datapath.c

/*
 * Copyright (c) 2007-2012 Nicira, Inc.
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of version 2 of the GNU General Public
 * License as published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful, but
 * WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
 * General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
 * 02110-1301, USA
 */

#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

#include <linux/init.h>
#include <linux/module.h>
#include <linux/if_arp.h>
#include <linux/if_vlan.h>
#include <linux/in.h>
#include <linux/ip.h>
#include <linux/jhash.h>
#include <linux/delay.h>
#include <linux/time.h>
#include <linux/etherdevice.h>
#include <linux/genetlink.h>
#include <linux/kernel.h>
#include <linux/kthread.h>
#include <linux/mutex.h>
#include <linux/percpu.h>
#include <linux/rcupdate.h>
#include <linux/tcp.h>
#include <linux/udp.h>
#include <linux/ethtool.h>
#include <linux/wait.h>
#include <asm/div64.h>
#include <linux/highmem.h>
#include <linux/netfilter_bridge.h>
#include <linux/netfilter_ipv4.h>
#include <linux/inetdevice.h>
#include <linux/list.h>
#include <linux/lockdep.h>
#include <linux/openvswitch.h>
#include <linux/rculist.h>
#include <linux/dmi.h>
#include <linux/workqueue.h>
#include <net/genetlink.h>
#include <net/net_namespace.h>
#include <net/netns/generic.h>

#include "datapath.h"
#include "flow.h"
#include "vport-internal_dev.h"
#include "vport-netdev.h"


#define REHASH_FLOW_INTERVAL (10 * 60 * HZ)
static void rehash_flow_table(struct work_struct *work);
static DECLARE_DELAYED_WORK(rehash_flow_wq, rehash_flow_table);

int ovs_net_id __read_mostly;

static void ovs_notify(struct sk_buff *skb, struct genl_info *info,
		       struct genl_multicast_group *grp)
{
	genl_notify(skb, genl_info_net(info), info->snd_portid,
		    grp->id, info->nlhdr, GFP_KERNEL);
}

/**
 * DOC: Locking:
 *
 * All writes e.g. Writes to device state (add/remove datapath, port, set
 * operations on vports, etc.), Writes to other state (flow table
 * modifications, set miscellaneous datapath parameters, etc.) are protected
 * by ovs_lock.
 *
 * Reads are protected by RCU.
 *
 * There are a few special cases (mostly stats) that have their own
 * synchronization but they nest under all of above and don't interact with
 * each other.
 *
 * The RTNL lock nests inside ovs_mutex.
 */

static DEFINE_MUTEX(ovs_mutex);

void ovs_lock(void)
{
	mutex_lock(&ovs_mutex);
}

void ovs_unlock(void)
{
	mutex_unlock(&ovs_mutex);
}

#ifdef CONFIG_LOCKDEP
int lockdep_ovsl_is_held(void)
{
	if (debug_locks)
		return lockdep_is_held(&ovs_mutex);
	else
		return 1;
}
#endif

static struct vport *new_vport(const struct vport_parms *);
static int queue_gso_packets(struct net *, int dp_ifindex, struct sk_buff *,
			     const struct dp_upcall_info *);
static int queue_userspace_packet(struct net *, int dp_ifindex,
				  struct sk_buff *,
				  const struct dp_upcall_info *);

/* Must be called with rcu_read_lock or ovs_mutex. */
static struct datapath *get_dp(struct net *net, int dp_ifindex)
{
	struct datapath *dp = NULL;
	struct net_device *dev;

	rcu_read_lock();
	dev = dev_get_by_index_rcu(net, dp_ifindex);
	if (dev) {
		struct vport *vport = ovs_internal_dev_get_vport(dev);
		if (vport)
			dp = vport->dp;
	}
	rcu_read_unlock();

	return dp;
}

/* Must be called with rcu_read_lock or ovs_mutex. */
const char *ovs_dp_name(const struct datapath *dp)
{
	struct vport *vport = ovs_vport_ovsl_rcu(dp, OVSP_LOCAL);
	return vport->ops->get_name(vport);
}

static int get_dpifindex(struct datapath *dp)
{
	struct vport *local;
	int ifindex;

	rcu_read_lock();

