Add support for vendor-defined and variable-length actions.
[sliver-openvswitch.git] / datapath / flow.c
index d4c4802..9dcec90 100644 (file)
@@ -11,6 +11,7 @@
 #include <linux/if_vlan.h>
 #include <net/llc_pdu.h>
 #include <linux/ip.h>
+#include <linux/jiffies.h>
 #include <linux/kernel.h>
 #include <linux/module.h>
 #include <linux/tcp.h>
@@ -28,46 +29,66 @@ struct kmem_cache *flow_cache;
 /* Internal function used to compare fields in flow. */
 static inline
 int flow_fields_match(const struct sw_flow_key *a, const struct sw_flow_key *b,
-               uint16_t w)
+                     uint32_t w, uint32_t src_mask, uint32_t dst_mask)
 {
        return ((w & OFPFW_IN_PORT || a->in_port == b->in_port)
                && (w & OFPFW_DL_VLAN || a->dl_vlan == b->dl_vlan)
                && (w & OFPFW_DL_SRC || !memcmp(a->dl_src, b->dl_src, ETH_ALEN))
                && (w & OFPFW_DL_DST || !memcmp(a->dl_dst, b->dl_dst, ETH_ALEN))
                && (w & OFPFW_DL_TYPE || a->dl_type == b->dl_type)
-               && (w & OFPFW_NW_SRC || a->nw_src == b->nw_src)
-               && (w & OFPFW_NW_DST || a->nw_dst == b->nw_dst)
+               && !((a->nw_src ^ b->nw_src) & src_mask)
+               && !((a->nw_dst ^ b->nw_dst) & dst_mask)
                && (w & OFPFW_NW_PROTO || a->nw_proto == b->nw_proto)
                && (w & OFPFW_TP_SRC || a->tp_src == b->tp_src)
                && (w & OFPFW_TP_DST || a->tp_dst == b->tp_dst));
 }
 
 /* Returns nonzero if 'a' and 'b' match, that is, if their fields are equal
- * modulo wildcards, zero otherwise. */
-int flow_matches(const struct sw_flow_key *a, const struct sw_flow_key *b)
+ * modulo wildcards in 'b', zero otherwise. */
+int flow_matches_1wild(const struct sw_flow_key *a,
+                      const struct sw_flow_key *b)
 {
-       return flow_fields_match(a, b, (a->wildcards | b->wildcards));
+       return flow_fields_match(a, b, b->wildcards,
+                                b->nw_src_mask, b->nw_dst_mask);
 }
-EXPORT_SYMBOL(flow_matches);
+EXPORT_SYMBOL(flow_matches_1wild);
 
-/* Returns nonzero if 't' (the table entry's key) and 'd' (the key 
- * describing the deletion) match, that is, if their fields are 
+/* Returns nonzero if 'a' and 'b' match, that is, if their fields are equal
+ * modulo wildcards in 'a' or 'b', zero otherwise. */
+int flow_matches_2wild(const struct sw_flow_key *a,
+                      const struct sw_flow_key *b)
+{
+       return flow_fields_match(a, b,
+                                a->wildcards | b->wildcards,
+                                a->nw_src_mask & b->nw_src_mask,
+                                a->nw_dst_mask & b->nw_dst_mask);
+}
+EXPORT_SYMBOL(flow_matches_2wild);
+
+/* Returns nonzero if 't' (the table entry's key) and 'd' (the key
+ * describing the match) match, that is, if their fields are
  * equal modulo wildcards, zero otherwise.  If 'strict' is nonzero, the
  * wildcards must match in both 't_key' and 'd_key'.  Note that the
  * table's wildcards are ignored unless 'strict' is set. */
-int flow_del_matches(const struct sw_flow_key *t, const struct sw_flow_key *d, int strict)
+int flow_matches_desc(const struct sw_flow_key *t, const struct sw_flow_key *d, 
+               int strict)
 {
-       if (strict && (t->wildcards != d->wildcards))
+       if (strict && d->wildcards != t->wildcards)
                return 0;
+       return flow_matches_1wild(t, d);
+}
+EXPORT_SYMBOL(flow_matches_desc);
 
