[mirror_ubuntu-eoan-kernel.git] / net / openvswitch / actions.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
 */

#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

#include <linux/skbuff.h>
#include <linux/in.h>
#include <linux/ip.h>
#include <linux/openvswitch.h>
#include <linux/sctp.h>
#include <linux/tcp.h>
#include <linux/udp.h>
#include <linux/in6.h>
#include <linux/if_arp.h>
#include <linux/if_vlan.h>
#include <net/ip.h>
#include <net/ipv6.h>
#include <net/checksum.h>
#include <net/dsfield.h>
#include <net/sctp/checksum.h>

#include "datapath.h"
#include "vport.h"

static int do_execute_actions(struct datapath *dp, struct sk_buff *skb,
			const struct nlattr *attr, int len, bool keep_skb);

static int make_writable(struct sk_buff *skb, int write_len)
{
	if (!skb_cloned(skb) || skb_clone_writable(skb, write_len))
		return 0;

	return pskb_expand_head(skb, 0, 0, GFP_ATOMIC);
}

/* remove VLAN header from packet and update csum accordingly. */
static int __pop_vlan_tci(struct sk_buff *skb, __be16 *current_tci)
{
	struct vlan_hdr *vhdr;
	int err;

	err = make_writable(skb, VLAN_ETH_HLEN);
	if (unlikely(err))
		return err;

	if (skb->ip_summed == CHECKSUM_COMPLETE)
		skb->csum = csum_sub(skb->csum, csum_partial(skb->data
					+ (2 * ETH_ALEN), VLAN_HLEN, 0));

	vhdr = (struct vlan_hdr *)(skb->data + ETH_HLEN);
	*current_tci = vhdr->h_vlan_TCI;

	memmove(skb->data + VLAN_HLEN, skb->data, 2 * ETH_ALEN);
	__skb_pull(skb, VLAN_HLEN);

	vlan_set_encap_proto(skb, vhdr);
	skb->mac_header += VLAN_HLEN;
	skb_reset_mac_len(skb);

	return 0;
}

static int pop_vlan(struct sk_buff *skb)
{
	__be16 tci;
	int err;

	if (likely(vlan_tx_tag_present(skb))) {
		skb->vlan_tci = 0;
	} else {
		if (unlikely(skb->protocol != htons(ETH_P_8021Q) ||
			     skb->len < VLAN_ETH_HLEN))
			return 0;

		err = __pop_vlan_tci(skb, &tci);
		if (err)
			return err;
	}
	/* move next vlan tag to hw accel tag */
	if (likely(skb->protocol != htons(ETH_P_8021Q) ||
		   skb->len < VLAN_ETH_HLEN))
		return 0;

	err = __pop_vlan_tci(skb, &tci);
	if (unlikely(err))
		return err;

	__vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q), ntohs(tci));
	return 0;
}

static int push_vlan(struct sk_buff *skb, const struct ovs_action_push_vlan *vlan)
{
	if (unlikely(vlan_tx_tag_present(skb))) {
		u16 current_tag;

		/* push down current VLAN tag */
		current_tag = vlan_tx_tag_get(skb);

		if (!__vlan_put_tag(skb, skb->vlan_proto, current_tag))
			return -ENOMEM;

		if (skb->ip_summed == CHECKSUM_COMPLETE)
			skb->csum = csum_add(skb->csum, csum_partial(skb->data
					+ (2 * ETH_ALEN), VLAN_HLEN, 0));

	}
	__vlan_hwaccel_put_tag(skb, vlan->vlan_tpid, ntohs(vlan->vlan_tci) & ~VLAN_TAG_PRESENT);
	return 0;
}

static int set_eth_addr(struct sk_buff *skb,
			const struct ovs_key_ethernet *eth_key)
{
	int err;
	err = make_writable(skb, ETH_HLEN);
	if (unlikely(err))
		return err;

	skb_postpull_rcsum(skb, eth_hdr(skb), ETH_ALEN * 2);

	memcpy(eth_hdr(skb)->h_source, eth_key->eth_src, ETH_ALEN);
	memcpy(eth_hdr(skb)->h_dest, eth_key->eth_dst, ETH_ALEN);

	ovs_skb_postpush_rcsum(skb, eth_hdr(skb), ETH_ALEN * 2);

	return 0;
}

static void set_ip_addr(struct sk_buff *skb, struct iphdr *nh,
				__be32 *addr, __be32 new_addr)
{
	int transport_len = skb->len - skb_transport_offset(skb);

	if (nh->protocol == IPPROTO_TCP) {
		if (likely(transport_len >= sizeof(struct tcphdr)))
			inet_proto_csum_replace4(&tcp_hdr(skb)->check, skb,
						 *addr, new_addr, 1);
	} else if (nh->protocol == IPPROTO_UDP) {
		if (likely(transport_len >= sizeof(struct udphdr))) {
			struct udphdr *uh = udp_hdr(skb);

			if (uh->check || skb->ip_summed == CHECKSUM_PARTIAL) {
				inet_proto_csum_replace4(&uh->check, skb,
							 *addr, new_addr, 1);
				if (!uh->check)
					uh->check = CSUM_MANGLED_0;
			}
		}
	}