	local = ovs_vport_rcu(dp, OVSP_LOCAL);
	if (local)
		ifindex = netdev_vport_priv(local)->dev->ifindex;
	else
		ifindex = 0;

	rcu_read_unlock();

	return ifindex;
}

static void destroy_dp_rcu(struct rcu_head *rcu)
{
	struct datapath *dp = container_of(rcu, struct datapath, rcu);

	ovs_flow_tbl_destroy((__force struct flow_table *)dp->table);
	free_percpu(dp->stats_percpu);
	release_net(ovs_dp_get_net(dp));
	kfree(dp->ports);
	kfree(dp);
}

static struct hlist_head *vport_hash_bucket(const struct datapath *dp,
					    u16 port_no)
{
	return &dp->ports[port_no & (DP_VPORT_HASH_BUCKETS - 1)];
}

struct vport *ovs_lookup_vport(const struct datapath *dp, u16 port_no)
{
	struct vport *vport;
	struct hlist_head *head;

	head = vport_hash_bucket(dp, port_no);
	hlist_for_each_entry_rcu(vport, head, dp_hash_node) {
		if (vport->port_no == port_no)
			return vport;
	}
	return NULL;
}

/* Called with ovs_mutex. */
static struct vport *new_vport(const struct vport_parms *parms)
{
	struct vport *vport;

	vport = ovs_vport_add(parms);
	if (!IS_ERR(vport)) {
		struct datapath *dp = parms->dp;
		struct hlist_head *head = vport_hash_bucket(dp, vport->port_no);

		hlist_add_head_rcu(&vport->dp_hash_node, head);
	}
	return vport;
}

void ovs_dp_detach_port(struct vport *p)
{
	ASSERT_OVSL();

	/* First drop references to device. */
	hlist_del_rcu(&p->dp_hash_node);

	/* Then destroy it. */
	ovs_vport_del(p);
}

/* Must be called with rcu_read_lock. */
void ovs_dp_process_received_packet(struct vport *p, struct sk_buff *skb)
{
	struct datapath *dp = p->dp;
	struct sw_flow *flow;
	struct dp_stats_percpu *stats;
	struct sw_flow_key key;
	u64 *stats_counter;
	int error;
	int key_len;

	stats = this_cpu_ptr(dp->stats_percpu);

	/* Extract flow from 'skb' into 'key'. */
	error = ovs_flow_extract(skb, p->port_no, &key, &key_len);
	if (unlikely(error)) {
		kfree_skb(skb);
		return;
	}

	/* Look up flow. */
	flow = ovs_flow_tbl_lookup(rcu_dereference(dp->table), &key, key_len);
	if (unlikely(!flow)) {
		struct dp_upcall_info upcall;

		upcall.cmd = OVS_PACKET_CMD_MISS;
		upcall.key = &key;
		upcall.userdata = NULL;
		upcall.portid = p->upcall_portid;
		ovs_dp_upcall(dp, skb, &upcall);
		consume_skb(skb);
		stats_counter = &stats->n_missed;
		goto out;
	}

	OVS_CB(skb)->flow = flow;

	stats_counter = &stats->n_hit;
	ovs_flow_used(OVS_CB(skb)->flow, skb);
	ovs_execute_actions(dp, skb);

out:
	/* Update datapath statistics. */
	u64_stats_update_begin(&stats->sync);
	(*stats_counter)++;
	u64_stats_update_end(&stats->sync);
}

static struct genl_family dp_packet_genl_family = {
	.id = GENL_ID_GENERATE,
	.hdrsize = sizeof(struct ovs_header),
	.name = OVS_PACKET_FAMILY,
	.version = OVS_PACKET_VERSION,
	.maxattr = OVS_PACKET_ATTR_MAX,
	.netnsok = true,
	.parallel_ops = true,
};

int ovs_dp_upcall(struct datapath *dp, struct sk_buff *skb,
		  const struct dp_upcall_info *upcall_info)
{
	struct dp_stats_percpu *stats;
	int dp_ifindex;
	int err;

	if (upcall_info->portid == 0) {
		err = -ENOTCONN;
		goto err;
	}

	dp_ifindex = get_dpifindex(dp);
	if (!dp_ifindex) {
		err = -ENODEV;
		goto err;
	}

