source))
				memcpy(info->h_source, path->dev->dev_addr, ETH_ALEN);

			info->xmit_type = FLOW_OFFLOAD_XMIT_DIRECT;
			break;
		default:
			info->indev = NULL;
			break;
		}
	}
	if (!info->outdev)
		info->outdev = info->indev;
}

static bool nft_flowtable_find_dev(const struct net_device *dev,
				   struct nft_flowtable *ft)
{
	struct nft_hook *hook;
	bool found = false;

	list_for_each_entry_rcu(hook, &ft->hook_list, list) {
		if (hook->ops.dev != dev)
			continue;

		found = true;
		break;
	}

	return found;
}

static void nft_dev_forward_path(struct nf_flow_route *route,
				 const struct nf_conn *ct,
				 enum ip_conntrack_dir dir,
				 struct nft_flowtable *ft)
{
	const struct dst_entry *dst = route->tuple[dir].dst;
	struct net_device_path_stack stack;
	struct nft_forward_info info = {};
	unsigned char ha[ETH_ALEN];
	int i;

	if (nft_dev_fill_forward_path(route, dst, ct, dir, ha, &stack) >= 0)
		nft_dev_path_info(&stack, &info, ha);

	if (!info.indev || !nft_flowtable_find_dev(info.indev, ft))
		return;

	route->tuple[!dir].in.ifindex = info.indev->ifindex;
	for (i = 0; i < info.num_encaps; i++) {
		route->tuple[!dir].in.encap[i].id = info.encap[i].id;
		route->tuple[!dir].in.encap[i].proto = info.encap[i].proto;
	}
	route->tuple[!dir].in.num_encaps = info.num_encaps;

	if (info.xmit_type == FLOW_OFFLOAD_XMIT_DIRECT) {
		memcpy(route->tuple[dir].out.h_source, info.h_source, ETH_ALEN);
		memcpy(route->tuple[dir].out.h_dest, info.h_dest, ETH_ALEN);
		route->tuple[dir].out.ifindex = info.outdev->ifindex;
		route->tuple[dir].xmit_type = info.xmit_type;
	}
}

static int nft_flow_route(const struct nft_pktinfo *pkt,
			  const struct nf_conn *ct,
			  struct nf_flow_route *route,
			  enum ip_conntrack_dir dir,
			  struct nft_flowtable *ft)
{
	struct dst_entry *this_dst = skb_dst(pkt->skb);
	struct dst_entry *other_dst = NULL;
	struct flowi fl;

	memset(&fl, 0, sizeof(fl));
	switch (nft_pf(pkt)) {
	case NFPROTO_IPV4:
		fl.u.ip4.daddr = ct->tuplehash[dir].tuple.src.u3.ip;
		fl.u.ip4.flowi4_oif = nft_in(pkt)->ifindex;
		break;
	case NFPROTO_IPV6:
		fl.u.ip6.daddr = ct->tuplehash[dir].tuple.src.u3.in6;
		fl.u.ip6.flowi6_oif = nft_in(pkt)->ifindex;
		break;
	}

	nf_route(nft_net(pkt), &other_dst, &fl, false, nft_pf(pkt));
	if (!other_dst)
		return -ENOENT;

	nft_default_forward_path(route, this_dst, dir);
	nft_default_forward_path(route, other_dst, !dir);

	if (route->tuple[dir].xmit_type	== FLOW_OFFLOAD_XMIT_NEIGH &&
	    route->tuple[!dir].xmit_type == FLOW_OFFLOAD_XMIT_NEIGH) {
		nft_dev_forward_path(route, ct, dir, ft);
		nft_dev_forward_path(route, ct, !dir, ft);
	}

	return 0;
}

static bool nft_flow_offload_skip(struct sk_buff *skb, int family)
{
	if (skb_sec_path(skb))
		return true;

	if (family == NFPROTO_IPV4) {
		const struct ip_options *opt;

		opt = &(IPCB(skb)->opt);