	csum_replace4(&nh->check, *addr, new_addr);
	skb->rxhash = 0;
	*addr = new_addr;
}

static void update_ipv6_checksum(struct sk_buff *skb, u8 l4_proto,
				 __be32 addr[4], const __be32 new_addr[4])
{
	int transport_len = skb->len - skb_transport_offset(skb);

	if (l4_proto == IPPROTO_TCP) {
		if (likely(transport_len >= sizeof(struct tcphdr)))
			inet_proto_csum_replace16(&tcp_hdr(skb)->check, skb,
						  addr, new_addr, 1);
	} else if (l4_proto == IPPROTO_UDP) {
		if (likely(transport_len >= sizeof(struct udphdr))) {
			struct udphdr *uh = udp_hdr(skb);

			if (uh->check || skb->ip_summed == CHECKSUM_PARTIAL) {
				inet_proto_csum_replace16(&uh->check, skb,
							  addr, new_addr, 1);
				if (!uh->check)
					uh->check = CSUM_MANGLED_0;
			}
		}
	}
}

static void set_ipv6_addr(struct sk_buff *skb, u8 l4_proto,
			  __be32 addr[4], const __be32 new_addr[4],
			  bool recalculate_csum)
{
	if (recalculate_csum)
		update_ipv6_checksum(skb, l4_proto, addr, new_addr);

	skb->rxhash = 0;
	memcpy(addr, new_addr, sizeof(__be32[4]));
}

static void set_ipv6_tc(struct ipv6hdr *nh, u8 tc)
{
	nh->priority = tc >> 4;
	nh->flow_lbl[0] = (nh->flow_lbl[0] & 0x0F) | ((tc & 0x0F) << 4);
}

static void set_ipv6_fl(struct ipv6hdr *nh, u32 fl)
{
	nh->flow_lbl[0] = (nh->flow_lbl[0] & 0xF0) | (fl & 0x000F0000) >> 16;
	nh->flow_lbl[1] = (fl & 0x0000FF00) >> 8;
	nh->flow_lbl[2] = fl & 0x000000FF;
}

static void set_ip_ttl(struct sk_buff *skb, struct iphdr *nh, u8 new_ttl)
{
	csum_replace2(&nh->check, htons(nh->ttl << 8), htons(new_ttl << 8));
	nh->ttl = new_ttl;
}

static int set_ipv4(struct sk_buff *skb, const struct ovs_key_ipv4 *ipv4_key)
{
	struct iphdr *nh;
	int err;

	err = make_writable(skb, skb_network_offset(skb) +
				 sizeof(struct iphdr));
	if (unlikely(err))
		return err;

	nh = ip_hdr(skb);

	if (ipv4_key->ipv4_src != nh->saddr)
		set_ip_addr(skb, nh, &nh->saddr, ipv4_key->ipv4_src);

	if (ipv4_key->ipv4_dst != nh->daddr)
		set_ip_addr(skb, nh, &nh->daddr, ipv4_key->ipv4_dst);

	if (ipv4_key->ipv4_tos != nh->tos)
		ipv4_change_dsfield(nh, 0, ipv4_key->ipv4_tos);

	if (ipv4_key->ipv4_ttl != nh->ttl)
		set_ip_ttl(skb, nh, ipv4_key->ipv4_ttl);

	return 0;
}

static int set_ipv6(struct sk_buff *skb, const struct ovs_key_ipv6 *ipv6_key)
{
	struct ipv6hdr *nh;
	int err;
	__be32 *saddr;
	__be32 *daddr;

	err = make_writable(skb, skb_network_offset(skb) +
			    sizeof(struct ipv6hdr));
	if (unlikely(err))
		return err;

	nh = ipv6_hdr(skb);
	saddr = (__be32 *)&nh->saddr;
	daddr = (__be32 *)&nh->daddr;

	if (memcmp(ipv6_key->ipv6_src, saddr, sizeof(ipv6_key->ipv6_src)))
		set_ipv6_addr(skb, ipv6_key->ipv6_proto, saddr,
			      ipv6_key->ipv6_src, true);

	if (memcmp(ipv6_key->ipv6_dst, daddr, sizeof(ipv6_key->ipv6_dst))) {
		unsigned int offset = 0;
		int flags = IP6_FH_F_SKIP_RH;
		bool recalc_csum = true;

		if (ipv6_ext_hdr(nh->nexthdr))
			recalc_csum = ipv6_find_hdr(skb, &offset,
						    NEXTHDR_ROUTING, NULL,
						    &flags) != NEXTHDR_ROUTING;

		set_ipv6_addr(skb, ipv6_key->ipv6_proto, daddr,
			      ipv6_key->ipv6_dst, recalc_csum);
	}

	set_ipv6_tc(nh, ipv6_key->ipv6_tclass);
	set_ipv6_fl(nh, ntohl(ipv6_key->ipv6_label));
	nh->hop_limit = ipv6_key->ipv6_hlimit;

	return 0;
}

/* Must follow make_writable() since that can move the skb data. */
static void set_tp_port(struct sk_buff *skb, __be16 *port,
			 __be16 new_port, __sum16 *check)
{
	inet_proto_csum_replace2(check, skb, *port, new_port, 0);
	*port = new_port;
	skb->rxhash = 0;
}