	if (!skb_is_gso(skb))
		err = queue_userspace_packet(ovs_dp_get_net(dp), dp_ifindex, skb, upcall_info);
	else
		err = queue_gso_packets(ovs_dp_get_net(dp), dp_ifindex, skb, upcall_info);
	if (err)
		goto err;

	return 0;

err:
	stats = this_cpu_ptr(dp->stats_percpu);

	u64_stats_update_begin(&stats->sync);
	stats->n_lost++;
	u64_stats_update_end(&stats->sync);

	return err;
}

static int queue_gso_packets(struct net *net, int dp_ifindex,
			     struct sk_buff *skb,
			     const struct dp_upcall_info *upcall_info)
{
	unsigned short gso_type = skb_shinfo(skb)->gso_type;
	struct dp_upcall_info later_info;
	struct sw_flow_key later_key;
	struct sk_buff *segs, *nskb;
	int err;

	segs = __skb_gso_segment(skb, NETIF_F_SG | NETIF_F_HW_CSUM, false);
	if (IS_ERR(segs))
		return PTR_ERR(segs);

	/* Queue all of the segments. */
	skb = segs;
	do {
		err = queue_userspace_packet(net, dp_ifindex, skb, upcall_info);
		if (err)
			break;

		if (skb == segs && gso_type & SKB_GSO_UDP) {
			/* The initial flow key extracted by ovs_flow_extract()
			 * in this case is for a first fragment, so we need to
			 * properly mark later fragments.
			 */
			later_key = *upcall_info->key;
			later_key.ip.frag = OVS_FRAG_TYPE_LATER;

			later_info = *upcall_info;
			later_info.key = &later_key;
			upcall_info = &later_info;
		}
	} while ((skb = skb->next));

	/* Free all of the segments. */
	skb = segs;
	do {
		nskb = skb->next;
		if (err)
			kfree_skb(skb);
		else
			consume_skb(skb);
	} while ((skb = nskb));
	return err;
}

static size_t key_attr_size(void)
{
	return    nla_total_size(4)   /* OVS_KEY_ATTR_PRIORITY */
		+ nla_total_size(4)   /* OVS_KEY_ATTR_IN_PORT */
		+ nla_total_size(4)   /* OVS_KEY_ATTR_SKB_MARK */
		+ nla_total_size(12)  /* OVS_KEY_ATTR_ETHERNET */
		+ nla_total_size(2)   /* OVS_KEY_ATTR_ETHERTYPE */
		+ nla_total_size(4)   /* OVS_KEY_ATTR_8021Q */
		+ nla_total_size(0)   /* OVS_KEY_ATTR_ENCAP */
		+ nla_total_size(2)   /* OVS_KEY_ATTR_ETHERTYPE */
		+ nla_total_size(40)  /* OVS_KEY_ATTR_IPV6 */
		+ nla_total_size(2)   /* OVS_KEY_ATTR_ICMPV6 */
		+ nla_total_size(28); /* OVS_KEY_ATTR_ND */
}

static size_t upcall_msg_size(const struct sk_buff *skb,
			      const struct nlattr *userdata)
{
	size_t size = NLMSG_ALIGN(sizeof(struct ovs_header))
		+ nla_total_size(skb->len) /* OVS_PACKET_ATTR_PACKET */
		+ nla_total_size(key_attr_size()); /* OVS_PACKET_ATTR_KEY */

	/* OVS_PACKET_ATTR_USERDATA */
	if (userdata)
		size += NLA_ALIGN(userdata->nla_len);

	return size;
}

static int queue_userspace_packet(struct net *net, int dp_ifindex,
				  struct sk_buff *skb,
				  const struct dp_upcall_info *upcall_info)
{
	struct ovs_header *upcall;
	struct sk_buff *nskb = NULL;
	struct sk_buff *user_skb; /* to be queued to userspace */
	struct nlattr *nla;
	int err;

	if (vlan_tx_tag_present(skb)) {
		nskb = skb_clone(skb, GFP_ATOMIC);
		if (!nskb)
			return -ENOMEM;