		if (unlikely(opt->optlen))
			return true;
	}

	return false;
}

static void nft_flow_offload_eval(const struct nft_expr *expr,
				  struct nft_regs *regs,
				  const struct nft_pktinfo *pkt)
{
	struct nft_flow_offload *priv = nft_expr_priv(expr);
	struct nf_flowtable *flowtable = &priv->flowtable->data;
	struct tcphdr _tcph, *tcph = NULL;
	struct nf_flow_route route = {};
	enum ip_conntrack_info ctinfo;
	struct flow_offload *flow;
	enum ip_conntrack_dir dir;
	struct nf_conn *ct;
	int ret;

	if (nft_flow_offload_skip(pkt->skb, nft_pf(pkt)))
		goto out;

	ct = nf_ct_get(pkt->skb, &ctinfo);
	if (!ct)
		goto out;

	switch (ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.protonum) {
	case IPPROTO_TCP:
		tcph = skb_header_pointer(pkt->skb, pkt->xt.thoff,
					  sizeof(_tcph), &_tcph);
		if (unlikely(!tcph || tcph->fin || tcph->rst))
			goto out;
		break;
	case IPPROTO_UDP:
		break;
	default:
		goto out;
	}

	if (nf_ct_ext_exist(ct, NF_CT_EXT_HELPER) ||
	    ct->status & (IPS_SEQ_ADJUST | IPS_NAT_CLASH))
		goto out;

	if (!nf_ct_is_confirmed(ct))
		goto out;

	if (test_and_set_bit(IPS_OFFLOAD_BIT, &ct->status))
		goto out;

	dir = CTINFO2DIR(ctinfo);
	if (nft_flow_route(pkt, ct, &route, dir, priv->flowtable) < 0)
		goto err_flow_route;

	flow = flow_offload_alloc(ct);
	if (!flow)
		goto err_flow_alloc;

	if (flow_offload_route_init(flow, &route) < 0)
		goto err_flow_add;

	if (tcph) {
		ct->proto.tcp.seen[0].flags |= IP_CT_TCP_FLAG_BE_LIBERAL;
		ct->proto.tcp.seen[1].flags |= IP_CT_TCP_FLAG_BE_LIBERAL;
	}

	ret = flow_offload_add(flowtable, flow);
	if (ret < 0)
		goto err_flow_add;

	dst_release(route.tuple[!dir].dst);
	return;

err_flow_add:
	flow_offload_free(flow);
err_flow_alloc:
	dst_release(route.tuple[!dir].dst);
err_flow_route:
	clear_bit(IPS_OFFLOAD_BIT, &ct->status);
out:
	regs->verdict.code = NFT_BREAK;
}

static int nft_flow_offload_validate(const struct nft_ctx *ctx,
				     const struct nft_expr *expr,
				     const struct nft_data **data)
{
	unsigned int hook_mask = (1 << NF_INET_FORWARD);

	return nft_chain_validate_hooks(ctx->chain, hook_mask);
}

static const struct nla_policy nft_flow_offload_policy[NFTA_FLOW_MAX + 1] = {
	[NFTA_FLOW_TABLE_NAME]	= { .type = NLA_STRING,
				    .len = NFT_NAME_MAXLEN - 1 },
};

static int nft_flow_offload_init(const struct nft_ctx *ctx,
				 const struct nft_expr *expr,
				 const struct nlattr * const tb[])
{
	struct nft_flow_offload *priv = nft_expr_priv(expr);
	u8 genmask = nft_genmask_next(ctx->net);
	struct nft_flowtable *flowtable;

	if (!tb[NFTA_FLOW_TABLE_NAME])
		return -EINVAL;

	flowtable = nft_flowtable_lookup(ctx->table, tb[NFTA_FLOW_TABLE_NAME],
					 genmask);
	if (IS_ERR(flowtable))
		return PTR_ERR(flowtable);

	priv->flowtable = flowtable;
	flowtable->use++;