static void set_udp_port(struct sk_buff *skb, __be16 *port, __be16 new_port)
{
	struct udphdr *uh = udp_hdr(skb);

	if (uh->check && skb->ip_summed != CHECKSUM_PARTIAL) {
		set_tp_port(skb, port, new_port, &uh->check);

		if (!uh->check)
			uh->check = CSUM_MANGLED_0;
	} else {
		*port = new_port;
		skb->rxhash = 0;
	}
}

static int set_udp(struct sk_buff *skb, const struct ovs_key_udp *udp_port_key)
{
	struct udphdr *uh;
	int err;

	err = make_writable(skb, skb_transport_offset(skb) +
				 sizeof(struct udphdr));
	if (unlikely(err))
		return err;

	uh = udp_hdr(skb);
	if (udp_port_key->udp_src != uh->source)
		set_udp_port(skb, &uh->source, udp_port_key->udp_src);

	if (udp_port_key->udp_dst != uh->dest)
		set_udp_port(skb, &uh->dest, udp_port_key->udp_dst);

	return 0;
}

static int set_tcp(struct sk_buff *skb, const struct ovs_key_tcp *tcp_port_key)
{
	struct tcphdr *th;
	int err;

	err = make_writable(skb, skb_transport_offset(skb) +
				 sizeof(struct tcphdr));
	if (unlikely(err))
		return err;

	th = tcp_hdr(skb);
	if (tcp_port_key->tcp_src != th->source)
		set_tp_port(skb, &th->source, tcp_port_key->tcp_src, &th->check);

	if (tcp_port_key->tcp_dst != th->dest)
		set_tp_port(skb, &th->dest, tcp_port_key->tcp_dst, &th->check);

	return 0;
}

static int set_sctp(struct sk_buff *skb,
		     const struct ovs_key_sctp *sctp_port_key)
{
	struct sctphdr *sh;
	int err;
	unsigned int sctphoff = skb_transport_offset(skb);

	err = make_writable(skb, sctphoff + sizeof(struct sctphdr));
	if (unlikely(err))
		return err;

	sh = sctp_hdr(skb);
	if (sctp_port_key->sctp_src != sh->source ||
	    sctp_port_key->sctp_dst != sh->dest) {
		__le32 old_correct_csum, new_csum, old_csum;

		old_csum = sh->checksum;
		old_correct_csum = sctp_compute_cksum(skb, sctphoff);

		sh->source = sctp_port_key->sctp_src;
		sh->dest = sctp_port_key->sctp_dst;

		new_csum = sctp_compute_cksum(skb, sctphoff);

		/* Carry any checksum errors through. */
		sh->checksum = old_csum ^ old_correct_csum ^ new_csum;

		skb->rxhash = 0;
	}

	return 0;
}

static int do_output(struct datapath *dp, struct sk_buff *skb, int out_port)
{
	struct vport *vport;

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

	vport = ovs_vport_rcu(dp, out_port);
	if (unlikely(!vport)) {
		kfree_skb(skb);
		return -ENODEV;
	}

	ovs_vport_send(vport, skb);
	return 0;
}

static int output_userspace(struct datapath *dp, struct sk_buff *skb,
			    const struct nlattr *attr)
{
	struct dp_upcall_info upcall;
	const struct nlattr *a;
	int rem;

	BUG_ON(!OVS_CB(skb)->pkt_key);

	upcall.cmd = OVS_PACKET_CMD_ACTION;
	upcall.key = OVS_CB(skb)->pkt_key;
	upcall.userdata = NULL;
	upcall.portid = 0;

	for (a = nla_data(attr), rem = nla_len(attr); rem > 0;
		 a = nla_next(a, &rem)) {
		switch (nla_type(a)) {
		case OVS_USERSPACE_ATTR_USERDATA:
			upcall.userdata = a;
			break;

		case OVS_USERSPACE_ATTR_PID:
			upcall.portid = nla_get_u32(a);
			break;
		}
	}

	return ovs_dp_upcall(dp, skb, &upcall);
}

static int sample(struct datapath *dp, struct sk_buff *skb,
		  const struct nlattr *attr)
{
	const struct nlattr *acts_list = NULL;
	const struct nlattr *a;
	int rem;

	for (a = nla_data(attr), rem = nla_len(attr); rem > 0;
		 a = nla_next(a, &rem)) {
		switch (nla_type(a)) {
		case OVS_SAMPLE_ATTR_PROBABILITY:
			if (net_random() >= nla_get_u32(a))
				return 0;
			break;

		case OVS_SAMPLE_ATTR_ACTIONS:
			acts_list = a;
			break;
		}
	}

	return do_execute_actions(dp, skb, nla_data(acts_list),
						 nla_len(acts_list), true);
}

static int execute_set_action(struct sk_buff *skb,
				 const struct nlattr *nested_attr)
{
	int err = 0;

	switch (nla_type(nested_attr)) {
	case OVS_KEY_ATTR_PRIORITY:
		skb->priority = nla_get_u32(nested_attr);
		break;

	case OVS_KEY_ATTR_SKB_MARK:
		skb->mark = nla_get_u32(nested_attr);
		break;

	case OVS_KEY_ATTR_IPV4_TUNNEL:
		OVS_CB(skb)->tun_key = nla_data(nested_attr);
		break;

