[deliverable/linux.git] / net / openvswitch / flow.c

/*
 * Copyright (c) 2007-2013 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
 */

#include "flow.h"
#include "datapath.h"
#include <linux/uaccess.h>
#include <linux/netdevice.h>
#include <linux/etherdevice.h>
#include <linux/if_ether.h>
#include <linux/if_vlan.h>
#include <net/llc_pdu.h>
#include <linux/kernel.h>
#include <linux/jhash.h>
#include <linux/jiffies.h>
#include <linux/llc.h>
#include <linux/module.h>
#include <linux/in.h>
#include <linux/rcupdate.h>
#include <linux/if_arp.h>
#include <linux/ip.h>
#include <linux/ipv6.h>
#include <linux/sctp.h>
#include <linux/tcp.h>
#include <linux/udp.h>
#include <linux/icmp.h>
#include <linux/icmpv6.h>
#include <linux/rculist.h>
#include <net/ip.h>
#include <net/ip_tunnels.h>
#include <net/ipv6.h>
#include <net/ndisc.h>

u64 ovs_flow_used_time(unsigned long flow_jiffies)
{
	struct timespec cur_ts;
	u64 cur_ms, idle_ms;

	ktime_get_ts(&cur_ts);
	idle_ms = jiffies_to_msecs(jiffies - flow_jiffies);
	cur_ms = (u64)cur_ts.tv_sec * MSEC_PER_SEC +
		 cur_ts.tv_nsec / NSEC_PER_MSEC;

	return cur_ms - idle_ms;
}

#define TCP_FLAGS_BE16(tp) (*(__be16 *)&tcp_flag_word(tp) & htons(0x0FFF))

void ovs_flow_used(struct sw_flow *flow, struct sk_buff *skb)
{
	__be16 tcp_flags = 0;

	if ((flow->key.eth.type == htons(ETH_P_IP) ||
	     flow->key.eth.type == htons(ETH_P_IPV6)) &&
	    flow->key.ip.proto == IPPROTO_TCP &&
	    likely(skb->len >= skb_transport_offset(skb) + sizeof(struct tcphdr))) {
		tcp_flags = TCP_FLAGS_BE16(tcp_hdr(skb));
	}

	spin_lock(&flow->lock);
	flow->used = jiffies;
	flow->packet_count++;
	flow->byte_count += skb->len;
	flow->tcp_flags |= tcp_flags;
	spin_unlock(&flow->lock);
}

static int check_header(struct sk_buff *skb, int len)
{
	if (unlikely(skb->len < len))
		return -EINVAL;
	if (unlikely(!pskb_may_pull(skb, len)))
		return -ENOMEM;
	return 0;
}

static bool arphdr_ok(struct sk_buff *skb)
{
	return pskb_may_pull(skb, skb_network_offset(skb) +
				  sizeof(struct arp_eth_header));
}

static int check_iphdr(struct sk_buff *skb)
{
	unsigned int nh_ofs = skb_network_offset(skb);
	unsigned int ip_len;
	int err;

	err = check_header(skb, nh_ofs + sizeof(struct iphdr));
	if (unlikely(err))
		return err;

	ip_len = ip_hdrlen(skb);
	if (unlikely(ip_len < sizeof(struct iphdr) ||
		     skb->len < nh_ofs + ip_len))
		return -EINVAL;

	skb_set_transport_header(skb, nh_ofs + ip_len);
	return 0;
}

static bool tcphdr_ok(struct sk_buff *skb)
{
	int th_ofs = skb_transport_offset(skb);
	int tcp_len;

	if (unlikely(!pskb_may_pull(skb, th_ofs + sizeof(struct tcphdr))))
		return false;

	tcp_len = tcp_hdrlen(skb);
	if (unlikely(tcp_len < sizeof(struct tcphdr) ||
		     skb->len < th_ofs + tcp_len))
		return false;

	return true;
}

static bool udphdr_ok(struct sk_buff *skb)
{
	return pskb_may_pull(skb, skb_transport_offset(skb) +
				  sizeof(struct udphdr));
}

static bool sctphdr_ok(struct sk_buff *skb)
{
	return pskb_may_pull(skb, skb_transport_offset(skb) +
				  sizeof(struct sctphdr));
}

static bool icmphdr_ok(struct sk_buff *skb)
{
	return pskb_may_pull(skb, skb_transport_offset(skb) +
				  sizeof(struct icmphdr));
}

static int parse_ipv6hdr(struct sk_buff *skb, struct sw_flow_key *key)
{
	unsigned int nh_ofs = skb_network_offset(skb);
	unsigned int nh_len;
	int payload_ofs;
	struct ipv6hdr *nh;
	uint8_t nexthdr;
	__be16 frag_off;
	int err;

	err = check_header(skb, nh_ofs + sizeof(*nh));
	if (unlikely(err))
		return err;

	nh = ipv6_hdr(skb);
	nexthdr = nh->nexthdr;
	payload_ofs = (u8 *)(nh + 1) - skb->data;

	key->ip.proto = NEXTHDR_NONE;
	key->ip.tos = ipv6_get_dsfield(nh);
	key->ip.ttl = nh->hop_limit;
	key->ipv6.label = *(__be32 *)nh & htonl(IPV6_FLOWINFO_FLOWLABEL);
	key->ipv6.addr.src = nh->saddr;
	key->ipv6.addr.dst = nh->daddr;

