[thirdparty/linux.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 bool nft_is_valid_ether_device(const struct net_device *dev)
{
	if (!dev || (dev->flags & IFF_LOOPBACK) || dev->type != ARPHRD_ETHER ||
	    dev->addr_len != ETH_ALEN || !is_valid_ether_addr(dev->dev_addr))
		return false;

	return true;
}

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;

	if (!nft_is_valid_ether_device(dev))
		goto out;

	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;

out:
	return dev_fill_forward_path(dev, ha, stack);
}

struct nft_forward_info {
	const struct net_device *indev;
	const struct net_device *outdev;
	const struct net_device *hw_outdev;
	struct id {
		__u16	id;
		__be16	proto;
	} encap[NF_FLOW_TABLE_ENCAP_MAX];
	u8 num_encaps;
	u8 ingress_vlans;
	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, struct nf_flowtable *flowtable)
{
	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_DSA:
		case DEV_PATH_VLAN:
		case DEV_PATH_PPPOE:
			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;
			if (path->type == DEV_PATH_DSA) {
				i = stack->num_paths;
				break;
			}

			/* DEV_PATH_VLAN and DEV_PATH_PPPOE */
			if (info->num_encaps >= NF_FLOW_TABLE_ENCAP_MAX) {
				info->indev = NULL;
				break;
			}
			if (!info->outdev)
				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++;
			if (path->type == DEV_PATH_PPPOE)
				memcpy(info->h_dest, path->encap.h_dest, ETH_ALEN);
			break;
		case DEV_PATH_BRIDGE:
			if (is_zero_ether_addr(info->h_source))
				memcpy(info->h_source, path->dev->dev_addr, ETH_ALEN);

			switch (path->bridge.vlan_mode) {
			case DEV_PATH_BR_VLAN_UNTAG_HW:
				info->ingress_vlans |= BIT(info->num_encaps - 1);
				break;
			case DEV_PATH_BR_VLAN_TAG:
				info->encap[info->num_encaps].id = path->bridge.vlan_id;
				info->encap[info->num_encaps].proto = path->bridge.vlan_proto;
				info->num_encaps++;
				break;
			case DEV_PATH_BR_VLAN_UNTAG:
				info->num_encaps--;
				break;
			case DEV_PATH_BR_VLAN_KEEP:
				break;
			}
			info->xmit_type = FLOW_OFFLOAD_XMIT_DIRECT;
			break;
		default:
			info->indev = NULL;
			break;
		}
	}
	if (!info->outdev)
		info->outdev = info->indev;

	info->hw_outdev = info->indev;

	if (nf_flowtable_hw_offload(flowtable) &&
	    nft_is_valid_ether_device(info->indev))
		info->xmit_type = FLOW_OFFLOAD_XMIT_DIRECT;
}

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, &ft->data);

	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;
	route->tuple[!dir].in.ingress_vlans = info.ingress_vlans;

	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].out.hw_ifindex = info.hw_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.saddr = ct->tuplehash[!dir].tuple.src.u3.ip;
		fl.u.ip4.flowi4_oif = nft_in(pkt)->ifindex;
		fl.u.ip4.flowi4_iif = this_dst->dev->ifindex;
		fl.u.ip4.flowi4_tos = RT_TOS(ip_hdr(pkt->skb)->tos);
		fl.u.ip4.flowi4_mark = pkt->skb->mark;
		fl.u.ip4.flowi4_flags = FLOWI_FLAG_ANYSRC;
		break;
	case NFPROTO_IPV6:
		fl.u.ip6.daddr = ct->tuplehash[dir].tuple.src.u3.in6;
		fl.u.ip6.saddr = ct->tuplehash[!dir].tuple.src.u3.in6;
		fl.u.ip6.flowi6_oif = nft_in(pkt)->ifindex;
		fl.u.ip6.flowi6_iif = this_dst->dev->ifindex;
		fl.u.ip6.flowlabel = ip6_flowinfo(ipv6_hdr(pkt->skb));
		fl.u.ip6.flowi6_mark = pkt->skb->mark;
		fl.u.ip6.flowi6_flags = FLOWI_FLAG_ANYSRC;
		break;
	}

