/*
- * Copyright (c) 2007-2011 Nicira Networks.
+ * Copyright (c) 2007-2011 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
#include <linux/in.h>
#include <linux/rcupdate.h>
#include <linux/if_arp.h>
-#include <linux/if_ether.h>
#include <linux/ip.h>
#include <linux/ipv6.h>
#include <linux/tcp.h>
#include "vlan.h"
static struct kmem_cache *flow_cache;
-static unsigned int hash_seed __read_mostly;
static int check_header(struct sk_buff *skb, int len)
{
(offsetof(struct sw_flow_key, field) + \
FIELD_SIZEOF(struct sw_flow_key, field))
-/**
- * skip_exthdr - skip any IPv6 extension headers
- * @skb: skbuff to parse
- * @start: offset of first extension header
- * @nexthdrp: Initially, points to the type of the extension header at @start.
- * This function updates it to point to the extension header at the final
- * offset.
- * @frag: Points to the @frag member in a &struct sw_flow_key. This
- * function sets an appropriate %OVS_FRAG_TYPE_* value.
- *
- * This is based on ipv6_skip_exthdr() but adds the updates to *@frag.
- *
- * When there is more than one fragment header, this version reports whether
- * the final fragment header that it examines is a first fragment.
- *
- * Returns the final payload offset, or -1 on error.
- */
-static int skip_exthdr(const struct sk_buff *skb, int start, u8 *nexthdrp,
- u8 *frag)
-{
- u8 nexthdr = *nexthdrp;
-
- while (ipv6_ext_hdr(nexthdr)) {
- struct ipv6_opt_hdr _hdr, *hp;
- int hdrlen;
-
- if (nexthdr == NEXTHDR_NONE)
- return -1;
- hp = skb_header_pointer(skb, start, sizeof(_hdr), &_hdr);
- if (hp == NULL)
- return -1;
- if (nexthdr == NEXTHDR_FRAGMENT) {
- __be16 _frag_off, *fp;
- fp = skb_header_pointer(skb,
- start+offsetof(struct frag_hdr,
- frag_off),
- sizeof(_frag_off),
- &_frag_off);
- if (fp == NULL)
- return -1;
-
- if (ntohs(*fp) & ~0x7) {
- *frag = OVS_FRAG_TYPE_LATER;
- break;
- }
- *frag = OVS_FRAG_TYPE_FIRST;
- hdrlen = 8;
- } else if (nexthdr == NEXTHDR_AUTH)
- hdrlen = (hp->hdrlen+2)<<2;
- else
- hdrlen = ipv6_optlen(hp);
-
- nexthdr = hp->nexthdr;
- start += hdrlen;
- }
-
- *nexthdrp = nexthdr;
- return start;
-}
-
static int parse_ipv6hdr(struct sk_buff *skb, struct sw_flow_key *key,
int *key_lenp)
{
int payload_ofs;
struct ipv6hdr *nh;
uint8_t nexthdr;
+ __be16 frag_off;
int err;
*key_lenp = SW_FLOW_KEY_OFFSET(ipv6.label);
key->ip.tos = ipv6_get_dsfield(nh);
key->ip.ttl = nh->hop_limit;
key->ipv6.label = *(__be32 *)nh & htonl(IPV6_FLOWINFO_FLOWLABEL);
- ipv6_addr_copy(&key->ipv6.addr.src, &nh->saddr);
- ipv6_addr_copy(&key->ipv6.addr.dst, &nh->daddr);
+ key->ipv6.addr.src = nh->saddr;
+ key->ipv6.addr.dst = nh->daddr;
- payload_ofs = skip_exthdr(skb, payload_ofs, &nexthdr, &key->ip.frag);
+ 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;
{
u8 tcp_flags = 0;
- if (flow->key.eth.type == htons(ETH_P_IP) &&
- flow->key.ip.proto == IPPROTO_TCP) {
+ 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))) {
u8 *tcp = (u8 *)tcp_hdr(skb);
tcp_flags = *(tcp + TCP_FLAGS_OFFSET) & TCP_FLAG_MASK;
}
int actions_len = nla_len(actions);
struct sw_flow_actions *sfa;
- /* At least DP_MAX_PORTS actions are required to be able to flood a
- * packet to every port. Factor of 2 allows for setting VLAN tags,