-       return flow_fields_match(t, d, d->wildcards);
+static uint32_t make_nw_mask(int n_wild_bits)
+{
+       n_wild_bits &= (1u << OFPFW_NW_SRC_BITS) - 1;
+       return n_wild_bits < 32 ? htonl(~((1u << n_wild_bits) - 1)) : 0;
 }
-EXPORT_SYMBOL(flow_del_matches);
 
 void flow_extract_match(struct sw_flow_key* to, const struct ofp_match* from)
 {
-       to->wildcards = ntohs(from->wildcards) & OFPFW_ALL;
-       memset(to->pad, '\0', sizeof(to->pad));
+       to->wildcards = ntohl(from->wildcards) & OFPFW_ALL;
+       to->pad = 0;
        to->in_port = from->in_port;
        to->dl_vlan = from->dl_vlan;
        memcpy(to->dl_src, from->dl_src, ETH_ALEN);
@@ -78,7 +99,7 @@ void flow_extract_match(struct sw_flow_key* to, const struct ofp_match* from)
        to->tp_src = to->tp_dst = 0;
 
 #define OFPFW_TP (OFPFW_TP_SRC | OFPFW_TP_DST)
-#define OFPFW_NW (OFPFW_NW_SRC | OFPFW_NW_DST | OFPFW_NW_PROTO)
+#define OFPFW_NW (OFPFW_NW_SRC_MASK | OFPFW_NW_DST_MASK | OFPFW_NW_PROTO)
        if (to->wildcards & OFPFW_DL_TYPE) {
                /* Can't sensibly match on network or transport headers if the
                 * data link type is unknown. */
@@ -108,39 +129,61 @@ void flow_extract_match(struct sw_flow_key* to, const struct ofp_match* from)
                 * instead of falling into table-linear. */
                to->wildcards &= ~(OFPFW_NW | OFPFW_TP);
        }
+
+       /* We set these late because code above adjusts to->wildcards. */
+       to->nw_src_mask = make_nw_mask(to->wildcards >> OFPFW_NW_SRC_SHIFT);
+       to->nw_dst_mask = make_nw_mask(to->wildcards >> OFPFW_NW_DST_SHIFT);
 }
 
 void flow_fill_match(struct ofp_match* to, const struct sw_flow_key* from)
 {
-       to->wildcards = htons(from->wildcards);
+       to->wildcards = htonl(from->wildcards);
        to->in_port   = from->in_port;
        to->dl_vlan   = from->dl_vlan;
        memcpy(to->dl_src, from->dl_src, ETH_ALEN);
        memcpy(to->dl_dst, from->dl_dst, ETH_ALEN);
        to->dl_type   = from->dl_type;
-       to->nw_src        = from->nw_src;
-       to->nw_dst        = from->nw_dst;
+       to->nw_src    = from->nw_src;
+       to->nw_dst    = from->nw_dst;
        to->nw_proto  = from->nw_proto;
-       to->tp_src        = from->tp_src;
-       to->tp_dst        = from->tp_dst;
-       memset(to->pad, '\0', sizeof(to->pad));
+       to->tp_src    = from->tp_src;
+       to->tp_dst    = from->tp_dst;
+       to->pad       = 0;
+}
+
+int flow_timeout(struct sw_flow *flow)
+{
+       if (flow->idle_timeout != OFP_FLOW_PERMANENT
+           && time_after(jiffies, flow->used + flow->idle_timeout * HZ))
+               return OFPER_IDLE_TIMEOUT;
+       else if (flow->hard_timeout != OFP_FLOW_PERMANENT
+                && time_after(jiffies,
+                              flow->init_time + flow->hard_timeout * HZ))
+               return OFPER_HARD_TIMEOUT;
+       else
+               return -1;
 }
+EXPORT_SYMBOL(flow_timeout);
 
-/* Allocates and returns a new flow with 'n_actions' action, using allocation
- * flags 'flags'.  Returns the new flow or a null pointer on failure. */
-struct sw_flow *flow_alloc(int n_actions, gfp_t flags)
+/* Allocates and returns a new flow with room for 'actions_len' actions, 
+ * using allocation flags 'flags'.  Returns the new flow or a null pointer 
+ * on failure. */
+struct sw_flow *flow_alloc(size_t actions_len, gfp_t flags)
 {
+       struct sw_flow_actions *sfa;
+       size_t size = sizeof *sfa + actions_len;
        struct sw_flow *flow = kmem_cache_alloc(flow_cache, flags);
        if (unlikely(!flow))
                return NULL;
 