	if (!dst_hold_safe(this_dst))
		return -ENOENT;

	nf_route(nft_net(pkt), &other_dst, &fl, false, nft_pf(pkt));
	if (!other_dst) {
		dst_release(this_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, nft_thoff(pkt),
					  sizeof(_tcph), &_tcph);
		if (unlikely(!tcph || tcph->fin || tcph->rst ||
			     !nf_conntrack_tcp_established(ct)))
			goto out;
		break;
	case IPPROTO_UDP:
		break;
#ifdef CONFIG_NF_CT_PROTO_GRE
	case IPPROTO_GRE: {
		struct nf_conntrack_tuple *tuple;

		if (ct->status & IPS_NAT_MASK)
			goto out;
		tuple = &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple;
		/* No support for GRE v1 */
		if (tuple->src.u.gre.key || tuple->dst.u.gre.key)
			goto out;
		break;
	}
#endif
	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;

	flow_offload_route_init(flow, &route);

	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;

	return;

err_flow_add:
	flow_offload_free(flow);
err_flow_alloc:
	dst_release(route.tuple[dir].dst);
	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);

	if (!nft_use_inc(&flowtable->use))
		return -EMFILE;

	priv->flowtable = flowtable;

	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);

	nft_use_inc_restore(&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, bool reset)
{
	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,
	.reduce		= NFT_REDUCE_READONLY,
};