- * etc. */
- if (actions_len > 2 * DP_MAX_PORTS * nla_total_size(4))
+ if (actions_len > MAX_ACTIONS_BUFSIZE)
return ERR_PTR(-EINVAL);
sfa = kmalloc(sizeof(*sfa) + actions_len, GFP_KERNEL);
static struct hlist_head *find_bucket(struct flow_table *table, u32 hash)
{
+ hash = jhash_1word(hash, table->hash_seed);
return flex_array_get(table->buckets,
(hash & (table->n_buckets - 1)));
}
}
table->n_buckets = new_size;
table->count = 0;
+ table->node_ver = 0;
+ table->keep_flows = false;
+ get_random_bytes(&table->hash_seed, sizeof(u32));
return table;
}
if (!table)
return;
+ if (table->keep_flows)
+ goto skip_flows;
+
for (i = 0; i < table->n_buckets; i++) {
struct sw_flow *flow;
struct hlist_head *head = flex_array_get(table->buckets, i);
struct hlist_node *node, *n;
+ int ver = table->node_ver;
- hlist_for_each_entry_safe(flow, node, n, head, hash_node) {
- hlist_del_init_rcu(&flow->hash_node);
+ hlist_for_each_entry_safe(flow, node, n, head, hash_node[ver]) {
+ hlist_del_rcu(&flow->hash_node[ver]);
flow_free(flow);
}
}
+skip_flows:
free_buckets(table->buckets);
kfree(table);
}
struct sw_flow *flow;
struct hlist_head *head;
struct hlist_node *n;
+ int ver;
int i;
+ ver = table->node_ver;
while (*bucket < table->n_buckets) {
i = 0;
head = flex_array_get(table->buckets, *bucket);
- hlist_for_each_entry_rcu(flow, n, head, hash_node) {
+ hlist_for_each_entry_rcu(flow, n, head, hash_node[ver]) {
if (i < *last) {
i++;
continue;
return NULL;
}
-struct flow_table *ovs_flow_tbl_expand(struct flow_table *table)
+static void flow_table_copy_flows(struct flow_table *old, struct flow_table *new)
{
- struct flow_table *new_table;
- int n_buckets = table->n_buckets * 2;
+ int old_ver;
int i;
- new_table = ovs_flow_tbl_alloc(n_buckets);
- if (!new_table)
- return ERR_PTR(-ENOMEM);
+ old_ver = old->node_ver;
+ new->node_ver = !old_ver;
- for (i = 0; i < table->n_buckets; i++) {
+ /* Insert in new table. */
+ for (i = 0; i < old->n_buckets; i++) {
struct sw_flow *flow;
struct hlist_head *head;
- struct hlist_node *n, *pos;
+ struct hlist_node *n;
- head = flex_array_get(table->buckets, i);
+ head = flex_array_get(old->buckets, i);
- hlist_for_each_entry_safe(flow, n, pos, head, hash_node) {
- hlist_del_init_rcu(&flow->hash_node);
- ovs_flow_tbl_insert(new_table, flow);
- }
+ hlist_for_each_entry(flow, n, head, hash_node[old_ver])
+ ovs_flow_tbl_insert(new, flow);
}
+ old->keep_flows = true;
+}
+
+static struct flow_table *__flow_tbl_rehash(struct flow_table *table, int n_buckets)
+{
+ struct flow_table *new_table;
+
+ new_table = ovs_flow_tbl_alloc(n_buckets);
+ if (!new_table)
+ return ERR_PTR(-ENOMEM);
+
+ flow_table_copy_flows(table, new_table);
return new_table;
}
+struct flow_table *ovs_flow_tbl_rehash(struct flow_table *table)
+{
+ return __flow_tbl_rehash(table, table->n_buckets);
+}
+
+struct flow_table *ovs_flow_tbl_expand(struct flow_table *table)
+{
+ return __flow_tbl_rehash(table, table->n_buckets * 2);
+}
+
/* RCU callback used by ovs_flow_deferred_free. */
static void rcu_free_flow_callback(struct rcu_head *rcu)
{
}
nd = (struct nd_msg *)skb_transport_header(skb);
- ipv6_addr_copy(&key->ipv6.nd.target, &nd->target);