		nskb = __vlan_put_tag(nskb, nskb->vlan_proto, vlan_tx_tag_get(nskb));
		if (!nskb)
			return -ENOMEM;

		nskb->vlan_tci = 0;
		skb = nskb;
	}

	if (nla_attr_size(skb->len) > USHRT_MAX) {
		err = -EFBIG;
		goto out;
	}

	user_skb = genlmsg_new(upcall_msg_size(skb, upcall_info->userdata), GFP_ATOMIC);
	if (!user_skb) {
		err = -ENOMEM;
		goto out;
	}

	upcall = genlmsg_put(user_skb, 0, 0, &dp_packet_genl_family,
			     0, upcall_info->cmd);
	upcall->dp_ifindex = dp_ifindex;

	nla = nla_nest_start(user_skb, OVS_PACKET_ATTR_KEY);
	ovs_flow_to_nlattrs(upcall_info->key, user_skb);
	nla_nest_end(user_skb, nla);

	if (upcall_info->userdata)
		__nla_put(user_skb, OVS_PACKET_ATTR_USERDATA,
			  nla_len(upcall_info->userdata),
			  nla_data(upcall_info->userdata));

	nla = __nla_reserve(user_skb, OVS_PACKET_ATTR_PACKET, skb->len);

	skb_copy_and_csum_dev(skb, nla_data(nla));

	genlmsg_end(user_skb, upcall);
	err = genlmsg_unicast(net, user_skb, upcall_info->portid);

out:
	kfree_skb(nskb);
	return err;
}

/* Called with ovs_mutex. */
static int flush_flows(struct datapath *dp)
{
	struct flow_table *old_table;
	struct flow_table *new_table;

	old_table = ovsl_dereference(dp->table);
	new_table = ovs_flow_tbl_alloc(TBL_MIN_BUCKETS);
	if (!new_table)
		return -ENOMEM;

	rcu_assign_pointer(dp->table, new_table);

	ovs_flow_tbl_deferred_destroy(old_table);
	return 0;
}

static int validate_actions(const struct nlattr *attr,
				const struct sw_flow_key *key, int depth);

static int validate_sample(const struct nlattr *attr,
				const struct sw_flow_key *key, int depth)
{
	const struct nlattr *attrs[OVS_SAMPLE_ATTR_MAX + 1];
	const struct nlattr *probability, *actions;
	const struct nlattr *a;
	int rem;

	memset(attrs, 0, sizeof(attrs));
	nla_for_each_nested(a, attr, rem) {
		int type = nla_type(a);
		if (!type || type > OVS_SAMPLE_ATTR_MAX || attrs[type])
			return -EINVAL;
		attrs[type] = a;
	}
	if (rem)
		return -EINVAL;

	probability = attrs[OVS_SAMPLE_ATTR_PROBABILITY];
	if (!probability || nla_len(probability) != sizeof(u32))
		return -EINVAL;

	actions = attrs[OVS_SAMPLE_ATTR_ACTIONS];
	if (!actions || (nla_len(actions) && nla_len(actions) < NLA_HDRLEN))
		return -EINVAL;
	return validate_actions(actions, key, depth + 1);
}

static int validate_tp_port(const struct sw_flow_key *flow_key)
{
	if (flow_key->eth.type == htons(ETH_P_IP)) {
		if (flow_key->ipv4.tp.src || flow_key->ipv4.tp.dst)
			return 0;
	} else if (flow_key->eth.type == htons(ETH_P_IPV6)) {
		if (flow_key->ipv6.tp.src || flow_key->ipv6.tp.dst)
			return 0;
	}

	return -EINVAL;
}

static int validate_set(const struct nlattr *a,
			const struct sw_flow_key *flow_key)
{
	const struct nlattr *ovs_key = nla_data(a);
	int key_type = nla_type(ovs_key);