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
45ca3e61 FF	39	static bool nft_is_valid_ether_device(const struct net_device *dev)
	40	{
	41	if (!dev \|\| (dev->flags & IFF_LOOPBACK) \|\| dev->type != ARPHRD_ETHER \|\|
	42	dev->addr_len != ETH_ALEN \|\| !is_valid_ether_addr(dev->dev_addr))
	43	return false;
	44
	45	return true;
	46	}
	47
c63a7cc4 PNA	48	static int nft_dev_fill_forward_path(const struct nf_flow_route *route,
	49	const struct dst_entry *dst_cache,
	50	const struct nf_conn *ct,
7a27f6ab	51	enum ip_conntrack_dir dir, u8 *ha,
c63a7cc4 PNA	52	struct net_device_path_stack *stack)
	53	{
	54	const void *daddr = &ct->tuplehash[!dir].tuple.src.u3;
	55	struct net_device *dev = dst_cache->dev;
c63a7cc4 PNA	56	struct neighbour *n;
	57	u8 nud_state;
	58
45ca3e61 FF	59	if (!nft_is_valid_ether_device(dev))
	60	goto out;
	61
c63a7cc4 PNA	62	n = dst_neigh_lookup(dst_cache, daddr);
	63	if (!n)
	64	return -1;
	65
	66	read_lock_bh(&n->lock);
	67	nud_state = n->nud_state;
	68	ether_addr_copy(ha, n->ha);
	69	read_unlock_bh(&n->lock);
	70	neigh_release(n);
	71
	72	if (!(nud_state & NUD_VALID))
	73	return -1;
	74
45ca3e61	75	out:
c63a7cc4 PNA	76	return dev_fill_forward_path(dev, ha, stack);
	77	}
	78
	79	struct nft_forward_info {
	80	const struct net_device *indev;
7a27f6ab	81	const struct net_device *outdev;
73f97025	82	const struct net_device *hw_outdev;
4cd91f7c PNA	83	struct id {
	84	__u16 id;
	85	__be16 proto;
	86	} encap[NF_FLOW_TABLE_ENCAP_MAX];
	87	u8 num_encaps;
26267bf9	88	u8 ingress_vlans;
7a27f6ab PNA	89	u8 h_source[ETH_ALEN];
	90	u8 h_dest[ETH_ALEN];
	91	enum flow_offload_xmit_type xmit_type;
c63a7cc4 PNA	92	};
	93
	94	static void nft_dev_path_info(const struct net_device_path_stack *stack,
7a27f6ab	95	struct nft_forward_info *info,
73f97025	96	unsigned char ha, struct nf_flowtable flowtable)
c63a7cc4 PNA	97	{
	98	const struct net_device_path *path;
	99	int i;
	100
7a27f6ab PNA	101	memcpy(info->h_dest, ha, ETH_ALEN);
7a27f6ab PNA	102
c63a7cc4 PNA	103	for (i = 0; i < stack->num_paths; i++) {
	104	path = &stack->path[i];
	105	switch (path->type) {
	106	case DEV_PATH_ETHERNET:
a11e7973	107	case DEV_PATH_DSA:
4cd91f7c	108	case DEV_PATH_VLAN:
72efd585	109	case DEV_PATH_PPPOE:
c63a7cc4	110	info->indev = path->dev;
7a27f6ab PNA	111	if (is_zero_ether_addr(info->h_source))
7a27f6ab PNA	112	memcpy(info->h_source, path->dev->dev_addr, ETH_ALEN);
4cd91f7c PNA	113
	114	if (path->type == DEV_PATH_ETHERNET)
	115	break;
a11e7973 PNA	116	if (path->type == DEV_PATH_DSA) {
	117	i = stack->num_paths;
	118	break;
	119	}
4cd91f7c	120
72efd585	121	/* DEV_PATH_VLAN and DEV_PATH_PPPOE */
4cd91f7c PNA	122	if (info->num_encaps >= NF_FLOW_TABLE_ENCAP_MAX) {
	123	info->indev = NULL;
	124	break;
	125	}
24560749 FF	126	if (!info->outdev)
24560749 FF	127	info->outdev = path->dev;
4cd91f7c PNA	128	info->encap[info->num_encaps].id = path->encap.id;
	129	info->encap[info->num_encaps].proto = path->encap.proto;
	130	info->num_encaps++;
72efd585 PNA	131	if (path->type == DEV_PATH_PPPOE)
72efd585 PNA	132	memcpy(info->h_dest, path->encap.h_dest, ETH_ALEN);
c63a7cc4	133	break;
c63a7cc4	134	case DEV_PATH_BRIDGE:
7a27f6ab PNA	135	if (is_zero_ether_addr(info->h_source))