	payload_ofs = ipv6_skip_exthdr(skb, payload_ofs, &nexthdr, &frag_off);
	if (unlikely(payload_ofs < 0))
		return -EINVAL;

	if (frag_off) {
		if (frag_off & htons(~0x7))
			key->ip.frag = OVS_FRAG_TYPE_LATER;
		else
			key->ip.frag = OVS_FRAG_TYPE_FIRST;
	}

	nh_len = payload_ofs - nh_ofs;
	skb_set_transport_header(skb, nh_ofs + nh_len);
	key->ip.proto = nexthdr;
	return nh_len;
}

static bool icmp6hdr_ok(struct sk_buff *skb)
{
	return pskb_may_pull(skb, skb_transport_offset(skb) +
				  sizeof(struct icmp6hdr));
}

static int parse_vlan(struct sk_buff *skb, struct sw_flow_key *key)
{
	struct qtag_prefix {
		__be16 eth_type; /* ETH_P_8021Q */
		__be16 tci;
	};
	struct qtag_prefix *qp;

	if (unlikely(skb->len < sizeof(struct qtag_prefix) + sizeof(__be16)))
		return 0;

	if (unlikely(!pskb_may_pull(skb, sizeof(struct qtag_prefix) +
					 sizeof(__be16))))
		return -ENOMEM;

	qp = (struct qtag_prefix *) skb->data;
	key->eth.tci = qp->tci | htons(VLAN_TAG_PRESENT);
	__skb_pull(skb, sizeof(struct qtag_prefix));

	return 0;
}

static __be16 parse_ethertype(struct sk_buff *skb)
{
	struct llc_snap_hdr {
		u8  dsap;  /* Always 0xAA */
		u8  ssap;  /* Always 0xAA */
		u8  ctrl;
		u8  oui[3];
		__be16 ethertype;
	};
	struct llc_snap_hdr *llc;
	__be16 proto;

	proto = *(__be16 *) skb->data;
	__skb_pull(skb, sizeof(__be16));

	if (ntohs(proto) >= ETH_P_802_3_MIN)
		return proto;

	if (skb->len < sizeof(struct llc_snap_hdr))
		return htons(ETH_P_802_2);

	if (unlikely(!pskb_may_pull(skb, sizeof(struct llc_snap_hdr))))
		return htons(0);

	llc = (struct llc_snap_hdr *) skb->data;
	if (llc->dsap != LLC_SAP_SNAP ||
	    llc->ssap != LLC_SAP_SNAP ||
	    (llc->oui[0] | llc->oui[1] | llc->oui[2]) != 0)
		return htons(ETH_P_802_2);

	__skb_pull(skb, sizeof(struct llc_snap_hdr));

	if (ntohs(llc->ethertype) >= ETH_P_802_3_MIN)
		return llc->ethertype;

	return htons(ETH_P_802_2);
}

static int parse_icmpv6(struct sk_buff *skb, struct sw_flow_key *key,
			int nh_len)
{
	struct icmp6hdr *icmp = icmp6_hdr(skb);

	/* The ICMPv6 type and code fields use the 16-bit transport port
	 * fields, so we need to store them in 16-bit network byte order.
	 */
	key->ipv6.tp.src = htons(icmp->icmp6_type);
	key->ipv6.tp.dst = htons(icmp->icmp6_code);

	if (icmp->icmp6_code == 0 &&
	    (icmp->icmp6_type == NDISC_NEIGHBOUR_SOLICITATION ||
	     icmp->icmp6_type == NDISC_NEIGHBOUR_ADVERTISEMENT)) {
		int icmp_len = skb->len - skb_transport_offset(skb);
		struct nd_msg *nd;
		int offset;

		/* In order to process neighbor discovery options, we need the
		 * entire packet.
		 */
		if (unlikely(icmp_len < sizeof(*nd)))
			return 0;

		if (unlikely(skb_linearize(skb)))
			return -ENOMEM;

		nd = (struct nd_msg *)skb_transport_header(skb);
		key->ipv6.nd.target = nd->target;

		icmp_len -= sizeof(*nd);
		offset = 0;
		while (icmp_len >= 8) {
			struct nd_opt_hdr *nd_opt =
				 (struct nd_opt_hdr *)(nd->opt + offset);
			int opt_len = nd_opt->nd_opt_len * 8;

			if (unlikely(!opt_len || opt_len > icmp_len))
				return 0;

			/* Store the link layer address if the appropriate
			 * option is provided.  It is considered an error if
			 * the same link layer option is specified twice.
			 */
			if (nd_opt->nd_opt_type == ND_OPT_SOURCE_LL_ADDR
			    && opt_len == 8) {
				if (unlikely(!is_zero_ether_addr(key->ipv6.nd.sll)))
					goto invalid;
				memcpy(key->ipv6.nd.sll,
				    &nd->opt[offset+sizeof(*nd_opt)], ETH_ALEN);
			} else if (nd_opt->nd_opt_type == ND_OPT_TARGET_LL_ADDR
				   && opt_len == 8) {
				if (unlikely(!is_zero_ether_addr(key->ipv6.nd.tll)))
					goto invalid;
				memcpy(key->ipv6.nd.tll,
				    &nd->opt[offset+sizeof(*nd_opt)], ETH_ALEN);
			}

			icmp_len -= opt_len;
			offset += opt_len;
		}
	}

	return 0;