+ key->ipv6.nd.target = nd->target;
key_len = SW_FLOW_KEY_OFFSET(ipv6.nd);
icmp_len -= sizeof(*nd);
u32 ovs_flow_hash(const struct sw_flow_key *key, int key_len)
{
- return jhash2((u32 *)key, DIV_ROUND_UP(key_len, sizeof(u32)), hash_seed);
+ return jhash2((u32 *)key, DIV_ROUND_UP(key_len, sizeof(u32)), 0);
}
struct sw_flow *ovs_flow_tbl_lookup(struct flow_table *table,
hash = ovs_flow_hash(key, key_len);
head = find_bucket(table, hash);
- hlist_for_each_entry_rcu(flow, n, head, hash_node) {
+ hlist_for_each_entry_rcu(flow, n, head, hash_node[table->node_ver]) {
if (flow->hash == hash &&
!memcmp(&flow->key, key, key_len)) {
struct hlist_head *head;
head = find_bucket(table, flow->hash);
- hlist_add_head_rcu(&flow->hash_node, head);
+ hlist_add_head_rcu(&flow->hash_node[table->node_ver], head);
table->count++;
}
void ovs_flow_tbl_remove(struct flow_table *table, struct sw_flow *flow)
{
- if (!hlist_unhashed(&flow->hash_node)) {
- hlist_del_init_rcu(&flow->hash_node);
- table->count--;
- BUG_ON(table->count < 0);
- }
+ hlist_del_rcu(&flow->hash_node[table->node_ver]);
+ table->count--;
+ BUG_ON(table->count < 0);
}
/* The size of the argument for each %OVS_KEY_ATTR_* Netlink attribute. */
swkey->phy.in_port = in_port;
attrs &= ~(1 << OVS_KEY_ATTR_IN_PORT);
} else {
- swkey->phy.in_port = USHRT_MAX;
+ swkey->phy.in_port = DP_MAX_PORTS;
}
if (attrs & (1ULL << OVS_KEY_ATTR_TUN_ID)) {
const struct nlattr *nla;
int rem;
- *in_port = USHRT_MAX;
+ *in_port = DP_MAX_PORTS;
*tun_id = 0;
*priority = 0;
struct ovs_key_ethernet *eth_key;
struct nlattr *nla, *encap;
- if (swkey->phy.priority)
- NLA_PUT_U32(skb, OVS_KEY_ATTR_PRIORITY, swkey->phy.priority);
+ if (swkey->phy.priority &&
+ nla_put_u32(skb, OVS_KEY_ATTR_PRIORITY, swkey->phy.priority))
+ goto nla_put_failure;
- if (swkey->phy.tun_id != cpu_to_be64(0))
- NLA_PUT_BE64(skb, OVS_KEY_ATTR_TUN_ID, swkey->phy.tun_id);
+ if (swkey->phy.tun_id != cpu_to_be64(0) &&
+ nla_put_be64(skb, OVS_KEY_ATTR_TUN_ID, swkey->phy.tun_id))
+ goto nla_put_failure;
- if (swkey->phy.in_port != USHRT_MAX)
- NLA_PUT_U32(skb, OVS_KEY_ATTR_IN_PORT, swkey->phy.in_port);
+ if (swkey->phy.in_port != DP_MAX_PORTS &&
+ nla_put_u32(skb, OVS_KEY_ATTR_IN_PORT, swkey->phy.in_port))
+ goto nla_put_failure;
nla = nla_reserve(skb, OVS_KEY_ATTR_ETHERNET, sizeof(*eth_key));
if (!nla)
memcpy(eth_key->eth_dst, swkey->eth.dst, ETH_ALEN);
if (swkey->eth.tci || swkey->eth.type == htons(ETH_P_8021Q)) {
- NLA_PUT_BE16(skb, OVS_KEY_ATTR_ETHERTYPE, htons(ETH_P_8021Q));
- NLA_PUT_BE16(skb, OVS_KEY_ATTR_VLAN, swkey->eth.tci);
+ if (nla_put_be16(skb, OVS_KEY_ATTR_ETHERTYPE, htons(ETH_P_8021Q)) ||
+ nla_put_be16(skb, OVS_KEY_ATTR_VLAN, swkey->eth.tci))
+ goto nla_put_failure;
encap = nla_nest_start(skb, OVS_KEY_ATTR_ENCAP);
if (!swkey->eth.tci)
goto unencap;
if (swkey->eth.type == htons(ETH_P_802_2))
goto unencap;
- NLA_PUT_BE16(skb, OVS_KEY_ATTR_ETHERTYPE, swkey->eth.type);
+ if (nla_put_be16(skb, OVS_KEY_ATTR_ETHERTYPE, swkey->eth.type))
+ goto nla_put_failure;
if (swkey->eth.type == htons(ETH_P_IP)) {
struct ovs_key_ipv4 *ipv4_key;
if (flow_cache == NULL)
return -ENOMEM;
- get_random_bytes(&hash_seed, sizeof(hash_seed));
-
return 0;
}