	136	memcpy(info->h_source, path->dev->dev_addr, ETH_ALEN);
	137
e990cef6	138	switch (path->bridge.vlan_mode) {
26267bf9 FF	139	case DEV_PATH_BR_VLAN_UNTAG_HW:
	140	info->ingress_vlans \|= BIT(info->num_encaps - 1);
	141	break;
e990cef6 PNA	142	case DEV_PATH_BR_VLAN_TAG:
	143	info->encap[info->num_encaps].id = path->bridge.vlan_id;
	144	info->encap[info->num_encaps].proto = path->bridge.vlan_proto;
	145	info->num_encaps++;
	146	break;
	147	case DEV_PATH_BR_VLAN_UNTAG:
	148	info->num_encaps--;
	149	break;
	150	case DEV_PATH_BR_VLAN_KEEP:
	151	break;
	152	}
7a27f6ab PNA	153	info->xmit_type = FLOW_OFFLOAD_XMIT_DIRECT;
7a27f6ab PNA	154	break;
c63a7cc4 PNA	155	default:
	156	info->indev = NULL;
	157	break;
	158	}
	159	}
7a27f6ab PNA	160	if (!info->outdev)
7a27f6ab PNA	161	info->outdev = info->indev;
73f97025 PNA	162
	163	info->hw_outdev = info->indev;
	164
	165	if (nf_flowtable_hw_offload(flowtable) &&
	166	nft_is_valid_ether_device(info->indev))
	167	info->xmit_type = FLOW_OFFLOAD_XMIT_DIRECT;
c63a7cc4 PNA	168	}
	169
	170	static bool nft_flowtable_find_dev(const struct net_device *dev,
	171	struct nft_flowtable *ft)
	172	{
	173	struct nft_hook *hook;
	174	bool found = false;
	175
	176	list_for_each_entry_rcu(hook, &ft->hook_list, list) {
	177	if (hook->ops.dev != dev)
	178	continue;
	179
	180	found = true;
	181	break;
	182	}
	183
	184	return found;
	185	}
	186
	187	static void nft_dev_forward_path(struct nf_flow_route *route,
	188	const struct nf_conn *ct,
	189	enum ip_conntrack_dir dir,
	190	struct nft_flowtable *ft)
	191	{
	192	const struct dst_entry *dst = route->tuple[dir].dst;
	193	struct net_device_path_stack stack;
	194	struct nft_forward_info info = {};
7a27f6ab	195	unsigned char ha[ETH_ALEN];
4cd91f7c	196	int i;
c63a7cc4	197
7a27f6ab	198	if (nft_dev_fill_forward_path(route, dst, ct, dir, ha, &stack) >= 0)
73f97025	199	nft_dev_path_info(&stack, &info, ha, &ft->data);
c63a7cc4 PNA	200
	201	if (!info.indev \|\| !nft_flowtable_find_dev(info.indev, ft))
	202	return;
	203
	204	route->tuple[!dir].in.ifindex = info.indev->ifindex;
4cd91f7c PNA	205	for (i = 0; i < info.num_encaps; i++) {
	206	route->tuple[!dir].in.encap[i].id = info.encap[i].id;
	207	route->tuple[!dir].in.encap[i].proto = info.encap[i].proto;
	208	}
	209	route->tuple[!dir].in.num_encaps = info.num_encaps;
26267bf9	210	route->tuple[!dir].in.ingress_vlans = info.ingress_vlans;
7a27f6ab PNA	211
	212	if (info.xmit_type == FLOW_OFFLOAD_XMIT_DIRECT) {
	213	memcpy(route->tuple[dir].out.h_source, info.h_source, ETH_ALEN);
	214	memcpy(route->tuple[dir].out.h_dest, info.h_dest, ETH_ALEN);
	215	route->tuple[dir].out.ifindex = info.outdev->ifindex;
73f97025	216	route->tuple[dir].out.hw_ifindex = info.hw_outdev->ifindex;
7a27f6ab PNA	217	route->tuple[dir].xmit_type = info.xmit_type;
7a27f6ab PNA	218	}
c63a7cc4 PNA	219	}
c63a7cc4 PNA	220
a3c90f7a PNA	221	static int nft_flow_route(const struct nft_pktinfo *pkt,
	222	const struct nf_conn *ct,
	223	struct nf_flow_route *route,
c63a7cc4 PNA	224	enum ip_conntrack_dir dir,
c63a7cc4 PNA	225	struct nft_flowtable *ft)
a3c90f7a PNA	226	{
	227	struct dst_entry *this_dst = skb_dst(pkt->skb);
	228	struct dst_entry *other_dst = NULL;
	229	struct flowi fl;
	230
	231	memset(&fl, 0, sizeof(fl));