	return nf_ct_netns_get(ctx->net, ctx->family);
}

static void nft_flow_offload_deactivate(const struct nft_ctx *ctx,
					const struct nft_expr *expr,
					enum nft_trans_phase phase)
{
	struct nft_flow_offload *priv = nft_expr_priv(expr);

	nf_tables_deactivate_flowtable(ctx, priv->flowtable, phase);
}

static void nft_flow_offload_activate(const struct nft_ctx *ctx,
				      const struct nft_expr *expr)
{
	struct nft_flow_offload *priv = nft_expr_priv(expr);

	priv->flowtable->use++;
}

static void nft_flow_offload_destroy(const struct nft_ctx *ctx,
				     const struct nft_expr *expr)
{
	nf_ct_netns_put(ctx->net, ctx->family);
}

static int nft_flow_offload_dump(struct sk_buff *skb, const struct nft_expr *expr)
{
	struct nft_flow_offload *priv = nft_expr_priv(expr);

	if (nla_put_string(skb, NFTA_FLOW_TABLE_NAME, priv->flowtable->name))
		goto nla_put_failure;

	return 0;

nla_put_failure:
	return -1;
}

static struct nft_expr_type nft_flow_offload_type;
static const struct nft_expr_ops nft_flow_offload_ops = {
	.type		= &nft_flow_offload_type,
	.size		= NFT_EXPR_SIZE(sizeof(struct nft_flow_offload)),
	.eval		= nft_flow_offload_eval,
	.init		= nft_flow_offload_init,
	.activate	= nft_flow_offload_activate,
	.deactivate	= nft_flow_offload_deactivate,
	.destroy	= nft_flow_offload_destroy,
	.validate	= nft_flow_offload_validate,
	.dump		= nft_flow_offload_dump,
};

static struct nft_expr_type nft_flow_offload_type __read_mostly = {
	.name		= "flow_offload",
	.ops		= &nft_flow_offload_ops,
	.policy		= nft_flow_offload_policy,
	.maxattr	= NFTA_FLOW_MAX,
	.owner		= THIS_MODULE,
};

static int flow_offload_netdev_event(struct notifier_block *this,
				     unsigned long event, void *ptr)
{
	struct net_device *dev = netdev_notifier_info_to_dev(ptr);

	if (event != NETDEV_DOWN)
		return NOTIFY_DONE;

	nf_flow_table_cleanup(dev);

	return NOTIFY_DONE;
}

static struct notifier_block flow_offload_netdev_notifier = {
	.notifier_call	= flow_offload_netdev_event,
};

static int __init nft_flow_offload_module_init(void)
{
	int err;

	err = register_netdevice_notifier(&flow_offload_netdev_notifier);
	if (err)
		goto err;

	err = nft_register_expr(&nft_flow_offload_type);
	if (err < 0)
		goto register_expr;

	return 0;

register_expr:
	unregister_netdevice_notifier(&flow_offload_netdev_notifier);
err:
	return err;
}

static void __exit nft_flow_offload_module_exit(void)
{
	nft_unregister_expr(&nft_flow_offload_type);
	unregister_netdevice_notifier(&flow_offload_netdev_notifier);
}

module_init(nft_flow_offload_module_init);
module_exit(nft_flow_offload_module_exit);