	case OVS_KEY_ATTR_ETHERNET:
		err = set_eth_addr(skb, nla_data(nested_attr));
		break;

	case OVS_KEY_ATTR_IPV4:
		err = set_ipv4(skb, nla_data(nested_attr));
		break;

	case OVS_KEY_ATTR_IPV6:
		err = set_ipv6(skb, nla_data(nested_attr));
		break;

	case OVS_KEY_ATTR_TCP:
		err = set_tcp(skb, nla_data(nested_attr));
		break;

	case OVS_KEY_ATTR_UDP:
		err = set_udp(skb, nla_data(nested_attr));
		break;

	case OVS_KEY_ATTR_SCTP:
		err = set_sctp(skb, nla_data(nested_attr));
		break;
	}

	return err;
}

/* Execute a list of actions against 'skb'. */
static int do_execute_actions(struct datapath *dp, struct sk_buff *skb,
			const struct nlattr *attr, int len, bool keep_skb)
{
	/* Every output action needs a separate clone of 'skb', but the common
	 * case is just a single output action, so that doing a clone and
	 * then freeing the original skbuff is wasteful.  So the following code
	 * is slightly obscure just to avoid that. */
	int prev_port = -1;
	const struct nlattr *a;
	int rem;

	for (a = attr, rem = len; rem > 0;
	     a = nla_next(a, &rem)) {
		int err = 0;

		if (prev_port != -1) {
			do_output(dp, skb_clone(skb, GFP_ATOMIC), prev_port);
			prev_port = -1;
		}

		switch (nla_type(a)) {
		case OVS_ACTION_ATTR_OUTPUT:
			prev_port = nla_get_u32(a);
			break;

		case OVS_ACTION_ATTR_USERSPACE:
			output_userspace(dp, skb, a);
			break;

		case OVS_ACTION_ATTR_PUSH_VLAN:
			err = push_vlan(skb, nla_data(a));
			if (unlikely(err)) /* skb already freed. */
				return err;
			break;

		case OVS_ACTION_ATTR_POP_VLAN:
			err = pop_vlan(skb);
			break;

		case OVS_ACTION_ATTR_SET:
			err = execute_set_action(skb, nla_data(a));
			break;

		case OVS_ACTION_ATTR_SAMPLE:
			err = sample(dp, skb, a);
			break;
		}

		if (unlikely(err)) {
			kfree_skb(skb);
			return err;
		}
	}

	if (prev_port != -1) {
		if (keep_skb)
			skb = skb_clone(skb, GFP_ATOMIC);

		do_output(dp, skb, prev_port);
	} else if (!keep_skb)
		consume_skb(skb);

	return 0;
}

/* Execute a list of actions against 'skb'. */
int ovs_execute_actions(struct datapath *dp, struct sk_buff *skb)
{
	struct sw_flow_actions *acts = rcu_dereference(OVS_CB(skb)->flow->sf_acts);