invalid:
	memset(&key->ipv6.nd.target, 0, sizeof(key->ipv6.nd.target));
	memset(key->ipv6.nd.sll, 0, sizeof(key->ipv6.nd.sll));
	memset(key->ipv6.nd.tll, 0, sizeof(key->ipv6.nd.tll));

	return 0;
}

/**
 * ovs_flow_extract - extracts a flow key from an Ethernet frame.
 * @skb: sk_buff that contains the frame, with skb->data pointing to the
 * Ethernet header
 * @in_port: port number on which @skb was received.
 * @key: output flow key
 *
 * The caller must ensure that skb->len >= ETH_HLEN.
 *
 * Returns 0 if successful, otherwise a negative errno value.
 *
 * Initializes @skb header pointers as follows:
 *
 *    - skb->mac_header: the Ethernet header.
 *
 *    - skb->network_header: just past the Ethernet header, or just past the
 *      VLAN header, to the first byte of the Ethernet payload.
 *
 *    - skb->transport_header: If key->eth.type is ETH_P_IP or ETH_P_IPV6
 *      on output, then just past the IP header, if one is present and
 *      of a correct length, otherwise the same as skb->network_header.
 *      For other key->eth.type values it is left untouched.
 */
int ovs_flow_extract(struct sk_buff *skb, u16 in_port, struct sw_flow_key *key)
{
	int error;
	struct ethhdr *eth;

	memset(key, 0, sizeof(*key));

	key->phy.priority = skb->priority;
	if (OVS_CB(skb)->tun_key)
		memcpy(&key->tun_key, OVS_CB(skb)->tun_key, sizeof(key->tun_key));
	key->phy.in_port = in_port;
	key->phy.skb_mark = skb->mark;

	skb_reset_mac_header(skb);

	/* Link layer.  We are guaranteed to have at least the 14 byte Ethernet
	 * header in the linear data area.
	 */
	eth = eth_hdr(skb);
	memcpy(key->eth.src, eth->h_source, ETH_ALEN);
	memcpy(key->eth.dst, eth->h_dest, ETH_ALEN);

	__skb_pull(skb, 2 * ETH_ALEN);
	/* We are going to push all headers that we pull, so no need to
	 * update skb->csum here.
	 */

	if (vlan_tx_tag_present(skb))
		key->eth.tci = htons(skb->vlan_tci);
	else if (eth->h_proto == htons(ETH_P_8021Q))
		if (unlikely(parse_vlan(skb, key)))
			return -ENOMEM;

	key->eth.type = parse_ethertype(skb);
	if (unlikely(key->eth.type == htons(0)))
		return -ENOMEM;

	skb_reset_network_header(skb);
	__skb_push(skb, skb->data - skb_mac_header(skb));

	/* Network layer. */
	if (key->eth.type == htons(ETH_P_IP)) {
		struct iphdr *nh;
		__be16 offset;

		error = check_iphdr(skb);
		if (unlikely(error)) {
			if (error == -EINVAL) {
				skb->transport_header = skb->network_header;
				error = 0;
			}
			return error;
		}

		nh = ip_hdr(skb);
		key->ipv4.addr.src = nh->saddr;
		key->ipv4.addr.dst = nh->daddr;

		key->ip.proto = nh->protocol;
		key->ip.tos = nh->tos;
		key->ip.ttl = nh->ttl;

		offset = nh->frag_off & htons(IP_OFFSET);
		if (offset) {
			key->ip.frag = OVS_FRAG_TYPE_LATER;
			return 0;
		}
		if (nh->frag_off & htons(IP_MF) ||
			 skb_shinfo(skb)->gso_type & SKB_GSO_UDP)
			key->ip.frag = OVS_FRAG_TYPE_FIRST;

		/* Transport layer. */
		if (key->ip.proto == IPPROTO_TCP) {
			if (tcphdr_ok(skb)) {
				struct tcphdr *tcp = tcp_hdr(skb);
				key->ipv4.tp.src = tcp->source;
				key->ipv4.tp.dst = tcp->dest;
				key->ipv4.tp.flags = TCP_FLAGS_BE16(tcp);
			}
		} else if (key->ip.proto == IPPROTO_UDP) {
			if (udphdr_ok(skb)) {
				struct udphdr *udp = udp_hdr(skb);
				key->ipv4.tp.src = udp->source;
				key->ipv4.tp.dst = udp->dest;
			}
		} else if (key->ip.proto == IPPROTO_SCTP) {
			if (sctphdr_ok(skb)) {
				struct sctphdr *sctp = sctp_hdr(skb);
				key->ipv4.tp.src = sctp->source;
				key->ipv4.tp.dst = sctp->dest;
			}
		} else if (key->ip.proto == IPPROTO_ICMP) {
			if (icmphdr_ok(skb)) {
				struct icmphdr *icmp = icmp_hdr(skb);
				/* The ICMP type and code fields use the 16-bit
				 * transport port fields, so we need to store
				 * them in 16-bit network byte order. */
				key->ipv4.tp.src = htons(icmp->type);
				key->ipv4.tp.dst = htons(icmp->code);
			}
		}

	} else if ((key->eth.type == htons(ETH_P_ARP) ||
		   key->eth.type == htons(ETH_P_RARP)) && arphdr_ok(skb)) {
		struct arp_eth_header *arp;

		arp = (struct arp_eth_header *)skb_network_header(skb);

		if (arp->ar_hrd == htons(ARPHRD_ETHER)
				&& arp->ar_pro == htons(ETH_P_IP)
				&& arp->ar_hln == ETH_ALEN
				&& arp->ar_pln == 4) {