MODULE_LICENSE("GPL");
MODULE_AUTHOR("Pablo Neira Ayuso <pablo@netfilter.org>");
MODULE_ALIAS_NFT_EXPR("flow_offload");
MODULE_DESCRIPTION("nftables hardware flow offload module");
Commit	Line	Data
09c434b8	1	// SPDX-License-Identifier: GPL-2.0-only
a3c90f7a PNA	2	#include <linux/kernel.h>
	3	#include <linux/module.h>
	4	#include <linux/init.h>
	5	#include <linux/netlink.h>
	6	#include <linux/netfilter.h>
	7	#include <linux/workqueue.h>
	8	#include <linux/spinlock.h>
40d102cd	9	#include <linux/netfilter/nf_conntrack_common.h>
a3c90f7a PNA	10	#include <linux/netfilter/nf_tables.h>
	11	#include <net/ip.h> /* for ipv4 options. */
	12	#include <net/netfilter/nf_tables.h>
	13	#include <net/netfilter/nf_tables_core.h>
	14	#include <net/netfilter/nf_conntrack_core.h>
40d102cd	15	#include <net/netfilter/nf_conntrack_extend.h>
a3c90f7a PNA	16	#include <net/netfilter/nf_flow_table.h>
	17
	18	struct nft_flow_offload {
	19	struct nft_flowtable *flowtable;
	20	};
	21
5139c0c0 PNA	22	static enum flow_offload_xmit_type nft_xmit_type(struct dst_entry *dst)
	23	{
	24	if (dst_xfrm(dst))
	25	return FLOW_OFFLOAD_XMIT_XFRM;
	26
	27	return FLOW_OFFLOAD_XMIT_NEIGH;
	28	}
	29
	30	static void nft_default_forward_path(struct nf_flow_route *route,
	31	struct dst_entry *dst_cache,
	32	enum ip_conntrack_dir dir)
	33	{
c63a7cc4	34	route->tuple[!dir].in.ifindex = dst_cache->dev->ifindex;
5139c0c0 PNA	35	route->tuple[dir].dst = dst_cache;
	36	route->tuple[dir].xmit_type = nft_xmit_type(dst_cache);
	37	}
	38
c63a7cc4 PNA	39	static int nft_dev_fill_forward_path(const struct nf_flow_route *route,
	40	const struct dst_entry *dst_cache,
	41	const struct nf_conn *ct,
7a27f6ab	42	enum ip_conntrack_dir dir, u8 *ha,
c63a7cc4 PNA	43	struct net_device_path_stack *stack)
	44	{
	45	const void *daddr = &ct->tuplehash[!dir].tuple.src.u3;
	46	struct net_device *dev = dst_cache->dev;
c63a7cc4 PNA	47	struct neighbour *n;
	48	u8 nud_state;
	49
	50	n = dst_neigh_lookup(dst_cache, daddr);
	51	if (!n)
	52	return -1;
	53
	54	read_lock_bh(&n->lock);
	55	nud_state = n->nud_state;
	56	ether_addr_copy(ha, n->ha);
	57	read_unlock_bh(&n->lock);
	58	neigh_release(n);
	59
	60	if (!(nud_state & NUD_VALID))
	61	return -1;
	62
	63	return dev_fill_forward_path(dev, ha, stack);
	64	}
	65
	66	struct nft_forward_info {
	67	const struct net_device *indev;
7a27f6ab	68	const struct net_device *outdev;
4cd91f7c PNA	69	struct id {
	70	__u16 id;
	71	__be16 proto;
	72	} encap[NF_FLOW_TABLE_ENCAP_MAX];
	73	u8 num_encaps;
7a27f6ab PNA	74	u8 h_source[ETH_ALEN];
	75	u8 h_dest[ETH_ALEN];
	76	enum flow_offload_xmit_type xmit_type;
c63a7cc4 PNA	77	};
	78
	79	static void nft_dev_path_info(const struct net_device_path_stack *stack,
7a27f6ab PNA	80	struct nft_forward_info *info,
7a27f6ab PNA	81	unsigned char *ha)
c63a7cc4 PNA	82	{
	83	const struct net_device_path *path;
	84	int i;
	85