	232	switch (nft_pf(pkt)) {
	233	case NFPROTO_IPV4:
a799aea0	234	fl.u.ip4.daddr = ct->tuplehash[dir].tuple.src.u3.ip;
97629b23	235	fl.u.ip4.saddr = ct->tuplehash[!dir].tuple.src.u3.ip;
10f4e765	236	fl.u.ip4.flowi4_oif = nft_in(pkt)->ifindex;
3412e164 SA	237	fl.u.ip4.flowi4_iif = this_dst->dev->ifindex;
	238	fl.u.ip4.flowi4_tos = RT_TOS(ip_hdr(pkt->skb)->tos);
	239	fl.u.ip4.flowi4_mark = pkt->skb->mark;
f1896d45	240	fl.u.ip4.flowi4_flags = FLOWI_FLAG_ANYSRC;
a3c90f7a PNA	241	break;
a3c90f7a PNA	242	case NFPROTO_IPV6:
a799aea0	243	fl.u.ip6.daddr = ct->tuplehash[dir].tuple.src.u3.in6;
97629b23	244	fl.u.ip6.saddr = ct->tuplehash[!dir].tuple.src.u3.in6;
10f4e765	245	fl.u.ip6.flowi6_oif = nft_in(pkt)->ifindex;
3412e164 SA	246	fl.u.ip6.flowi6_iif = this_dst->dev->ifindex;
	247	fl.u.ip6.flowlabel = ip6_flowinfo(ipv6_hdr(pkt->skb));
	248	fl.u.ip6.flowi6_mark = pkt->skb->mark;
f1896d45	249	fl.u.ip6.flowi6_flags = FLOWI_FLAG_ANYSRC;
a3c90f7a PNA	250	break;
	251	}
	252
fa502c86 PNA	253	if (!dst_hold_safe(this_dst))
	254	return -ENOENT;
	255
a3c90f7a	256	nf_route(nft_net(pkt), &other_dst, &fl, false, nft_pf(pkt));
fa502c86 PNA	257	if (!other_dst) {
fa502c86 PNA	258	dst_release(this_dst);
a3c90f7a	259	return -ENOENT;
fa502c86	260	}
a3c90f7a	261
5139c0c0 PNA	262	nft_default_forward_path(route, this_dst, dir);
5139c0c0 PNA	263	nft_default_forward_path(route, other_dst, !dir);
a3c90f7a	264
c63a7cc4 PNA	265	if (route->tuple[dir].xmit_type == FLOW_OFFLOAD_XMIT_NEIGH &&
	266	route->tuple[!dir].xmit_type == FLOW_OFFLOAD_XMIT_NEIGH) {
	267	nft_dev_forward_path(route, ct, dir, ft);
	268	nft_dev_forward_path(route, ct, !dir, ft);
	269	}
	270
a3c90f7a PNA	271	return 0;
	272	}
	273
69aeb538	274	static bool nft_flow_offload_skip(struct sk_buff *skb, int family)
a3c90f7a	275	{
a3c90f7a PNA	276	if (skb_sec_path(skb))
	277	return true;
	278
69aeb538 FW	279	if (family == NFPROTO_IPV4) {
	280	const struct ip_options *opt;
	281
	282	opt = &(IPCB(skb)->opt);
	283
	284	if (unlikely(opt->optlen))
	285	return true;
	286	}
	287
a3c90f7a PNA	288	return false;
	289	}
	290
	291	static void nft_flow_offload_eval(const struct nft_expr *expr,
	292	struct nft_regs *regs,
	293	const struct nft_pktinfo *pkt)
	294	{
	295	struct nft_flow_offload *priv = nft_expr_priv(expr);
	296	struct nf_flowtable *flowtable = &priv->flowtable->data;
dfe42be1	297	struct tcphdr _tcph, *tcph = NULL;
c63a7cc4	298	struct nf_flow_route route = {};
a3c90f7a	299	enum ip_conntrack_info ctinfo;
a3c90f7a PNA	300	struct flow_offload *flow;
	301	enum ip_conntrack_dir dir;
	302	struct nf_conn *ct;
	303	int ret;
	304
69aeb538	305	if (nft_flow_offload_skip(pkt->skb, nft_pf(pkt)))
a3c90f7a PNA	306	goto out;
	307
	308	ct = nf_ct_get(pkt->skb, &ctinfo);
	309	if (!ct)
	310	goto out;
	311
	312	switch (ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.protonum) {
	313	case IPPROTO_TCP:
2d7b4ace	314	tcph = skb_header_pointer(pkt->skb, nft_thoff(pkt),
dfe42be1	315	sizeof(_tcph), &_tcph);
e5eaac2b PNA	316	if (unlikely(!tcph \|\| tcph->fin \|\| tcph->rst \|\|
e5eaac2b PNA	317	!nf_conntrack_tcp_established(ct)))
dfe42be1	318	goto out;
8437a620	319	break;
a3c90f7a PNA	320	case IPPROTO_UDP:
a3c90f7a PNA	321	break;
4e8d9584 TM	322	#ifdef CONFIG_NF_CT_PROTO_GRE