			/* We only match on the lower 8 bits of the opcode. */
			if (ntohs(arp->ar_op) <= 0xff)
				key->ip.proto = ntohs(arp->ar_op);
			memcpy(&key->ipv4.addr.src, arp->ar_sip, sizeof(key->ipv4.addr.src));
			memcpy(&key->ipv4.addr.dst, arp->ar_tip, sizeof(key->ipv4.addr.dst));
			memcpy(key->ipv4.arp.sha, arp->ar_sha, ETH_ALEN);
			memcpy(key->ipv4.arp.tha, arp->ar_tha, ETH_ALEN);
		}
	} else if (key->eth.type == htons(ETH_P_IPV6)) {
		int nh_len;             /* IPv6 Header + Extensions */

		nh_len = parse_ipv6hdr(skb, key);
		if (unlikely(nh_len < 0)) {
			if (nh_len == -EINVAL) {
				skb->transport_header = skb->network_header;
				error = 0;
			} else {
				error = nh_len;
			}
			return error;
		}

		if (key->ip.frag == OVS_FRAG_TYPE_LATER)
			return 0;
		if (skb_shinfo(skb)->gso_type & SKB_GSO_UDP)
			key->ip.frag = OVS_FRAG_TYPE_FIRST;

		/* Transport layer. */
		if (key->ip.proto == NEXTHDR_TCP) {
			if (tcphdr_ok(skb)) {
				struct tcphdr *tcp = tcp_hdr(skb);
				key->ipv6.tp.src = tcp->source;
				key->ipv6.tp.dst = tcp->dest;
				key->ipv6.tp.flags = TCP_FLAGS_BE16(tcp);
			}
		} else if (key->ip.proto == NEXTHDR_UDP) {
			if (udphdr_ok(skb)) {
				struct udphdr *udp = udp_hdr(skb);
				key->ipv6.tp.src = udp->source;
				key->ipv6.tp.dst = udp->dest;
			}
		} else if (key->ip.proto == NEXTHDR_SCTP) {
			if (sctphdr_ok(skb)) {
				struct sctphdr *sctp = sctp_hdr(skb);
				key->ipv6.tp.src = sctp->source;
				key->ipv6.tp.dst = sctp->dest;
			}
		} else if (key->ip.proto == NEXTHDR_ICMP) {
			if (icmp6hdr_ok(skb)) {
				error = parse_icmpv6(skb, key, nh_len);
				if (error)
					return error;
			}
		}
	}