-       flow->n_actions = n_actions;
-       flow->actions = kmalloc(n_actions * sizeof *flow->actions,
-                               flags);
-       if (unlikely(!flow->actions) && n_actions > 0) {
+       sfa = kmalloc(size, flags);
+       if (unlikely(!sfa)) {
                kmem_cache_free(flow_cache, flow);
                return NULL;
        }
+       sfa->actions_len = actions_len;
+       flow->sf_acts = sfa;
+
        return flow;
 }
 
@@ -149,13 +192,13 @@ void flow_free(struct sw_flow *flow)
 {
        if (unlikely(!flow))
                return;
-       kfree(flow->actions);
+       kfree(flow->sf_acts);
        kmem_cache_free(flow_cache, flow);
 }
 EXPORT_SYMBOL(flow_free);
 
 /* RCU callback used by flow_deferred_free. */
-static void rcu_callback(struct rcu_head *rcu)
+static void rcu_free_flow_callback(struct rcu_head *rcu)
 {
        struct sw_flow *flow = container_of(rcu, struct sw_flow, rcu);
        flow_free(flow);
@@ -165,14 +208,53 @@ static void rcu_callback(struct rcu_head *rcu)
  * The caller must hold rcu_read_lock for this to be sensible. */
 void flow_deferred_free(struct sw_flow *flow)
 {
-       call_rcu(&flow->rcu, rcu_callback);
+       call_rcu(&flow->rcu, rcu_free_flow_callback);
 }
 EXPORT_SYMBOL(flow_deferred_free);
 
+/* RCU callback used by flow_deferred_free_acts. */
+static void rcu_free_acts_callback(struct rcu_head *rcu)
+{
+       struct sw_flow_actions *sf_acts = container_of(rcu, 
+                       struct sw_flow_actions, rcu);
+       kfree(sf_acts);
+}
+
+/* Schedules 'sf_acts' to be freed after the next RCU grace period.
+ * The caller must hold rcu_read_lock for this to be sensible. */
+void flow_deferred_free_acts(struct sw_flow_actions *sf_acts)
+{
+       call_rcu(&sf_acts->rcu, rcu_free_acts_callback);
+}
+EXPORT_SYMBOL(flow_deferred_free_acts);
+
+/* Copies 'actions' into a newly allocated structure for use by 'flow'
+ * and safely frees the structure that defined the previous actions. */
+void flow_replace_acts(struct sw_flow *flow, 
+               const struct ofp_action_header *actions, size_t actions_len)
+{
+       struct sw_flow_actions *sfa;
+       struct sw_flow_actions *orig_sfa = flow->sf_acts;
+       size_t size = sizeof *sfa + actions_len;
+
+       sfa = kmalloc(size, GFP_ATOMIC);
+       if (unlikely(!sfa))
+               return;
+
+       sfa->actions_len = actions_len;
+       memcpy(sfa->actions, actions, actions_len);
+
+       rcu_assign_pointer(flow->sf_acts, sfa);
+       flow_deferred_free_acts(orig_sfa);
+
+       return;
+}
+EXPORT_SYMBOL(flow_replace_acts);
+
 /* Prints a representation of 'key' to the kernel log. */
 void print_flow(const struct sw_flow_key *key)
 {
-       printk("wild%04x port%04x:vlan%04x mac%02x:%02x:%02x:%02x:%02x:%02x"
+       printk("wild%08x port%04x:vlan%04x mac%02x:%02x:%02x:%02x:%02x:%02x"
                        "->%02x:%02x:%02x:%02x:%02x:%02x "
                        "proto%04x ip%u.%u.%u.%u->%u.%u.%u.%u port%d->%d\n",
                        key->wildcards, ntohs(key->in_port), ntohs(key->dl_vlan),
@@ -193,18 +275,38 @@ void print_flow(const struct sw_flow_key *key)
 }
 EXPORT_SYMBOL(print_flow);
 