	/* There can be only one key in a action */
	if (nla_total_size(nla_len(ovs_key)) != nla_len(a))
		return -EINVAL;

	if (key_type > OVS_KEY_ATTR_MAX ||
	    nla_len(ovs_key) != ovs_key_lens[key_type])
		return -EINVAL;

	switch (key_type) {
	const struct ovs_key_ipv4 *ipv4_key;
	const struct ovs_key_ipv6 *ipv6_key;

	case OVS_KEY_ATTR_PRIORITY:
	case OVS_KEY_ATTR_SKB_MARK:
	case OVS_KEY_ATTR_ETHERNET:
		break;

	case OVS_KEY_ATTR_IPV4:
		if (flow_key->eth.type != htons(ETH_P_IP))
			return -EINVAL;

		if (!flow_key->ip.proto)
			return -EINVAL;

		ipv4_key = nla_data(ovs_key);
		if (ipv4_key->ipv4_proto != flow_key->ip.proto)
			return -EINVAL;

		if (ipv4_key->ipv4_frag != flow_key->ip.frag)
			return -EINVAL;

		break;

	case OVS_KEY_ATTR_IPV6:
		if (flow_key->eth.type != htons(ETH_P_IPV6))
			return -EINVAL;

		if (!flow_key->ip.proto)
			return -EINVAL;

		ipv6_key = nla_data(ovs_key);
		if (ipv6_key->ipv6_proto != flow_key->ip.proto)
			return -EINVAL;

		if (ipv6_key->ipv6_frag != flow_key->ip.frag)
			return -EINVAL;

		if (ntohl(ipv6_key->ipv6_label) & 0xFFF00000)
			return -EINVAL;

		break;

	case OVS_KEY_ATTR_TCP:
		if (flow_key->ip.proto != IPPROTO_TCP)
			return -EINVAL;

		return validate_tp_port(flow_key);

	case OVS_KEY_ATTR_UDP:
		if (flow_key->ip.proto != IPPROTO_UDP)
			return -EINVAL;

		return validate_tp_port(flow_key);

	default:
		return -EINVAL;
	}

	return 0;
}

static int validate_userspace(const struct nlattr *attr)
{
	static const struct nla_policy userspace_policy[OVS_USERSPACE_ATTR_MAX + 1] =	{
		[OVS_USERSPACE_ATTR_PID] = {.type = NLA_U32 },
		[OVS_USERSPACE_ATTR_USERDATA] = {.type = NLA_UNSPEC },
	};
	struct nlattr *a[OVS_USERSPACE_ATTR_MAX + 1];
	int error;

	error = nla_parse_nested(a, OVS_USERSPACE_ATTR_MAX,
				 attr, userspace_policy);
	if (error)
		return error;

	if (!a[OVS_USERSPACE_ATTR_PID] ||
	    !nla_get_u32(a[OVS_USERSPACE_ATTR_PID]))
		return -EINVAL;

	return 0;
}

static int validate_actions(const struct nlattr *attr,
				const struct sw_flow_key *key,  int depth)
{
	const struct nlattr *a;
	int rem, err;

	if (depth >= SAMPLE_ACTION_DEPTH)
		return -EOVERFLOW;

	nla_for_each_nested(a, attr, rem) {
		/* Expected argument lengths, (u32)-1 for variable length. */
		static const u32 action_lens[OVS_ACTION_ATTR_MAX + 1] = {
			[OVS_ACTION_ATTR_OUTPUT] = sizeof(u32),
			[OVS_ACTION_ATTR_USERSPACE] = (u32)-1,
			[OVS_ACTION_ATTR_PUSH_VLAN] = sizeof(struct ovs_action_push_vlan),
			[OVS_ACTION_ATTR_POP_VLAN] = 0,
			[OVS_ACTION_ATTR_SET] = (u32)-1,
			[OVS_ACTION_ATTR_SAMPLE] = (u32)-1
		};
		const struct ovs_action_push_vlan *vlan;
		int type = nla_type(a);

		if (type > OVS_ACTION_ATTR_MAX ||
		    (action_lens[type] != nla_len(a) &&
		     action_lens[type] != (u32)-1))
			return -EINVAL;

		switch (type) {
		case OVS_ACTION_ATTR_UNSPEC:
			return -EINVAL;

		case OVS_ACTION_ATTR_USERSPACE:
			err = validate_userspace(a);
			if (err)
				return err;
			break;

		case OVS_ACTION_ATTR_OUTPUT:
			if (nla_get_u32(a) >= DP_MAX_PORTS)
				return -EINVAL;
			break;


		case OVS_ACTION_ATTR_POP_VLAN:
			break;