	return 0;
}
Commit	Line	Data
ccb1352e	1	/*
03f0d916	2	* Copyright (c) 2007-2013 Nicira, Inc.
ccb1352e JG	3	*
	4	* This program is free software; you can redistribute it and/or
	5	* modify it under the terms of version 2 of the GNU General Public
	6	* License as published by the Free Software Foundation.
	7	*
	8	* This program is distributed in the hope that it will be useful, but
	9	* WITHOUT ANY WARRANTY; without even the implied warranty of
	10	* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
	11	* General Public License for more details.
	12	*
	13	* You should have received a copy of the GNU General Public License
	14	* along with this program; if not, write to the Free Software
	15	* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
	16	* 02110-1301, USA
	17	*/
	18
	19	#include "flow.h"
	20	#include "datapath.h"
	21	#include <linux/uaccess.h>
	22	#include <linux/netdevice.h>
	23	#include <linux/etherdevice.h>
	24	#include <linux/if_ether.h>
	25	#include <linux/if_vlan.h>
	26	#include <net/llc_pdu.h>
	27	#include <linux/kernel.h>
	28	#include <linux/jhash.h>
	29	#include <linux/jiffies.h>
	30	#include <linux/llc.h>
	31	#include <linux/module.h>
	32	#include <linux/in.h>
	33	#include <linux/rcupdate.h>
	34	#include <linux/if_arp.h>
ccb1352e JG	35	#include <linux/ip.h>
ccb1352e JG	36	#include <linux/ipv6.h>
a175a723	37	#include <linux/sctp.h>
ccb1352e JG	38	#include <linux/tcp.h>
	39	#include <linux/udp.h>
	40	#include <linux/icmp.h>
	41	#include <linux/icmpv6.h>
	42	#include <linux/rculist.h>
	43	#include <net/ip.h>
7d5437c7	44	#include <net/ip_tunnels.h>
ccb1352e JG	45	#include <net/ipv6.h>
	46	#include <net/ndisc.h>
	47
e6445719	48	u64 ovs_flow_used_time(unsigned long flow_jiffies)
03f0d916	49	{
e6445719 PS	50	struct timespec cur_ts;
e6445719 PS	51	u64 cur_ms, idle_ms;
03f0d916	52
e6445719 PS	53	ktime_get_ts(&cur_ts);
	54	idle_ms = jiffies_to_msecs(jiffies - flow_jiffies);
	55	cur_ms = (u64)cur_ts.tv_sec * MSEC_PER_SEC +
	56	cur_ts.tv_nsec / NSEC_PER_MSEC;
03f0d916	57
e6445719	58	return cur_ms - idle_ms;
5828cd9a AZ	59	}
5828cd9a AZ	60
df23e9f6	61	#define TCP_FLAGS_BE16(tp) ((__be16 )&tcp_flag_word(tp) & htons(0x0FFF))
03f0d916	62
e6445719	63	void ovs_flow_used(struct sw_flow flow, struct sk_buff skb)
03f0d916	64	{
df23e9f6	65	__be16 tcp_flags = 0;
03f0d916	66
e6445719 PS	67	if ((flow->key.eth.type == htons(ETH_P_IP) \|\|
	68	flow->key.eth.type == htons(ETH_P_IPV6)) &&
	69	flow->key.ip.proto == IPPROTO_TCP &&
	70	likely(skb->len >= skb_transport_offset(skb) + sizeof(struct tcphdr))) {
df23e9f6	71	tcp_flags = TCP_FLAGS_BE16(tcp_hdr(skb));
03f0d916 AZ	72	}
03f0d916 AZ	73
e6445719 PS	74	spin_lock(&flow->lock);
	75	flow->used = jiffies;
	76	flow->packet_count++;
	77	flow->byte_count += skb->len;
	78	flow->tcp_flags \|= tcp_flags;
	79	spin_unlock(&flow->lock);
03f0d916 AZ	80	}
03f0d916 AZ	81
ccb1352e JG	82	static int check_header(struct sk_buff *skb, int len)
	83	{
	84	if (unlikely(skb->len < len))
	85	return -EINVAL;
	86	if (unlikely(!pskb_may_pull(skb, len)))
	87	return -ENOMEM;
	88	return 0;
	89	}
	90
	91	static bool arphdr_ok(struct sk_buff *skb)
	92	{
	93	return pskb_may_pull(skb, skb_network_offset(skb) +
	94	sizeof(struct arp_eth_header));
	95	}
	96
	97	static int check_iphdr(struct sk_buff *skb)
	98	{
	99	unsigned int nh_ofs = skb_network_offset(skb);
	100	unsigned int ip_len;
	101	int err;
	102
	103	err = check_header(skb, nh_ofs + sizeof(struct iphdr));
	104	if (unlikely(err))
	105	return err;
	106
	107	ip_len = ip_hdrlen(skb);
	108	if (unlikely(ip_len < sizeof(struct iphdr) \|\|
	109	skb->len < nh_ofs + ip_len))
	110	return -EINVAL;
	111
	112	skb_set_transport_header(skb, nh_ofs + ip_len);
	113	return 0;
	114	}
	115
	116	static bool tcphdr_ok(struct sk_buff *skb)
	117	{
	118	int th_ofs = skb_transport_offset(skb);
	119	int tcp_len;
	120
	121	if (unlikely(!pskb_may_pull(skb, th_ofs + sizeof(struct tcphdr))))
	122	return false;
	123
	124	tcp_len = tcp_hdrlen(skb);
	125	if (unlikely(tcp_len < sizeof(struct tcphdr) \|\|
	126	skb->len < th_ofs + tcp_len))
	127	return false;
	128
	129	return true;
	130	}
	131
	132	static bool udphdr_ok(struct sk_buff *skb)
	133	{