	OVS_CB(skb)->tun_key = NULL;
	return do_execute_actions(dp, skb, acts->actions,
					 acts->actions_len, false);
}
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	#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
	20
	21	#include <linux/skbuff.h>
	22	#include <linux/in.h>
	23	#include <linux/ip.h>
	24	#include <linux/openvswitch.h>
a175a723	25	#include <linux/sctp.h>
ccb1352e JG	26	#include <linux/tcp.h>
	27	#include <linux/udp.h>
	28	#include <linux/in6.h>
	29	#include <linux/if_arp.h>
	30	#include <linux/if_vlan.h>
	31	#include <net/ip.h>
3fdbd1ce	32	#include <net/ipv6.h>
ccb1352e JG	33	#include <net/checksum.h>
ccb1352e JG	34	#include <net/dsfield.h>
a175a723	35	#include <net/sctp/checksum.h>
ccb1352e JG	36
	37	#include "datapath.h"
	38	#include "vport.h"
	39
	40	static int do_execute_actions(struct datapath dp, struct sk_buff skb,
	41	const struct nlattr *attr, int len, bool keep_skb);
	42
	43	static int make_writable(struct sk_buff *skb, int write_len)
	44	{
	45	if (!skb_cloned(skb) \|\| skb_clone_writable(skb, write_len))
	46	return 0;
	47
	48	return pskb_expand_head(skb, 0, 0, GFP_ATOMIC);
	49	}
	50
39855b5b	51	/* remove VLAN header from packet and update csum accordingly. */
ccb1352e JG	52	static int __pop_vlan_tci(struct sk_buff skb, __be16 current_tci)
	53	{
	54	struct vlan_hdr *vhdr;
	55	int err;
	56
	57	err = make_writable(skb, VLAN_ETH_HLEN);
	58	if (unlikely(err))
	59	return err;
	60
	61	if (skb->ip_summed == CHECKSUM_COMPLETE)
	62	skb->csum = csum_sub(skb->csum, csum_partial(skb->data
7b024082	63	+ (2 * ETH_ALEN), VLAN_HLEN, 0));
ccb1352e JG	64
	65	vhdr = (struct vlan_hdr *)(skb->data + ETH_HLEN);
	66	*current_tci = vhdr->h_vlan_TCI;
	67
	68	memmove(skb->data + VLAN_HLEN, skb->data, 2 * ETH_ALEN);
	69	__skb_pull(skb, VLAN_HLEN);
	70
	71	vlan_set_encap_proto(skb, vhdr);
	72	skb->mac_header += VLAN_HLEN;
	73	skb_reset_mac_len(skb);
	74
	75	return 0;
	76	}
	77
	78	static int pop_vlan(struct sk_buff *skb)
	79	{
	80	__be16 tci;
	81	int err;
	82
	83	if (likely(vlan_tx_tag_present(skb))) {
	84	skb->vlan_tci = 0;
	85	} else {
	86	if (unlikely(skb->protocol != htons(ETH_P_8021Q) \|\|
	87	skb->len < VLAN_ETH_HLEN))
	88	return 0;
	89
	90	err = __pop_vlan_tci(skb, &tci);
	91	if (err)
	92	return err;
	93	}
	94	/* move next vlan tag to hw accel tag */
	95	if (likely(skb->protocol != htons(ETH_P_8021Q) \|\|
	96	skb->len < VLAN_ETH_HLEN))
	97	return 0;
	98
	99	err = __pop_vlan_tci(skb, &tci);
	100	if (unlikely(err))
	101	return err;
	102
86a9bad3	103	__vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q), ntohs(tci));
ccb1352e JG	104	return 0;
	105	}
	106
	107	static int push_vlan(struct sk_buff skb, const struct ovs_action_push_vlan vlan)
	108	{
	109	if (unlikely(vlan_tx_tag_present(skb))) {
	110	u16 current_tag;
	111
	112	/* push down current VLAN tag */
	113	current_tag = vlan_tx_tag_get(skb);
	114
86a9bad3	115	if (!__vlan_put_tag(skb, skb->vlan_proto, current_tag))
ccb1352e JG	116	return -ENOMEM;
	117
	118	if (skb->ip_summed == CHECKSUM_COMPLETE)
	119	skb->csum = csum_add(skb->csum, csum_partial(skb->data
7b024082	120	+ (2 * ETH_ALEN), VLAN_HLEN, 0));
ccb1352e JG	121
ccb1352e JG	122	}
86a9bad3	123	__vlan_hwaccel_put_tag(skb, vlan->vlan_tpid, ntohs(vlan->vlan_tci) & ~VLAN_TAG_PRESENT);
ccb1352e JG	124	return 0;
	125	}
	126
	127	static int set_eth_addr(struct sk_buff *skb,
	128	const struct ovs_key_ethernet *eth_key)
	129	{
	130	int err;
	131	err = make_writable(skb, ETH_HLEN);
	132	if (unlikely(err))
	133	return err;
	134
b34df5e8 PS	135	skb_postpull_rcsum(skb, eth_hdr(skb), ETH_ALEN * 2);
b34df5e8 PS	136
ccb1352e JG	137	memcpy(eth_hdr(skb)->h_source, eth_key->eth_src, ETH_ALEN);
	138	memcpy(eth_hdr(skb)->h_dest, eth_key->eth_dst, ETH_ALEN);
	139
b34df5e8 PS	140	ovs_skb_postpush_rcsum(skb, eth_hdr(skb), ETH_ALEN * 2);
b34df5e8 PS	141
ccb1352e JG	142	return 0;
	143	}
	144
	145	static void set_ip_addr(struct sk_buff skb, struct iphdr nh,
	146	__be32 *addr, __be32 new_addr)
	147	{
	148	int transport_len = skb->len - skb_transport_offset(skb);