7a27f6ab PNA	86	memcpy(info->h_dest, ha, ETH_ALEN);
7a27f6ab PNA	87
c63a7cc4 PNA	88	for (i = 0; i < stack->num_paths; i++) {
	89	path = &stack->path[i];
	90	switch (path->type) {
	91	case DEV_PATH_ETHERNET:
4cd91f7c	92	case DEV_PATH_VLAN:
c63a7cc4	93	info->indev = path->dev;
7a27f6ab PNA	94	if (is_zero_ether_addr(info->h_source))
7a27f6ab PNA	95	memcpy(info->h_source, path->dev->dev_addr, ETH_ALEN);
4cd91f7c PNA	96
	97	if (path->type == DEV_PATH_ETHERNET)
	98	break;
	99
	100	/* DEV_PATH_VLAN */
	101	if (info->num_encaps >= NF_FLOW_TABLE_ENCAP_MAX) {
	102	info->indev = NULL;
	103	break;
	104	}
	105	info->outdev = path->dev;
	106	info->encap[info->num_encaps].id = path->encap.id;
	107	info->encap[info->num_encaps].proto = path->encap.proto;
	108	info->num_encaps++;
c63a7cc4	109	break;
c63a7cc4	110	case DEV_PATH_BRIDGE:
7a27f6ab PNA	111	if (is_zero_ether_addr(info->h_source))
	112	memcpy(info->h_source, path->dev->dev_addr, ETH_ALEN);
	113
	114	info->xmit_type = FLOW_OFFLOAD_XMIT_DIRECT;
	115	break;
c63a7cc4 PNA	116	default:
	117	info->indev = NULL;
	118	break;
	119	}
	120	}
7a27f6ab PNA	121	if (!info->outdev)
7a27f6ab PNA	122	info->outdev = info->indev;
c63a7cc4 PNA	123	}
	124
	125	static bool nft_flowtable_find_dev(const struct net_device *dev,
	126	struct nft_flowtable *ft)
	127	{
	128	struct nft_hook *hook;
	129	bool found = false;
	130
	131	list_for_each_entry_rcu(hook, &ft->hook_list, list) {
	132	if (hook->ops.dev != dev)
	133	continue;
	134
	135	found = true;
	136	break;
	137	}
	138
	139	return found;
	140	}
	141
	142	static void nft_dev_forward_path(struct nf_flow_route *route,
	143	const struct nf_conn *ct,
	144	enum ip_conntrack_dir dir,
	145	struct nft_flowtable *ft)
	146	{
	147	const struct dst_entry *dst = route->tuple[dir].dst;
	148	struct net_device_path_stack stack;
	149	struct nft_forward_info info = {};
7a27f6ab	150	unsigned char ha[ETH_ALEN];
4cd91f7c	151	int i;
c63a7cc4	152
7a27f6ab PNA	153	if (nft_dev_fill_forward_path(route, dst, ct, dir, ha, &stack) >= 0)
7a27f6ab PNA	154	nft_dev_path_info(&stack, &info, ha);
c63a7cc4 PNA	155
	156	if (!info.indev \|\| !nft_flowtable_find_dev(info.indev, ft))
	157	return;
	158
	159	route->tuple[!dir].in.ifindex = info.indev->ifindex;
4cd91f7c PNA	160	for (i = 0; i < info.num_encaps; i++) {
	161	route->tuple[!dir].in.encap[i].id = info.encap[i].id;
	162	route->tuple[!dir].in.encap[i].proto = info.encap[i].proto;
	163	}
	164	route->tuple[!dir].in.num_encaps = info.num_encaps;
7a27f6ab PNA	165
	166	if (info.xmit_type == FLOW_OFFLOAD_XMIT_DIRECT) {
	167	memcpy(route->tuple[dir].out.h_source, info.h_source, ETH_ALEN);
	168	memcpy(route->tuple[dir].out.h_dest, info.h_dest, ETH_ALEN);
	169	route->tuple[dir].out.ifindex = info.outdev->ifindex;
	170	route->tuple[dir].xmit_type = info.xmit_type;
	171	}
c63a7cc4 PNA	172	}
c63a7cc4 PNA	173