	134	return pskb_may_pull(skb, skb_transport_offset(skb) +
	135	sizeof(struct udphdr));
	136	}
	137
a175a723 JS	138	static bool sctphdr_ok(struct sk_buff *skb)
	139	{
	140	return pskb_may_pull(skb, skb_transport_offset(skb) +
	141	sizeof(struct sctphdr));
	142	}
	143
ccb1352e JG	144	static bool icmphdr_ok(struct sk_buff *skb)
	145	{
	146	return pskb_may_pull(skb, skb_transport_offset(skb) +
	147	sizeof(struct icmphdr));
	148	}
	149
03f0d916	150	static int parse_ipv6hdr(struct sk_buff skb, struct sw_flow_key key)
ccb1352e JG	151	{
	152	unsigned int nh_ofs = skb_network_offset(skb);
	153	unsigned int nh_len;
	154	int payload_ofs;
	155	struct ipv6hdr *nh;
	156	uint8_t nexthdr;
	157	__be16 frag_off;
	158	int err;
	159
ccb1352e JG	160	err = check_header(skb, nh_ofs + sizeof(*nh));
	161	if (unlikely(err))
	162	return err;
	163
	164	nh = ipv6_hdr(skb);
	165	nexthdr = nh->nexthdr;
	166	payload_ofs = (u8 *)(nh + 1) - skb->data;
	167
	168	key->ip.proto = NEXTHDR_NONE;
	169	key->ip.tos = ipv6_get_dsfield(nh);
	170	key->ip.ttl = nh->hop_limit;
	171	key->ipv6.label = (__be32 )nh & htonl(IPV6_FLOWINFO_FLOWLABEL);
	172	key->ipv6.addr.src = nh->saddr;
	173	key->ipv6.addr.dst = nh->daddr;
	174
	175	payload_ofs = ipv6_skip_exthdr(skb, payload_ofs, &nexthdr, &frag_off);
	176	if (unlikely(payload_ofs < 0))
	177	return -EINVAL;
	178
	179	if (frag_off) {
	180	if (frag_off & htons(~0x7))
	181	key->ip.frag = OVS_FRAG_TYPE_LATER;
	182	else
	183	key->ip.frag = OVS_FRAG_TYPE_FIRST;
	184	}
	185
	186	nh_len = payload_ofs - nh_ofs;
	187	skb_set_transport_header(skb, nh_ofs + nh_len);
	188	key->ip.proto = nexthdr;
	189	return nh_len;
	190	}
	191
	192	static bool icmp6hdr_ok(struct sk_buff *skb)
	193	{
	194	return pskb_may_pull(skb, skb_transport_offset(skb) +
	195	sizeof(struct icmp6hdr));
	196	}
	197
ccb1352e JG	198	static int parse_vlan(struct sk_buff skb, struct sw_flow_key key)
	199	{
	200	struct qtag_prefix {
	201	__be16 eth_type; /* ETH_P_8021Q */
	202	__be16 tci;
	203	};
	204	struct qtag_prefix *qp;
	205
	206	if (unlikely(skb->len < sizeof(struct qtag_prefix) + sizeof(__be16)))
	207	return 0;
	208
	209	if (unlikely(!pskb_may_pull(skb, sizeof(struct qtag_prefix) +
	210	sizeof(__be16))))
	211	return -ENOMEM;
	212
	213	qp = (struct qtag_prefix *) skb->data;
	214	key->eth.tci = qp->tci \| htons(VLAN_TAG_PRESENT);
	215	__skb_pull(skb, sizeof(struct qtag_prefix));
	216
	217	return 0;
	218	}
	219
	220	static __be16 parse_ethertype(struct sk_buff *skb)
	221	{
	222	struct llc_snap_hdr {
	223	u8 dsap; /* Always 0xAA */
	224	u8 ssap; /* Always 0xAA */
	225	u8 ctrl;
	226	u8 oui[3];
	227	__be16 ethertype;
	228	};
	229	struct llc_snap_hdr *llc;
	230	__be16 proto;
	231
	232	proto = (__be16 ) skb->data;
	233	__skb_pull(skb, sizeof(__be16));
	234
e5c5d22e	235	if (ntohs(proto) >= ETH_P_802_3_MIN)
ccb1352e JG	236	return proto;
	237
	238	if (skb->len < sizeof(struct llc_snap_hdr))
	239	return htons(ETH_P_802_2);
	240
	241	if (unlikely(!pskb_may_pull(skb, sizeof(struct llc_snap_hdr))))
	242	return htons(0);
	243
	244	llc = (struct llc_snap_hdr *) skb->data;
	245	if (llc->dsap != LLC_SAP_SNAP \|\|
	246	llc->ssap != LLC_SAP_SNAP \|\|
	247	(llc->oui[0] \| llc->oui[1] \| llc->oui[2]) != 0)
	248	return htons(ETH_P_802_2);
	249
	250	__skb_pull(skb, sizeof(struct llc_snap_hdr));
17b682a0	251
e5c5d22e	252	if (ntohs(llc->ethertype) >= ETH_P_802_3_MIN)
17b682a0 RL	253	return llc->ethertype;
	254
	255	return htons(ETH_P_802_2);
ccb1352e JG	256	}
	257
	258	static int parse_icmpv6(struct sk_buff skb, struct sw_flow_key key,
03f0d916	259	int nh_len)
ccb1352e JG	260	{
ccb1352e JG	261	struct icmp6hdr *icmp = icmp6_hdr(skb);
ccb1352e JG	262
	263	/* The ICMPv6 type and code fields use the 16-bit transport port
	264	* fields, so we need to store them in 16-bit network byte order.
	265	*/
	266	key->ipv6.tp.src = htons(icmp->icmp6_type);
	267	key->ipv6.tp.dst = htons(icmp->icmp6_code);
ccb1352e JG	268
	269	if (icmp->icmp6_code == 0 &&
	270	(icmp->icmp6_type == NDISC_NEIGHBOUR_SOLICITATION \|\|
	271	icmp->icmp6_type == NDISC_NEIGHBOUR_ADVERTISEMENT)) {
	272	int icmp_len = skb->len - skb_transport_offset(skb);
	273	struct nd_msg *nd;
	274	int offset;
	275
ccb1352e JG	276	/* In order to process neighbor discovery options, we need the