		case OVS_ACTION_ATTR_PUSH_VLAN:
			vlan = nla_data(a);
			if (vlan->vlan_tpid != htons(ETH_P_8021Q))
				return -EINVAL;
			if (!(vlan->vlan_tci & htons(VLAN_TAG_PRESENT)))
				return -EINVAL;
			break;

		case OVS_ACTION_ATTR_SET:
			err = validate_set(a, key);
			if (err)
				return err;
			break;

		case OVS_ACTION_ATTR_SAMPLE:
			err = validate_sample(a, key, depth);
			if (err)
				return err;
			break;

		default:
			return -EINVAL;
		}
	}

	if (rem > 0)
		return -EINVAL;

	return 0;
}

static void clear_stats(struct sw_flow *flow)
{
	flow->used = 0;
	flow->tcp_flags = 0;
	flow->packet_count = 0;
	flow->byte_count = 0;
}

static int ovs_packet_cmd_execute(struct sk_buff *skb, struct genl_info *info)
{
	struct ovs_header *ovs_header = info->userhdr;
	struct nlattr **a = info->attrs;
	struct sw_flow_actions *acts;
	struct sk_buff *packet;
	struct sw_flow *flow;
	struct datapath *dp;
	struct ethhdr *eth;
	int len;
	int err;
	int key_len;

	err = -EINVAL;
	if (!a[OVS_PACKET_ATTR_PACKET] || !a[OVS_PACKET_ATTR_KEY] ||
	    !a[OVS_PACKET_ATTR_ACTIONS])
		goto err;

	len = nla_len(a[OVS_PACKET_ATTR_PACKET]);
	packet = __dev_alloc_skb(NET_IP_ALIGN + len, GFP_KERNEL);
	err = -ENOMEM;
	if (!packet)
		goto err;
	skb_reserve(packet, NET_IP_ALIGN);

	nla_memcpy(__skb_put(packet, len), a[OVS_PACKET_ATTR_PACKET], len);

	skb_reset_mac_header(packet);
	eth = eth_hdr(packet);

	/* Normally, setting the skb 'protocol' field would be handled by a
	 * call to eth_type_trans(), but it assumes there's a sending
	 * device, which we may not have. */
	if (ntohs(eth->h_proto) >= ETH_P_802_3_MIN)
		packet->protocol = eth->h_proto;
	else
		packet->protocol = htons(ETH_P_802_2);

	/* Build an sw_flow for sending this packet. */
	flow = ovs_flow_alloc();
	err = PTR_ERR(flow);
	if (IS_ERR(flow))
		goto err_kfree_skb;

	err = ovs_flow_extract(packet, -1, &flow->key, &key_len);
	if (err)
		goto err_flow_free;

	err = ovs_flow_metadata_from_nlattrs(flow, a[OVS_PACKET_ATTR_KEY]);
	if (err)
		goto err_flow_free;

	err = validate_actions(a[OVS_PACKET_ATTR_ACTIONS], &flow->key, 0);
	if (err)
		goto err_flow_free;

	flow->hash = ovs_flow_hash(&flow->key, key_len);

	acts = ovs_flow_actions_alloc(a[OVS_PACKET_ATTR_ACTIONS]);
	err = PTR_ERR(acts);
	if (IS_ERR(acts))
		goto err_flow_free;
	rcu_assign_pointer(flow->sf_acts, acts);

	OVS_CB(packet)->flow = flow;
	packet->priority = flow->key.phy.priority;
	packet->mark = flow->key.phy.skb_mark;

	rcu_read_lock();
	dp = get_dp(sock_net(skb->sk), ovs_header->dp_ifindex);
	err = -ENODEV;
	if (!dp)
		goto err_unlock;

	local_bh_disable();
	err = ovs_execute_actions(dp, packet);
	local_bh_enable();
	rcu_read_unlock();

	ovs_flow_free(flow);
	return err;

err_unlock:
	rcu_read_unlock();
err_flow_free:
	ovs_flow_free(flow);
err_kfree_skb:
	kfree_skb(packet);
err:
	return err;
}