a3c90f7a PNA	174	static int nft_flow_route(const struct nft_pktinfo *pkt,
	175	const struct nf_conn *ct,
	176	struct nf_flow_route *route,
c63a7cc4 PNA	177	enum ip_conntrack_dir dir,
c63a7cc4 PNA	178	struct nft_flowtable *ft)
a3c90f7a PNA	179	{
	180	struct dst_entry *this_dst = skb_dst(pkt->skb);
	181	struct dst_entry *other_dst = NULL;
	182	struct flowi fl;
	183
	184	memset(&fl, 0, sizeof(fl));
	185	switch (nft_pf(pkt)) {
	186	case NFPROTO_IPV4:
a799aea0	187	fl.u.ip4.daddr = ct->tuplehash[dir].tuple.src.u3.ip;
10f4e765	188	fl.u.ip4.flowi4_oif = nft_in(pkt)->ifindex;
a3c90f7a PNA	189	break;
a3c90f7a PNA	190	case NFPROTO_IPV6:
a799aea0	191	fl.u.ip6.daddr = ct->tuplehash[dir].tuple.src.u3.in6;
10f4e765	192	fl.u.ip6.flowi6_oif = nft_in(pkt)->ifindex;
a3c90f7a PNA	193	break;
	194	}
	195
	196	nf_route(nft_net(pkt), &other_dst, &fl, false, nft_pf(pkt));
	197	if (!other_dst)
	198	return -ENOENT;
	199
5139c0c0 PNA	200	nft_default_forward_path(route, this_dst, dir);
5139c0c0 PNA	201	nft_default_forward_path(route, other_dst, !dir);
a3c90f7a	202
c63a7cc4 PNA	203	if (route->tuple[dir].xmit_type == FLOW_OFFLOAD_XMIT_NEIGH &&
	204	route->tuple[!dir].xmit_type == FLOW_OFFLOAD_XMIT_NEIGH) {
	205	nft_dev_forward_path(route, ct, dir, ft);
	206	nft_dev_forward_path(route, ct, !dir, ft);
	207	}
	208
a3c90f7a PNA	209	return 0;
	210	}
	211
69aeb538	212	static bool nft_flow_offload_skip(struct sk_buff *skb, int family)
a3c90f7a	213	{
a3c90f7a PNA	214	if (skb_sec_path(skb))
	215	return true;
	216
69aeb538 FW	217	if (family == NFPROTO_IPV4) {
	218	const struct ip_options *opt;
	219
	220	opt = &(IPCB(skb)->opt);
	221
	222	if (unlikely(opt->optlen))
	223	return true;
	224	}
	225
a3c90f7a PNA	226	return false;
	227	}
	228
	229	static void nft_flow_offload_eval(const struct nft_expr *expr,
	230	struct nft_regs *regs,
	231	const struct nft_pktinfo *pkt)
	232	{
	233	struct nft_flow_offload *priv = nft_expr_priv(expr);
	234	struct nf_flowtable *flowtable = &priv->flowtable->data;
dfe42be1	235	struct tcphdr _tcph, *tcph = NULL;
c63a7cc4	236	struct nf_flow_route route = {};
a3c90f7a	237	enum ip_conntrack_info ctinfo;
a3c90f7a PNA	238	struct flow_offload *flow;
	239	enum ip_conntrack_dir dir;
	240	struct nf_conn *ct;
	241	int ret;
	242
69aeb538	243	if (nft_flow_offload_skip(pkt->skb, nft_pf(pkt)))
a3c90f7a PNA	244	goto out;
	245
	246	ct = nf_ct_get(pkt->skb, &ctinfo);
	247	if (!ct)
	248	goto out;
	249
	250	switch (ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.protonum) {
	251	case IPPROTO_TCP:
dfe42be1 PNA	252	tcph = skb_header_pointer(pkt->skb, pkt->xt.thoff,
	253	sizeof(_tcph), &_tcph);
	254	if (unlikely(!tcph \|\| tcph->fin \|\| tcph->rst))
	255	goto out;
8437a620	256	break;
a3c90f7a PNA	257	case IPPROTO_UDP:
	258	break;
	259	default:
	260	goto out;
	261	}
	262
91a9048f	263	if (nf_ct_ext_exist(ct, NF_CT_EXT_HELPER) \|\|
c4617214	264	ct->status & (IPS_SEQ_ADJUST \| IPS_NAT_CLASH))