	277	* entire packet.
	278	*/
	279	if (unlikely(icmp_len < sizeof(*nd)))
03f0d916 AZ	280	return 0;
	281
	282	if (unlikely(skb_linearize(skb)))
	283	return -ENOMEM;
ccb1352e JG	284
	285	nd = (struct nd_msg *)skb_transport_header(skb);
	286	key->ipv6.nd.target = nd->target;
ccb1352e JG	287
	288	icmp_len -= sizeof(*nd);
	289	offset = 0;
	290	while (icmp_len >= 8) {
	291	struct nd_opt_hdr *nd_opt =
	292	(struct nd_opt_hdr *)(nd->opt + offset);
	293	int opt_len = nd_opt->nd_opt_len * 8;
	294
	295	if (unlikely(!opt_len \|\| opt_len > icmp_len))
03f0d916	296	return 0;
ccb1352e JG	297
	298	/* Store the link layer address if the appropriate
	299	* option is provided. It is considered an error if
	300	* the same link layer option is specified twice.
	301	*/
	302	if (nd_opt->nd_opt_type == ND_OPT_SOURCE_LL_ADDR
	303	&& opt_len == 8) {
	304	if (unlikely(!is_zero_ether_addr(key->ipv6.nd.sll)))
	305	goto invalid;
	306	memcpy(key->ipv6.nd.sll,
	307	&nd->opt[offset+sizeof(*nd_opt)], ETH_ALEN);
	308	} else if (nd_opt->nd_opt_type == ND_OPT_TARGET_LL_ADDR
	309	&& opt_len == 8) {
	310	if (unlikely(!is_zero_ether_addr(key->ipv6.nd.tll)))
	311	goto invalid;
	312	memcpy(key->ipv6.nd.tll,
	313	&nd->opt[offset+sizeof(*nd_opt)], ETH_ALEN);
	314	}
	315
	316	icmp_len -= opt_len;
	317	offset += opt_len;
	318	}
	319	}
	320
03f0d916	321	return 0;
ccb1352e JG	322
	323	invalid:
	324	memset(&key->ipv6.nd.target, 0, sizeof(key->ipv6.nd.target));
	325	memset(key->ipv6.nd.sll, 0, sizeof(key->ipv6.nd.sll));
	326	memset(key->ipv6.nd.tll, 0, sizeof(key->ipv6.nd.tll));
	327
03f0d916	328	return 0;
ccb1352e JG	329	}
	330
	331	/**
	332	* ovs_flow_extract - extracts a flow key from an Ethernet frame.
	333	* @skb: sk_buff that contains the frame, with skb->data pointing to the
	334	* Ethernet header
	335	* @in_port: port number on which @skb was received.
	336	* @key: output flow key
ccb1352e JG	337	*
	338	* The caller must ensure that skb->len >= ETH_HLEN.
	339	*
	340	* Returns 0 if successful, otherwise a negative errno value.
	341	*
	342	* Initializes @skb header pointers as follows:
	343	*
	344	* - skb->mac_header: the Ethernet header.
	345	*
	346	* - skb->network_header: just past the Ethernet header, or just past the
	347	* VLAN header, to the first byte of the Ethernet payload.
	348	*
34d94f21	349	* - skb->transport_header: If key->eth.type is ETH_P_IP or ETH_P_IPV6
ccb1352e JG	350	* on output, then just past the IP header, if one is present and
ccb1352e JG	351	* of a correct length, otherwise the same as skb->network_header.
34d94f21	352	* For other key->eth.type values it is left untouched.
ccb1352e	353	*/
03f0d916	354	int ovs_flow_extract(struct sk_buff skb, u16 in_port, struct sw_flow_key key)
ccb1352e	355	{
03f0d916	356	int error;
ccb1352e JG	357	struct ethhdr *eth;
	358
	359	memset(key, 0, sizeof(*key));
	360
	361	key->phy.priority = skb->priority;
7d5437c7 PS	362	if (OVS_CB(skb)->tun_key)
7d5437c7 PS	363	memcpy(&key->tun_key, OVS_CB(skb)->tun_key, sizeof(key->tun_key));
ccb1352e	364	key->phy.in_port = in_port;
39c7caeb	365	key->phy.skb_mark = skb->mark;
ccb1352e JG	366
	367	skb_reset_mac_header(skb);
	368
	369	/* Link layer. We are guaranteed to have at least the 14 byte Ethernet
	370	* header in the linear data area.
	371	*/
	372	eth = eth_hdr(skb);
	373	memcpy(key->eth.src, eth->h_source, ETH_ALEN);
	374	memcpy(key->eth.dst, eth->h_dest, ETH_ALEN);
	375
	376	__skb_pull(skb, 2 * ETH_ALEN);
b34df5e8 PS	377	/* We are going to push all headers that we pull, so no need to
	378	* update skb->csum here.
	379	*/
ccb1352e JG	380
	381	if (vlan_tx_tag_present(skb))
	382	key->eth.tci = htons(skb->vlan_tci);
	383	else if (eth->h_proto == htons(ETH_P_8021Q))
	384	if (unlikely(parse_vlan(skb, key)))
	385	return -ENOMEM;
	386
	387	key->eth.type = parse_ethertype(skb);
	388	if (unlikely(key->eth.type == htons(0)))
	389	return -ENOMEM;
	390
	391	skb_reset_network_header(skb);
	392	__skb_push(skb, skb->data - skb_mac_header(skb));
	393
	394	/* Network layer. */
	395	if (key->eth.type == htons(ETH_P_IP)) {
	396	struct iphdr *nh;
	397	__be16 offset;
	398
ccb1352e JG	399	error = check_iphdr(skb);
	400	if (unlikely(error)) {