	149
	150	if (nh->protocol == IPPROTO_TCP) {
	151	if (likely(transport_len >= sizeof(struct tcphdr)))
	152	inet_proto_csum_replace4(&tcp_hdr(skb)->check, skb,
	153	*addr, new_addr, 1);
	154	} else if (nh->protocol == IPPROTO_UDP) {
81e5d41d JG	155	if (likely(transport_len >= sizeof(struct udphdr))) {
	156	struct udphdr *uh = udp_hdr(skb);
	157
	158	if (uh->check \|\| skb->ip_summed == CHECKSUM_PARTIAL) {
	159	inet_proto_csum_replace4(&uh->check, skb,
	160	*addr, new_addr, 1);
	161	if (!uh->check)
	162	uh->check = CSUM_MANGLED_0;
	163	}
	164	}
ccb1352e JG	165	}
	166
	167	csum_replace4(&nh->check, *addr, new_addr);
	168	skb->rxhash = 0;
	169	*addr = new_addr;
	170	}
	171
3fdbd1ce AA	172	static void update_ipv6_checksum(struct sk_buff *skb, u8 l4_proto,
	173	__be32 addr[4], const __be32 new_addr[4])
	174	{
	175	int transport_len = skb->len - skb_transport_offset(skb);
	176
	177	if (l4_proto == IPPROTO_TCP) {
	178	if (likely(transport_len >= sizeof(struct tcphdr)))
	179	inet_proto_csum_replace16(&tcp_hdr(skb)->check, skb,
	180	addr, new_addr, 1);
	181	} else if (l4_proto == IPPROTO_UDP) {
	182	if (likely(transport_len >= sizeof(struct udphdr))) {
	183	struct udphdr *uh = udp_hdr(skb);
	184
	185	if (uh->check \|\| skb->ip_summed == CHECKSUM_PARTIAL) {
	186	inet_proto_csum_replace16(&uh->check, skb,
	187	addr, new_addr, 1);
	188	if (!uh->check)
	189	uh->check = CSUM_MANGLED_0;
	190	}
	191	}
	192	}
	193	}
	194
	195	static void set_ipv6_addr(struct sk_buff *skb, u8 l4_proto,
	196	__be32 addr[4], const __be32 new_addr[4],
	197	bool recalculate_csum)
	198	{
	199	if (recalculate_csum)
	200	update_ipv6_checksum(skb, l4_proto, addr, new_addr);
	201
	202	skb->rxhash = 0;
	203	memcpy(addr, new_addr, sizeof(__be32[4]));
	204	}
	205
	206	static void set_ipv6_tc(struct ipv6hdr *nh, u8 tc)
	207	{
	208	nh->priority = tc >> 4;
	209	nh->flow_lbl[0] = (nh->flow_lbl[0] & 0x0F) \| ((tc & 0x0F) << 4);
	210	}
	211
	212	static void set_ipv6_fl(struct ipv6hdr *nh, u32 fl)
	213	{
	214	nh->flow_lbl[0] = (nh->flow_lbl[0] & 0xF0) \| (fl & 0x000F0000) >> 16;
	215	nh->flow_lbl[1] = (fl & 0x0000FF00) >> 8;
	216	nh->flow_lbl[2] = fl & 0x000000FF;
	217	}
	218
ccb1352e JG	219	static void set_ip_ttl(struct sk_buff skb, struct iphdr nh, u8 new_ttl)
	220	{
	221	csum_replace2(&nh->check, htons(nh->ttl << 8), htons(new_ttl << 8));
	222	nh->ttl = new_ttl;
	223	}
	224
	225	static int set_ipv4(struct sk_buff skb, const struct ovs_key_ipv4 ipv4_key)
	226	{
	227	struct iphdr *nh;
	228	int err;
	229
	230	err = make_writable(skb, skb_network_offset(skb) +
	231	sizeof(struct iphdr));
	232	if (unlikely(err))
	233	return err;
	234
	235	nh = ip_hdr(skb);
	236
	237	if (ipv4_key->ipv4_src != nh->saddr)
	238	set_ip_addr(skb, nh, &nh->saddr, ipv4_key->ipv4_src);
	239
	240	if (ipv4_key->ipv4_dst != nh->daddr)
	241	set_ip_addr(skb, nh, &nh->daddr, ipv4_key->ipv4_dst);
	242
	243	if (ipv4_key->ipv4_tos != nh->tos)
	244	ipv4_change_dsfield(nh, 0, ipv4_key->ipv4_tos);
	245
	246	if (ipv4_key->ipv4_ttl != nh->ttl)
	247	set_ip_ttl(skb, nh, ipv4_key->ipv4_ttl);
	248
	249	return 0;
	250	}
	251
3fdbd1ce AA	252	static int set_ipv6(struct sk_buff skb, const struct ovs_key_ipv6 ipv6_key)
	253	{
	254	struct ipv6hdr *nh;
	255	int err;
	256	__be32 *saddr;
	257	__be32 *daddr;
	258
	259	err = make_writable(skb, skb_network_offset(skb) +
	260	sizeof(struct ipv6hdr));
	261	if (unlikely(err))
	262	return err;
	263
	264	nh = ipv6_hdr(skb);
	265	saddr = (__be32 *)&nh->saddr;
	266	daddr = (__be32 *)&nh->daddr;
	267
	268	if (memcmp(ipv6_key->ipv6_src, saddr, sizeof(ipv6_key->ipv6_src)))
	269	set_ipv6_addr(skb, ipv6_key->ipv6_proto, saddr,
	270	ipv6_key->ipv6_src, true);
	271
	272	if (memcmp(ipv6_key->ipv6_dst, daddr, sizeof(ipv6_key->ipv6_dst))) {
	273	unsigned int offset = 0;
	274	int flags = IP6_FH_F_SKIP_RH;
	275	bool recalc_csum = true;
	276
	277	if (ipv6_ext_hdr(nh->nexthdr))
	278	recalc_csum = ipv6_find_hdr(skb, &offset,
	279	NEXTHDR_ROUTING, NULL,
	280	&flags) != NEXTHDR_ROUTING;
	281