a3c90f7a PNA	265	goto out;
a3c90f7a PNA	266
270a8a29	267	if (!nf_ct_is_confirmed(ct))
a3c90f7a PNA	268	goto out;
	269
	270	if (test_and_set_bit(IPS_OFFLOAD_BIT, &ct->status))
	271	goto out;
	272
	273	dir = CTINFO2DIR(ctinfo);
c63a7cc4	274	if (nft_flow_route(pkt, ct, &route, dir, priv->flowtable) < 0)
a3c90f7a PNA	275	goto err_flow_route;
a3c90f7a PNA	276
f1363e05	277	flow = flow_offload_alloc(ct);
a3c90f7a PNA	278	if (!flow)
	279	goto err_flow_alloc;
	280
f1363e05 PNA	281	if (flow_offload_route_init(flow, &route) < 0)
	282	goto err_flow_add;
	283
dfe42be1	284	if (tcph) {
8437a620 FW	285	ct->proto.tcp.seen[0].flags \|= IP_CT_TCP_FLAG_BE_LIBERAL;
	286	ct->proto.tcp.seen[1].flags \|= IP_CT_TCP_FLAG_BE_LIBERAL;
	287	}
	288
a3c90f7a PNA	289	ret = flow_offload_add(flowtable, flow);
	290	if (ret < 0)
	291	goto err_flow_add;
	292
26a302af	293	dst_release(route.tuple[!dir].dst);
a3c90f7a PNA	294	return;
	295
	296	err_flow_add:
	297	flow_offload_free(flow);
	298	err_flow_alloc:
	299	dst_release(route.tuple[!dir].dst);
	300	err_flow_route:
	301	clear_bit(IPS_OFFLOAD_BIT, &ct->status);
	302	out:
	303	regs->verdict.code = NFT_BREAK;
	304	}
	305
	306	static int nft_flow_offload_validate(const struct nft_ctx *ctx,
	307	const struct nft_expr *expr,
	308	const struct nft_data **data)
	309	{
	310	unsigned int hook_mask = (1 << NF_INET_FORWARD);
	311
	312	return nft_chain_validate_hooks(ctx->chain, hook_mask);
	313	}
	314
14c41586 PNA	315	static const struct nla_policy nft_flow_offload_policy[NFTA_FLOW_MAX + 1] = {
	316	[NFTA_FLOW_TABLE_NAME] = { .type = NLA_STRING,
	317	.len = NFT_NAME_MAXLEN - 1 },
	318	};
	319
a3c90f7a PNA	320	static int nft_flow_offload_init(const struct nft_ctx *ctx,
	321	const struct nft_expr *expr,
	322	const struct nlattr * const tb[])
	323	{
	324	struct nft_flow_offload *priv = nft_expr_priv(expr);
	325	u8 genmask = nft_genmask_next(ctx->net);
	326	struct nft_flowtable *flowtable;
	327
	328	if (!tb[NFTA_FLOW_TABLE_NAME])
	329	return -EINVAL;
	330
cac20fcd PNA	331	flowtable = nft_flowtable_lookup(ctx->table, tb[NFTA_FLOW_TABLE_NAME],
cac20fcd PNA	332	genmask);
a3c90f7a PNA	333	if (IS_ERR(flowtable))
	334	return PTR_ERR(flowtable);
	335
	336	priv->flowtable = flowtable;
	337	flowtable->use++;
	338
36596dad	339	return nf_ct_netns_get(ctx->net, ctx->family);
a3c90f7a PNA	340	}
a3c90f7a PNA	341
9b05b6e1 LGL	342	static void nft_flow_offload_deactivate(const struct nft_ctx *ctx,
	343	const struct nft_expr *expr,
	344	enum nft_trans_phase phase)
	345	{
	346	struct nft_flow_offload *priv = nft_expr_priv(expr);
	347
	348	nf_tables_deactivate_flowtable(ctx, priv->flowtable, phase);
	349	}
	350
	351	static void nft_flow_offload_activate(const struct nft_ctx *ctx,
	352	const struct nft_expr *expr)
	353	{
	354	struct nft_flow_offload *priv = nft_expr_priv(expr);
	355
	356	priv->flowtable->use++;
	357	}