static const struct nla_policy packet_policy[OVS_PACKET_ATTR_MAX + 1] = {
	[OVS_PACKET_ATTR_PACKET] = { .len = ETH_HLEN },
	[OVS_PACKET_ATTR_KEY] = { .type = NLA_NESTED },
	[OVS_PACKET_ATTR_ACTIONS] = { .type = NLA_NESTED },
};

static struct genl_ops dp_packet_genl_ops[] = {
	{ .cmd = OVS_PACKET_CMD_EXECUTE,
	  .flags = GENL_ADMIN_PERM, /* Requires CAP_NET_ADMIN privilege. */
	  .policy = packet_policy,
	  .doit = ovs_packet_cmd_execute
	}
};

static void get_dp_stats(struct datapath *dp, struct ovs_dp_stats *stats)
{
	int i;
	struct flow_table *table = ovsl_dereference(dp->table);

	stats->n_flows = ovs_flow_tbl_count(table);

	stats->n_hit = stats->n_missed = stats->n_lost = 0;
	for_each_possible_cpu(i) {
		const struct dp_stats_percpu *percpu_stats;
		struct dp_stats_percpu local_stats;
		unsigned int start;

		percpu_stats = per_cpu_ptr(dp->stats_percpu, i);

		do {
			start = u64_stats_fetch_begin_bh(&percpu_stats->sync);
			local_stats = *percpu_stats;
		} while (u64_stats_fetch_retry_bh(&percpu_stats->sync, start));

		stats->n_hit += local_stats.n_hit;
		stats->n_missed += local_stats.n_missed;
		stats->n_lost += local_stats.n_lost;
	}
}

static const struct nla_policy flow_policy[OVS_FLOW_ATTR_MAX + 1] = {
	[OVS_FLOW_ATTR_KEY] = { .type = NLA_NESTED },
	[OVS_FLOW_ATTR_ACTIONS] = { .type = NLA_NESTED },
	[OVS_FLOW_ATTR_CLEAR] = { .type = NLA_FLAG },
};

static struct genl_family dp_flow_genl_family = {
	.id = GENL_ID_GENERATE,
	.hdrsize = sizeof(struct ovs_header),
	.name = OVS_FLOW_FAMILY,
	.version = OVS_FLOW_VERSION,
	.maxattr = OVS_FLOW_ATTR_MAX,
	.netnsok = true,
	.parallel_ops = true,
};

static struct genl_multicast_group ovs_dp_flow_multicast_group = {
	.name = OVS_FLOW_MCGROUP
};

static size_t ovs_flow_cmd_msg_size(const struct sw_flow_actions *acts)
{
	return NLMSG_ALIGN(sizeof(struct ovs_header))
		+ nla_total_size(key_attr_size()) /* OVS_FLOW_ATTR_KEY */
		+ nla_total_size(sizeof(struct ovs_flow_stats)) /* OVS_FLOW_ATTR_STATS */
		+ nla_total_size(1) /* OVS_FLOW_ATTR_TCP_FLAGS */
		+ nla_total_size(8) /* OVS_FLOW_ATTR_USED */
		+ nla_total_size(acts->actions_len); /* OVS_FLOW_ATTR_ACTIONS */
}

/* Called with ovs_mutex. */
static int ovs_flow_cmd_fill_info(struct sw_flow *flow, struct datapath *dp,
				  struct sk_buff *skb, u32 portid,
				  u32 seq, u32 flags, u8 cmd)
{
	const int skb_orig_len = skb->len;
	const struct sw_flow_actions *sf_acts;
	struct ovs_flow_stats stats;
	struct ovs_header *ovs_header;
	struct nlattr *nla;
	unsigned long used;
	u8 tcp_flags;
	int err;

	sf_acts = ovsl_dereference(flow->sf_acts);

	ovs_header = genlmsg_put(skb, portid, seq, &dp_flow_genl_family, flags, cmd);
	if (!ovs_header)
		return -EMSGSIZE;

	ovs_header->dp_ifindex = get_dpifindex(dp);

	nla = nla_nest_start(skb, OVS_FLOW_ATTR_KEY);
	if (!nla)
		goto nla_put_failure;
	err = ovs_flow_to_nlattrs(&flow->key, skb);
	if (err)
		goto error;
	nla_nest_end(skb, nla);