	282	set_ipv6_addr(skb, ipv6_key->ipv6_proto, daddr,
	283	ipv6_key->ipv6_dst, recalc_csum);
	284	}
	285
	286	set_ipv6_tc(nh, ipv6_key->ipv6_tclass);
	287	set_ipv6_fl(nh, ntohl(ipv6_key->ipv6_label));
	288	nh->hop_limit = ipv6_key->ipv6_hlimit;
	289
	290	return 0;
	291	}
	292
ccb1352e JG	293	/* Must follow make_writable() since that can move the skb data. */
	294	static void set_tp_port(struct sk_buff skb, __be16 port,
	295	__be16 new_port, __sum16 *check)
	296	{
	297	inet_proto_csum_replace2(check, skb, *port, new_port, 0);
	298	*port = new_port;
	299	skb->rxhash = 0;
	300	}
	301
81e5d41d JG	302	static void set_udp_port(struct sk_buff skb, __be16 port, __be16 new_port)
	303	{
	304	struct udphdr *uh = udp_hdr(skb);
	305
	306	if (uh->check && skb->ip_summed != CHECKSUM_PARTIAL) {
	307	set_tp_port(skb, port, new_port, &uh->check);
	308
	309	if (!uh->check)
	310	uh->check = CSUM_MANGLED_0;
	311	} else {
	312	*port = new_port;
	313	skb->rxhash = 0;
	314	}
	315	}
	316
	317	static int set_udp(struct sk_buff skb, const struct ovs_key_udp udp_port_key)
ccb1352e JG	318	{
	319	struct udphdr *uh;
	320	int err;
	321
	322	err = make_writable(skb, skb_transport_offset(skb) +
	323	sizeof(struct udphdr));
	324	if (unlikely(err))
	325	return err;
	326
	327	uh = udp_hdr(skb);
	328	if (udp_port_key->udp_src != uh->source)
81e5d41d	329	set_udp_port(skb, &uh->source, udp_port_key->udp_src);
ccb1352e JG	330
ccb1352e JG	331	if (udp_port_key->udp_dst != uh->dest)
81e5d41d	332	set_udp_port(skb, &uh->dest, udp_port_key->udp_dst);
ccb1352e JG	333
	334	return 0;
	335	}
	336
81e5d41d	337	static int set_tcp(struct sk_buff skb, const struct ovs_key_tcp tcp_port_key)
ccb1352e JG	338	{
	339	struct tcphdr *th;
	340	int err;
	341
	342	err = make_writable(skb, skb_transport_offset(skb) +
	343	sizeof(struct tcphdr));
	344	if (unlikely(err))
	345	return err;
	346
	347	th = tcp_hdr(skb);
	348	if (tcp_port_key->tcp_src != th->source)
	349	set_tp_port(skb, &th->source, tcp_port_key->tcp_src, &th->check);
	350
	351	if (tcp_port_key->tcp_dst != th->dest)
	352	set_tp_port(skb, &th->dest, tcp_port_key->tcp_dst, &th->check);
	353
	354	return 0;
	355	}
	356
a175a723 JS	357	static int set_sctp(struct sk_buff *skb,
	358	const struct ovs_key_sctp *sctp_port_key)
	359	{
	360	struct sctphdr *sh;
	361	int err;
	362	unsigned int sctphoff = skb_transport_offset(skb);
	363
	364	err = make_writable(skb, sctphoff + sizeof(struct sctphdr));
	365	if (unlikely(err))
	366	return err;
	367
	368	sh = sctp_hdr(skb);
	369	if (sctp_port_key->sctp_src != sh->source \|\|
	370	sctp_port_key->sctp_dst != sh->dest) {
	371	__le32 old_correct_csum, new_csum, old_csum;
	372
	373	old_csum = sh->checksum;
	374	old_correct_csum = sctp_compute_cksum(skb, sctphoff);
	375
	376	sh->source = sctp_port_key->sctp_src;
	377	sh->dest = sctp_port_key->sctp_dst;
	378
	379	new_csum = sctp_compute_cksum(skb, sctphoff);
	380
	381	/* Carry any checksum errors through. */
	382	sh->checksum = old_csum ^ old_correct_csum ^ new_csum;
	383
	384	skb->rxhash = 0;
	385	}
	386
	387	return 0;
	388	}
	389
ccb1352e JG	390	static int do_output(struct datapath dp, struct sk_buff skb, int out_port)
	391	{
	392	struct vport *vport;
	393
	394	if (unlikely(!skb))
	395	return -ENOMEM;
	396
15eac2a7	397	vport = ovs_vport_rcu(dp, out_port);
ccb1352e JG	398	if (unlikely(!vport)) {
	399	kfree_skb(skb);
	400	return -ENODEV;
	401	}
	402
	403	ovs_vport_send(vport, skb);
	404	return 0;
	405	}
	406
	407	static int output_userspace(struct datapath dp, struct sk_buff skb,
	408	const struct nlattr *attr)
	409	{
	410	struct dp_upcall_info upcall;
	411	const struct nlattr *a;
	412	int rem;
	413
03f0d916 AZ	414	BUG_ON(!OVS_CB(skb)->pkt_key);
03f0d916 AZ	415
ccb1352e	416	upcall.cmd = OVS_PACKET_CMD_ACTION;
03f0d916	417	upcall.key = OVS_CB(skb)->pkt_key;
ccb1352e	418	upcall.userdata = NULL;
15e47304	419	upcall.portid = 0;
ccb1352e JG	420
	421	for (a = nla_data(attr), rem = nla_len(attr); rem > 0;
	422	a = nla_next(a, &rem)) {
	423	switch (nla_type(a)) {
	424	case OVS_USERSPACE_ATTR_USERDATA:
	425	upcall.userdata = a;
	426	break;
	427
	428	case OVS_USERSPACE_ATTR_PID:
15e47304	429	upcall.portid = nla_get_u32(a);
ccb1352e JG	430	break;
	431	}
	432	}
	433
	434	return ovs_dp_upcall(dp, skb, &upcall);
	435	}
	436
	437	static int sample(struct datapath dp, struct sk_buff skb,
	438	const struct nlattr *attr)
	439	{
	440	const struct nlattr *acts_list = NULL;
	441	const struct nlattr *a;
	442	int rem;
	443
	444	for (a = nla_data(attr), rem = nla_len(attr); rem > 0;
	445	a = nla_next(a, &rem)) {
	446	switch (nla_type(a)) {
	447	case OVS_SAMPLE_ATTR_PROBABILITY:
	448	if (net_random() >= nla_get_u32(a))
	449	return 0;
	450	break;
	451
	452	case OVS_SAMPLE_ATTR_ACTIONS:
	453	acts_list = a;
	454	break;
	455	}
	456	}
	457
	458	return do_execute_actions(dp, skb, nla_data(acts_list),
	459	nla_len(acts_list), true);
	460	}
	461
	462	static int execute_set_action(struct sk_buff *skb,
	463	const struct nlattr *nested_attr)
	464	{
	465	int err = 0;
	466
	467	switch (nla_type(nested_attr)) {
	468	case OVS_KEY_ATTR_PRIORITY:
	469	skb->priority = nla_get_u32(nested_attr);
	470	break;
	471
39c7caeb AA	472	case OVS_KEY_ATTR_SKB_MARK:
	473	skb->mark = nla_get_u32(nested_attr);
	474	break;
	475
7d5437c7 PS	476	case OVS_KEY_ATTR_IPV4_TUNNEL:
	477	OVS_CB(skb)->tun_key = nla_data(nested_attr);
	478	break;
	479
ccb1352e JG	480	case OVS_KEY_ATTR_ETHERNET:
	481	err = set_eth_addr(skb, nla_data(nested_attr));
	482	break;
	483
	484	case OVS_KEY_ATTR_IPV4:
	485	err = set_ipv4(skb, nla_data(nested_attr));
	486	break;
	487
3fdbd1ce AA	488	case OVS_KEY_ATTR_IPV6:
	489	err = set_ipv6(skb, nla_data(nested_attr));
	490	break;
	491
ccb1352e	492	case OVS_KEY_ATTR_TCP:
81e5d41d	493	err = set_tcp(skb, nla_data(nested_attr));
ccb1352e JG	494	break;
	495
	496	case OVS_KEY_ATTR_UDP:
81e5d41d	497	err = set_udp(skb, nla_data(nested_attr));
ccb1352e	498	break;
a175a723 JS	499
	500	case OVS_KEY_ATTR_SCTP:
	501	err = set_sctp(skb, nla_data(nested_attr));
	502	break;
ccb1352e JG	503	}
	504
	505	return err;
	506	}
	507
	508	/* Execute a list of actions against 'skb'. */
	509	static int do_execute_actions(struct datapath dp, struct sk_buff skb,
	510	const struct nlattr *attr, int len, bool keep_skb)
	511	{
	512	/* Every output action needs a separate clone of 'skb', but the common
	513	* case is just a single output action, so that doing a clone and
	514	* then freeing the original skbuff is wasteful. So the following code
	515	* is slightly obscure just to avoid that. */
	516	int prev_port = -1;
	517	const struct nlattr *a;
	518	int rem;
	519
	520	for (a = attr, rem = len; rem > 0;
	521	a = nla_next(a, &rem)) {
	522	int err = 0;
	523
	524	if (prev_port != -1) {
	525	do_output(dp, skb_clone(skb, GFP_ATOMIC), prev_port);
	526	prev_port = -1;
	527	}
	528
	529	switch (nla_type(a)) {
	530	case OVS_ACTION_ATTR_OUTPUT:
	531	prev_port = nla_get_u32(a);
	532	break;
	533
	534	case OVS_ACTION_ATTR_USERSPACE:
	535	output_userspace(dp, skb, a);
	536	break;
	537
	538	case OVS_ACTION_ATTR_PUSH_VLAN:
	539	err = push_vlan(skb, nla_data(a));
	540	if (unlikely(err)) /* skb already freed. */
	541	return err;
	542	break;
	543
	544	case OVS_ACTION_ATTR_POP_VLAN:
	545	err = pop_vlan(skb);
	546	break;
	547
	548	case OVS_ACTION_ATTR_SET:
	549	err = execute_set_action(skb, nla_data(a));
	550	break;
	551
	552	case OVS_ACTION_ATTR_SAMPLE:
	553	err = sample(dp, skb, a);
	554	break;
	555	}
	556
	557	if (unlikely(err)) {
	558	kfree_skb(skb);
	559	return err;
	560	}
	561	}
	562
	563	if (prev_port != -1) {
	564	if (keep_skb)
	565	skb = skb_clone(skb, GFP_ATOMIC);
	566
567	do_output(dp, skb, prev_port);
568	} else if (!keep_skb)
569	consume_skb(skb);
570
571	return 0;
572	}
573
574	/* Execute a list of actions against 'skb'. */
575	int ovs_execute_actions(struct datapath dp, struct sk_buff skb)
576	{
577	struct sw_flow_actions *acts = rcu_dereference(OVS_CB(skb)->flow->sf_acts);
578
36bf5cc6	579	OVS_CB(skb)->tun_key = NULL;
ccb1352e JG	580	return do_execute_actions(dp, skb, acts->actions,
	581	acts->actions_len, false);
	582	}