	358
a3c90f7a PNA	359	static void nft_flow_offload_destroy(const struct nft_ctx *ctx,
	360	const struct nft_expr *expr)
	361	{
36596dad	362	nf_ct_netns_put(ctx->net, ctx->family);
a3c90f7a PNA	363	}
	364
	365	static int nft_flow_offload_dump(struct sk_buff skb, const struct nft_expr expr)
	366	{
	367	struct nft_flow_offload *priv = nft_expr_priv(expr);
	368
	369	if (nla_put_string(skb, NFTA_FLOW_TABLE_NAME, priv->flowtable->name))
	370	goto nla_put_failure;
	371
	372	return 0;
	373
	374	nla_put_failure:
	375	return -1;
	376	}
	377
	378	static struct nft_expr_type nft_flow_offload_type;
	379	static const struct nft_expr_ops nft_flow_offload_ops = {
	380	.type = &nft_flow_offload_type,
	381	.size = NFT_EXPR_SIZE(sizeof(struct nft_flow_offload)),
	382	.eval = nft_flow_offload_eval,
	383	.init = nft_flow_offload_init,
9b05b6e1 LGL	384	.activate = nft_flow_offload_activate,
9b05b6e1 LGL	385	.deactivate = nft_flow_offload_deactivate,
a3c90f7a PNA	386	.destroy = nft_flow_offload_destroy,
	387	.validate = nft_flow_offload_validate,
	388	.dump = nft_flow_offload_dump,
	389	};
	390
	391	static struct nft_expr_type nft_flow_offload_type __read_mostly = {
	392	.name = "flow_offload",
	393	.ops = &nft_flow_offload_ops,
14c41586	394	.policy = nft_flow_offload_policy,
a3c90f7a PNA	395	.maxattr = NFTA_FLOW_MAX,
	396	.owner = THIS_MODULE,
	397	};
	398
a3c90f7a PNA	399	static int flow_offload_netdev_event(struct notifier_block *this,
	400	unsigned long event, void *ptr)
	401	{
	402	struct net_device *dev = netdev_notifier_info_to_dev(ptr);
	403
	404	if (event != NETDEV_DOWN)
	405	return NOTIFY_DONE;
	406
5f1be84a	407	nf_flow_table_cleanup(dev);
a3c90f7a PNA	408
	409	return NOTIFY_DONE;
	410	}
	411
	412	static struct notifier_block flow_offload_netdev_notifier = {
	413	.notifier_call = flow_offload_netdev_event,
	414	};
	415
	416	static int __init nft_flow_offload_module_init(void)
	417	{
	418	int err;
	419
584eab29 TY	420	err = register_netdevice_notifier(&flow_offload_netdev_notifier);
	421	if (err)
	422	goto err;
a3c90f7a PNA	423
	424	err = nft_register_expr(&nft_flow_offload_type);
	425	if (err < 0)
	426	goto register_expr;
	427
	428	return 0;
	429
	430	register_expr:
	431	unregister_netdevice_notifier(&flow_offload_netdev_notifier);
584eab29	432	err:
a3c90f7a PNA	433	return err;
	434	}
	435
	436	static void __exit nft_flow_offload_module_exit(void)
	437	{
a3c90f7a PNA	438	nft_unregister_expr(&nft_flow_offload_type);
a3c90f7a PNA	439	unregister_netdevice_notifier(&flow_offload_netdev_notifier);
a3c90f7a PNA	440	}
	441
	442	module_init(nft_flow_offload_module_init);
	443	module_exit(nft_flow_offload_module_exit);
	444
	445	MODULE_LICENSE("GPL");
	446	MODULE_AUTHOR("Pablo Neira Ayuso <pablo@netfilter.org>");
	447	MODULE_ALIAS_NFT_EXPR("flow_offload");
4cacc395	448	MODULE_DESCRIPTION("nftables hardware flow offload module");