	323	case IPPROTO_GRE: {
	324	struct nf_conntrack_tuple *tuple;
	325
	326	if (ct->status & IPS_NAT_MASK)
	327	goto out;
	328	tuple = &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple;
	329	/* No support for GRE v1 */
	330	if (tuple->src.u.gre.key \|\| tuple->dst.u.gre.key)
	331	goto out;
	332	break;
	333	}
	334	#endif
a3c90f7a PNA	335	default:
	336	goto out;
	337	}
	338
91a9048f	339	if (nf_ct_ext_exist(ct, NF_CT_EXT_HELPER) \|\|
c4617214	340	ct->status & (IPS_SEQ_ADJUST \| IPS_NAT_CLASH))
a3c90f7a PNA	341	goto out;
a3c90f7a PNA	342
270a8a29	343	if (!nf_ct_is_confirmed(ct))
a3c90f7a PNA	344	goto out;
	345
	346	if (test_and_set_bit(IPS_OFFLOAD_BIT, &ct->status))
	347	goto out;
	348
	349	dir = CTINFO2DIR(ctinfo);
c63a7cc4	350	if (nft_flow_route(pkt, ct, &route, dir, priv->flowtable) < 0)
a3c90f7a PNA	351	goto err_flow_route;
a3c90f7a PNA	352
f1363e05	353	flow = flow_offload_alloc(ct);
a3c90f7a PNA	354	if (!flow)
	355	goto err_flow_alloc;
	356
fa502c86	357	flow_offload_route_init(flow, &route);
f1363e05	358
dfe42be1	359	if (tcph) {
8437a620 FW	360	ct->proto.tcp.seen[0].flags \|= IP_CT_TCP_FLAG_BE_LIBERAL;
	361	ct->proto.tcp.seen[1].flags \|= IP_CT_TCP_FLAG_BE_LIBERAL;
	362	}
	363
a3c90f7a PNA	364	ret = flow_offload_add(flowtable, flow);
	365	if (ret < 0)
	366	goto err_flow_add;
	367
	368	return;
	369
	370	err_flow_add:
	371	flow_offload_free(flow);
	372	err_flow_alloc:
fa502c86	373	dst_release(route.tuple[dir].dst);
a3c90f7a PNA	374	dst_release(route.tuple[!dir].dst);
	375	err_flow_route:
	376	clear_bit(IPS_OFFLOAD_BIT, &ct->status);
	377	out:
	378	regs->verdict.code = NFT_BREAK;
	379	}
	380
	381	static int nft_flow_offload_validate(const struct nft_ctx *ctx,
	382	const struct nft_expr *expr,
	383	const struct nft_data **data)
	384	{
	385	unsigned int hook_mask = (1 << NF_INET_FORWARD);
	386
	387	return nft_chain_validate_hooks(ctx->chain, hook_mask);
	388	}
	389
14c41586 PNA	390	static const struct nla_policy nft_flow_offload_policy[NFTA_FLOW_MAX + 1] = {
	391	[NFTA_FLOW_TABLE_NAME] = { .type = NLA_STRING,
	392	.len = NFT_NAME_MAXLEN - 1 },
	393	};
	394
a3c90f7a PNA	395	static int nft_flow_offload_init(const struct nft_ctx *ctx,
	396	const struct nft_expr *expr,
	397	const struct nlattr * const tb[])
	398	{
	399	struct nft_flow_offload *priv = nft_expr_priv(expr);
	400	u8 genmask = nft_genmask_next(ctx->net);
	401	struct nft_flowtable *flowtable;
	402
	403	if (!tb[NFTA_FLOW_TABLE_NAME])
	404	return -EINVAL;
	405
cac20fcd PNA	406	flowtable = nft_flowtable_lookup(ctx->table, tb[NFTA_FLOW_TABLE_NAME],
cac20fcd PNA	407	genmask);
a3c90f7a PNA	408	if (IS_ERR(flowtable))
	409	return PTR_ERR(flowtable);
	410
1689f259 PNA	411	if (!nft_use_inc(&flowtable->use))
	412	return -EMFILE;
	413
a3c90f7a	414	priv->flowtable = flowtable;
a3c90f7a	415
36596dad	416	return nf_ct_netns_get(ctx->net, ctx->family);
a3c90f7a PNA	417	}
a3c90f7a PNA	418
9b05b6e1 LGL	419	static void nft_flow_offload_deactivate(const struct nft_ctx *ctx,
	420	const struct nft_expr *expr,
	421	enum nft_trans_phase phase)
	422	{
	423	struct nft_flow_offload *priv = nft_expr_priv(expr);
	424
	425	nf_tables_deactivate_flowtable(ctx, priv->flowtable, phase);
	426	}
	427