	spin_lock_bh(&flow->lock);
	used = flow->used;
	stats.n_packets = flow->packet_count;
	stats.n_bytes = flow->byte_count;
	tcp_flags = flow->tcp_flags;
	spin_unlock_bh(&flow->lock);

	if (used &&
	    nla_put_u64(skb, OVS_FLOW_ATTR_USED, ovs_flow_used_time(used)))
		goto nla_put_failure;

	if (stats.n_packets &&
	    nla_put(skb, OVS_FLOW_ATTR_STATS,
		    sizeof(struct ovs_flow_stats), &stats))
		goto nla_put_failure;

	if (tcp_flags &&
	    nla_put_u8(skb, OVS_FLOW_ATTR_TCP_FLAGS, tcp_flags))
		goto nla_put_failure;

	/* If OVS_FLOW_ATTR_ACTIONS doesn't fit, skip dumping the actions if
	 * this is the first flow to be dumped into 'skb'.  This is unusual for
	 * Netlink but individual action lists can be longer than
	 * NLMSG_GOODSIZE and thus entirely undumpable if we didn't do this.
	 * The userspace caller can always fetch the actions separately if it
	 * really wants them.  (Most userspace callers in fact don't care.)
	 *
	 * This can only fail for dump operations because the skb is always
	 * properly sized for single flows.
	 */
	err = nla_put(skb, OVS_FLOW_ATTR_ACTIONS, sf_acts->actions_len,
		      sf_acts->actions);
	if (err < 0 && skb_orig_len)
		goto error;

	return genlmsg_end(skb, ovs_header);

nla_put_failure:
	err = -EMSGSIZE;
error:
	genlmsg_cancel(skb, ovs_header);
	return err;
}

static struct sk_buff *ovs_flow_cmd_alloc_info(struct sw_flow *flow)
{
	const struct sw_flow_actions *sf_acts;

	sf_acts = ovsl_dereference(flow->sf_acts);

	return genlmsg_new(ovs_flow_cmd_msg_size(sf_acts), GFP_KERNEL);
}

static struct sk_buff *ovs_flow_cmd_build_info(struct sw_flow *flow,
					       struct datapath *dp,
					       u32 portid, u32 seq, u8 cmd)
{
	struct sk_buff *skb;
	int retval;

	skb = ovs_flow_cmd_alloc_info(flow);
	if (!skb)
		return ERR_PTR(-ENOMEM);

	retval = ovs_flow_cmd_fill_info(flow, dp, skb, portid, seq, 0, cmd);
	BUG_ON(retval < 0);
	return skb;
}

static int ovs_flow_cmd_new_or_set(struct sk_buff *skb, struct genl_info *info)
{
	struct nlattr **a = info->attrs;
	struct ovs_header *ovs_header = info->userhdr;
	struct sw_flow_key key;
	struct sw_flow *flow;
	struct sk_buff *reply;
	struct datapath *dp;
	struct flow_table *table;
	int error;
	int key_len;

	/* Extract key. */
	error = -EINVAL;
	if (!a[OVS_FLOW_ATTR_KEY])
		goto error;
	error = ovs_flow_from_nlattrs(&key, &key_len, a[OVS_FLOW_ATTR_KEY]);
	if (error)
		goto error;

	/* Validate actions. */
	if (a[OVS_FLOW_ATTR_ACTIONS]) {
		error = validate_actions(a[OVS_FLOW_ATTR_ACTIONS], &key,  0);
		if (error)
			goto error;
	} else if (info->genlhdr->cmd == OVS_FLOW_CMD_NEW) {
		error = -EINVAL;
		goto error;
	}

	ovs_lock();
	dp = get_dp(sock_net(skb->sk), ovs_header->dp_ifindex);
	error = -ENODEV;
	if (!dp)
		goto err_unlock_ovs;

	table = ovsl_dereference(dp->table);
	flow = ovs_flow_tbl_lookup(table, &key, key_len);
	if (!flow) {
		struct sw_flow_actions *acts;

		/* Bail out if we're not allowed to create a new flow. */
		error = -ENOENT;
		if (info->genlhdr->cmd == OVS_FLOW_CMD_SET)
			goto err_unlock_ovs;