+static int tcphdr_ok(struct sk_buff *skb)
+{
+       int th_ofs = skb_transport_offset(skb);
+       if (skb->len >= th_ofs + sizeof(struct tcphdr)) {
+               int tcp_len = tcp_hdrlen(skb);
+               return (tcp_len >= sizeof(struct tcphdr)
+                       && skb->len >= th_ofs + tcp_len);
+       }
+       return 0;
+}
+
+static int udphdr_ok(struct sk_buff *skb)
+{
+       int th_ofs = skb_transport_offset(skb);
+       return skb->len >= th_ofs + sizeof(struct udphdr);
+}
+
 /* Parses the Ethernet frame in 'skb', which was received on 'in_port',
- * and initializes 'key' to match. */
-void flow_extract(struct sk_buff *skb, uint16_t in_port,
-                 struct sw_flow_key *key)
+ * and initializes 'key' to match.  Returns 1 if 'skb' contains an IP
+ * fragment, 0 otherwise. */
+int flow_extract(struct sk_buff *skb, uint16_t in_port,
+                struct sw_flow_key *key)
 {
        struct ethhdr *mac;
-       struct udphdr *th;
        int nh_ofs, th_ofs;
+       int retval = 0;
 
        key->in_port = htons(in_port);
+       key->pad = 0;
        key->wildcards = 0;
-       memset(key->pad, '\0', sizeof(key->pad));
+       key->nw_src_mask = 0;
+       key->nw_dst_mask = 0;
 
        /* This code doesn't check that skb->len is long enough to contain the
         * MAC or network header.  With a 46-byte minimum length frame this
@@ -223,7 +325,7 @@ void flow_extract(struct sk_buff *skb, uint16_t in_port,
                if (snap_get_ethertype(skb, &key->dl_type) != -EINVAL) {
                        nh_ofs += sizeof(struct snap_hdr);
                } else {
-                       key->dl_type = OFP_DL_TYPE_NOT_ETH_TYPE;
+                       key->dl_type = htons(OFP_DL_TYPE_NOT_ETH_TYPE);
                        nh_ofs += sizeof(struct llc_pdu_un);
                }
        }
@@ -251,25 +353,50 @@ void flow_extract(struct sk_buff *skb, uint16_t in_port,
                skb_set_transport_header(skb, th_ofs);
 
                /* Transport layer. */
-               if ((key->nw_proto != IPPROTO_TCP && key->nw_proto != IPPROTO_UDP)
-                   || skb->len < th_ofs + sizeof(struct udphdr)
-                   || nh->frag_off & htons(IP_MF | IP_OFFSET)) {
+               if (!(nh->frag_off & htons(IP_MF | IP_OFFSET))) {
+                       if (key->nw_proto == IPPROTO_TCP) {
+                               if (tcphdr_ok(skb)) {
+                                       struct tcphdr *tcp = tcp_hdr(skb);
+                                       key->tp_src = tcp->source;
+                                       key->tp_dst = tcp->dest;
+                               } else {
+                                       /* Avoid tricking other code into
+                                        * thinking that this packet has an L4
+                                        * header. */
+                                       goto no_proto;
+                               }
+                       } else if (key->nw_proto == IPPROTO_UDP) {
+                               if (udphdr_ok(skb)) {
+                                       struct udphdr *udp = udp_hdr(skb);
+                                       key->tp_src = udp->source;
+                                       key->tp_dst = udp->dest;
+                               } else {
+                                       /* Avoid tricking other code into
+                                        * thinking that this packet has an L4
+                                        * header. */
+                                       goto no_proto;
+                               }
+                       } else {
+                               goto no_th;
+                       }
+               } else {
+                       retval = 1;
                        goto no_th;
                }
-               th = udp_hdr(skb);
-               key->tp_src = th->source;
-               key->tp_dst = th->dest;
 
-               return;
+               return 0;
        }
 
        key->nw_src = 0;
        key->nw_dst = 0;
+
+no_proto:
        key->nw_proto = 0;
 
 no_th:
        key->tp_src = 0;
        key->tp_dst = 0;
+       return retval;
 }
 
 /* Initializes the flow module.