	401	if (error == -EINVAL) {
	402	skb->transport_header = skb->network_header;
	403	error = 0;
	404	}
03f0d916	405	return error;
ccb1352e JG	406	}
	407
	408	nh = ip_hdr(skb);
	409	key->ipv4.addr.src = nh->saddr;
	410	key->ipv4.addr.dst = nh->daddr;
	411
	412	key->ip.proto = nh->protocol;
	413	key->ip.tos = nh->tos;
	414	key->ip.ttl = nh->ttl;
	415
	416	offset = nh->frag_off & htons(IP_OFFSET);
	417	if (offset) {
	418	key->ip.frag = OVS_FRAG_TYPE_LATER;
03f0d916	419	return 0;
ccb1352e JG	420	}
	421	if (nh->frag_off & htons(IP_MF) \|\|
	422	skb_shinfo(skb)->gso_type & SKB_GSO_UDP)
	423	key->ip.frag = OVS_FRAG_TYPE_FIRST;
	424
	425	/* Transport layer. */
	426	if (key->ip.proto == IPPROTO_TCP) {
ccb1352e JG	427	if (tcphdr_ok(skb)) {
	428	struct tcphdr *tcp = tcp_hdr(skb);
	429	key->ipv4.tp.src = tcp->source;
	430	key->ipv4.tp.dst = tcp->dest;
5eb26b15	431	key->ipv4.tp.flags = TCP_FLAGS_BE16(tcp);
ccb1352e JG	432	}
ccb1352e JG	433	} else if (key->ip.proto == IPPROTO_UDP) {
ccb1352e JG	434	if (udphdr_ok(skb)) {
	435	struct udphdr *udp = udp_hdr(skb);
	436	key->ipv4.tp.src = udp->source;
	437	key->ipv4.tp.dst = udp->dest;
	438	}
a175a723 JS	439	} else if (key->ip.proto == IPPROTO_SCTP) {
	440	if (sctphdr_ok(skb)) {
	441	struct sctphdr *sctp = sctp_hdr(skb);
	442	key->ipv4.tp.src = sctp->source;
	443	key->ipv4.tp.dst = sctp->dest;
	444	}
ccb1352e	445	} else if (key->ip.proto == IPPROTO_ICMP) {
ccb1352e JG	446	if (icmphdr_ok(skb)) {
	447	struct icmphdr *icmp = icmp_hdr(skb);
	448	/* The ICMP type and code fields use the 16-bit
	449	* transport port fields, so we need to store
	450	* them in 16-bit network byte order. */
	451	key->ipv4.tp.src = htons(icmp->type);
	452	key->ipv4.tp.dst = htons(icmp->code);
	453	}
	454	}
	455
c0618533 MM	456	} else if ((key->eth.type == htons(ETH_P_ARP) \|\|
c0618533 MM	457	key->eth.type == htons(ETH_P_RARP)) && arphdr_ok(skb)) {
ccb1352e JG	458	struct arp_eth_header *arp;
	459
	460	arp = (struct arp_eth_header *)skb_network_header(skb);
	461
	462	if (arp->ar_hrd == htons(ARPHRD_ETHER)
	463	&& arp->ar_pro == htons(ETH_P_IP)
	464	&& arp->ar_hln == ETH_ALEN
	465	&& arp->ar_pln == 4) {
	466
	467	/* We only match on the lower 8 bits of the opcode. */
	468	if (ntohs(arp->ar_op) <= 0xff)
	469	key->ip.proto = ntohs(arp->ar_op);
d04d3829 MM	470	memcpy(&key->ipv4.addr.src, arp->ar_sip, sizeof(key->ipv4.addr.src));
	471	memcpy(&key->ipv4.addr.dst, arp->ar_tip, sizeof(key->ipv4.addr.dst));
	472	memcpy(key->ipv4.arp.sha, arp->ar_sha, ETH_ALEN);
	473	memcpy(key->ipv4.arp.tha, arp->ar_tha, ETH_ALEN);
ccb1352e JG	474	}
	475	} else if (key->eth.type == htons(ETH_P_IPV6)) {
	476	int nh_len; /* IPv6 Header + Extensions */
	477
03f0d916	478	nh_len = parse_ipv6hdr(skb, key);
ccb1352e	479	if (unlikely(nh_len < 0)) {
03f0d916	480	if (nh_len == -EINVAL) {
ccb1352e	481	skb->transport_header = skb->network_header;
03f0d916 AZ	482	error = 0;
03f0d916 AZ	483	} else {
ccb1352e	484	error = nh_len;
03f0d916 AZ	485	}
03f0d916 AZ	486	return error;
ccb1352e JG	487	}
	488
	489	if (key->ip.frag == OVS_FRAG_TYPE_LATER)
03f0d916	490	return 0;
ccb1352e JG	491	if (skb_shinfo(skb)->gso_type & SKB_GSO_UDP)
	492	key->ip.frag = OVS_FRAG_TYPE_FIRST;
	493
	494	/* Transport layer. */
	495	if (key->ip.proto == NEXTHDR_TCP) {
ccb1352e JG	496	if (tcphdr_ok(skb)) {
	497	struct tcphdr *tcp = tcp_hdr(skb);
	498	key->ipv6.tp.src = tcp->source;
	499	key->ipv6.tp.dst = tcp->dest;
5eb26b15	500	key->ipv6.tp.flags = TCP_FLAGS_BE16(tcp);
ccb1352e JG	501	}
ccb1352e JG	502	} else if (key->ip.proto == NEXTHDR_UDP) {
ccb1352e JG	503	if (udphdr_ok(skb)) {
	504	struct udphdr *udp = udp_hdr(skb);
	505	key->ipv6.tp.src = udp->source;
	506	key->ipv6.tp.dst = udp->dest;
	507	}
a175a723 JS	508	} else if (key->ip.proto == NEXTHDR_SCTP) {
	509	if (sctphdr_ok(skb)) {
	510	struct sctphdr *sctp = sctp_hdr(skb);
	511	key->ipv6.tp.src = sctp->source;
	512	key->ipv6.tp.dst = sctp->dest;
	513	}
ccb1352e	514	} else if (key->ip.proto == NEXTHDR_ICMP) {
ccb1352e	515	if (icmp6hdr_ok(skb)) {
03f0d916 AZ	516	error = parse_icmpv6(skb, key, nh_len);
	517	if (error)
	518	return error;
ccb1352e JG	519	}
	520	}
	521	}
	522
03f0d916	523	return 0;
ccb1352e	524	}