	428	static void nft_flow_offload_activate(const struct nft_ctx *ctx,
	429	const struct nft_expr *expr)
	430	{
	431	struct nft_flow_offload *priv = nft_expr_priv(expr);
	432
1689f259	433	nft_use_inc_restore(&priv->flowtable->use);
9b05b6e1 LGL	434	}
9b05b6e1 LGL	435
a3c90f7a PNA	436	static void nft_flow_offload_destroy(const struct nft_ctx *ctx,
	437	const struct nft_expr *expr)
	438	{
36596dad	439	nf_ct_netns_put(ctx->net, ctx->family);
a3c90f7a PNA	440	}
a3c90f7a PNA	441
7d34aa3e PS	442	static int nft_flow_offload_dump(struct sk_buff *skb,
7d34aa3e PS	443	const struct nft_expr *expr, bool reset)
a3c90f7a PNA	444	{
	445	struct nft_flow_offload *priv = nft_expr_priv(expr);
	446
	447	if (nla_put_string(skb, NFTA_FLOW_TABLE_NAME, priv->flowtable->name))
	448	goto nla_put_failure;
	449
	450	return 0;
	451
	452	nla_put_failure:
	453	return -1;
	454	}
	455
	456	static struct nft_expr_type nft_flow_offload_type;
	457	static const struct nft_expr_ops nft_flow_offload_ops = {
	458	.type = &nft_flow_offload_type,
	459	.size = NFT_EXPR_SIZE(sizeof(struct nft_flow_offload)),
	460	.eval = nft_flow_offload_eval,
	461	.init = nft_flow_offload_init,
9b05b6e1 LGL	462	.activate = nft_flow_offload_activate,
9b05b6e1 LGL	463	.deactivate = nft_flow_offload_deactivate,
a3c90f7a PNA	464	.destroy = nft_flow_offload_destroy,
	465	.validate = nft_flow_offload_validate,
	466	.dump = nft_flow_offload_dump,
b2d30654	467	.reduce = NFT_REDUCE_READONLY,
a3c90f7a PNA	468	};
	469
	470	static struct nft_expr_type nft_flow_offload_type __read_mostly = {
	471	.name = "flow_offload",
	472	.ops = &nft_flow_offload_ops,
14c41586	473	.policy = nft_flow_offload_policy,
a3c90f7a PNA	474	.maxattr = NFTA_FLOW_MAX,
	475	.owner = THIS_MODULE,
	476	};
	477
a3c90f7a PNA	478	static int flow_offload_netdev_event(struct notifier_block *this,
	479	unsigned long event, void *ptr)
	480	{
	481	struct net_device *dev = netdev_notifier_info_to_dev(ptr);
	482
	483	if (event != NETDEV_DOWN)
	484	return NOTIFY_DONE;
	485
5f1be84a	486	nf_flow_table_cleanup(dev);
a3c90f7a PNA	487
	488	return NOTIFY_DONE;
	489	}
	490
	491	static struct notifier_block flow_offload_netdev_notifier = {
	492	.notifier_call = flow_offload_netdev_event,
	493	};
	494
	495	static int __init nft_flow_offload_module_init(void)
	496	{
	497	int err;
	498
584eab29 TY	499	err = register_netdevice_notifier(&flow_offload_netdev_notifier);
	500	if (err)
	501	goto err;
a3c90f7a PNA	502
	503	err = nft_register_expr(&nft_flow_offload_type);
	504	if (err < 0)
	505	goto register_expr;
	506
	507	return 0;
	508
	509	register_expr:
	510	unregister_netdevice_notifier(&flow_offload_netdev_notifier);
584eab29	511	err:
a3c90f7a PNA	512	return err;
	513	}
	514
	515	static void __exit nft_flow_offload_module_exit(void)
	516	{
a3c90f7a PNA	517	nft_unregister_expr(&nft_flow_offload_type);
a3c90f7a PNA	518	unregister_netdevice_notifier(&flow_offload_netdev_notifier);
a3c90f7a PNA	519	}
	520
	521	module_init(nft_flow_offload_module_init);
	522	module_exit(nft_flow_offload_module_exit);
	523
	524	MODULE_LICENSE("GPL");
	525	MODULE_AUTHOR("Pablo Neira Ayuso <pablo@netfilter.org>");
	526	MODULE_ALIAS_NFT_EXPR("flow_offload");
4cacc395	527	MODULE_DESCRIPTION("nftables hardware flow offload module");