datapath: Get packet metadata from userspace in odp_packet_cmd_execute().

Until now, the tun_id and in_port have been lost when a packet is sent from
the kernel to userspace and then back to the kernel.  I didn't think that
this was a problem, but recent behavior made me look closer and see that
it makes a difference if sFlow is turned on or if an
ODP_ATTR_ACTION_CONTROLLER action is present.  We could possibly kluge
around those, but for future-proofing it seems better to pass the packet
metadata from userspace to the kernel.  That is what this commit does.

This commit introduces a user-kernel protocol break.  We could avoid that,
if it is desirable, by making ODP_PACKET_ATTR_KEY optional for
ODP_PACKET_CMD_EXECUTE commands.

Signed-off-by: Ben Pfaff <blp@nicira.com>
Acked-by: Jesse Gross <jesse@nicira.com>

datapath/datapath.c

@@ -655,7 +655,8 @@ static int odp_packet_cmd_execute(struct sk_buff *skb, struct genl_info *info)
 	int key_len;
 
 	err = -EINVAL;
-	if (!a[ODP_PACKET_ATTR_PACKET] || !a[ODP_PACKET_ATTR_ACTIONS] ||
+	if (!a[ODP_PACKET_ATTR_PACKET] || !a[ODP_PACKET_ATTR_KEY] ||
+	    !a[ODP_PACKET_ATTR_ACTIONS] ||
 	    nla_len(a[ODP_PACKET_ATTR_PACKET]) < ETH_HLEN)
 		goto err;
 
@@ -694,6 +695,12 @@ static int odp_packet_cmd_execute(struct sk_buff *skb, struct genl_info *info)
 		goto err_flow_put;
 	flow->tbl_node.hash = flow_hash(&flow->key, key_len);
 
+	err = flow_metadata_from_nlattrs(&flow->key.eth.in_port,
+					 &flow->key.eth.tun_id,
+					 a[ODP_PACKET_ATTR_KEY]);
+	if (err)
+		goto err_flow_put;
+
 	acts = flow_actions_alloc(a[ODP_PACKET_ATTR_ACTIONS]);
 	err = PTR_ERR(acts);
 	if (IS_ERR(acts))
@@ -725,6 +732,7 @@ err:
 
 static const struct nla_policy packet_policy[ODP_PACKET_ATTR_MAX + 1] = {
 	[ODP_PACKET_ATTR_PACKET] = { .type = NLA_UNSPEC },
+	[ODP_PACKET_ATTR_KEY] = { .type = NLA_NESTED },
 	[ODP_PACKET_ATTR_ACTIONS] = { .type = NLA_NESTED },
 };
 

datapath/flow.c

@@ -597,6 +597,23 @@ int flow_cmp(const struct tbl_node *node, void *key2_, int len)
 	return !memcmp(key1, key2, len);
 }
 
+/* The size of the argument for each %ODP_KEY_ATTR_* Netlink attribute.  */
+static const u32 key_lens[ODP_KEY_ATTR_MAX + 1] = {
+	[ODP_KEY_ATTR_TUN_ID] = 8,
+	[ODP_KEY_ATTR_IN_PORT] = 4,
+	[ODP_KEY_ATTR_ETHERNET] = sizeof(struct odp_key_ethernet),
+	[ODP_KEY_ATTR_8021Q] = sizeof(struct odp_key_8021q),
+	[ODP_KEY_ATTR_ETHERTYPE] = 2,
+	[ODP_KEY_ATTR_IPV4] = sizeof(struct odp_key_ipv4),
+	[ODP_KEY_ATTR_IPV6] = sizeof(struct odp_key_ipv6),
+	[ODP_KEY_ATTR_TCP] = sizeof(struct odp_key_tcp),
+	[ODP_KEY_ATTR_UDP] = sizeof(struct odp_key_udp),
+	[ODP_KEY_ATTR_ICMP] = sizeof(struct odp_key_icmp),
+	[ODP_KEY_ATTR_ICMPV6] = sizeof(struct odp_key_icmpv6),
+	[ODP_KEY_ATTR_ARP] = sizeof(struct odp_key_arp),
+	[ODP_KEY_ATTR_ND] = sizeof(struct odp_key_nd),
+};
+
 /**
  * flow_from_nlattrs - parses Netlink attributes into a flow key.
  * @swkey: receives the extracted flow key.
@@ -625,22 +642,6 @@ int flow_from_nlattrs(struct sw_flow_key *swkey, int *key_lenp,
 
 	prev_type = ODP_KEY_ATTR_UNSPEC;
 	nla_for_each_nested(nla, attr, rem) {
-		static const u32 key_lens[ODP_KEY_ATTR_MAX + 1] = {
-			[ODP_KEY_ATTR_TUN_ID] = 8,
-			[ODP_KEY_ATTR_IN_PORT] = 4,
-			[ODP_KEY_ATTR_ETHERNET] = sizeof(struct odp_key_ethernet),
-			[ODP_KEY_ATTR_8021Q] = sizeof(struct odp_key_8021q),
-			[ODP_KEY_ATTR_ETHERTYPE] = 2,
-			[ODP_KEY_ATTR_IPV4] = sizeof(struct odp_key_ipv4),
-			[ODP_KEY_ATTR_IPV6] = sizeof(struct odp_key_ipv6),
-			[ODP_KEY_ATTR_TCP] = sizeof(struct odp_key_tcp),
-			[ODP_KEY_ATTR_UDP] = sizeof(struct odp_key_udp),
-			[ODP_KEY_ATTR_ICMP] = sizeof(struct odp_key_icmp),
-			[ODP_KEY_ATTR_ICMPV6] = sizeof(struct odp_key_icmpv6),
-			[ODP_KEY_ATTR_ARP] = sizeof(struct odp_key_arp),
-			[ODP_KEY_ATTR_ND] = sizeof(struct odp_key_nd),
-		};
-
 		const struct odp_key_ethernet *eth_key;
 		const struct odp_key_8021q *q_key;
 		const struct odp_key_ipv4 *ipv4_key;
@@ -868,6 +869,62 @@ ok:
 	return error;
 }
 
+/**
+ * flow_metadata_from_nlattrs - parses Netlink attributes into a flow key.
+ * @in_port: receives the extracted input port.
+ * @tun_id: receives the extracted tunnel ID.
+ * @key: Netlink attribute holding nested %ODP_KEY_ATTR_* Netlink attribute
+ * sequence.
+ *
+ * This parses a series of Netlink attributes that form a flow key, which must
+ * take the same form accepted by flow_from_nlattrs(), but only enough of it to
+ * get the metadata, that is, the parts of the flow key that cannot be
+ * extracted from the packet itself.
+ */
+int flow_metadata_from_nlattrs(u16 *in_port, __be64 *tun_id,
+			       const struct nlattr *attr)
+{
+	const struct nlattr *nla;
+	u16 prev_type;
+	int rem;
+
+	*tun_id = 0;
+
+	prev_type = ODP_KEY_ATTR_UNSPEC;
+	nla_for_each_nested(nla, attr, rem) {
+                int type = nla_type(nla);
+
+                if (type > ODP_KEY_ATTR_MAX || nla_len(nla) != key_lens[type])
+                        return -EINVAL;
+
+		switch (TRANSITION(prev_type, type)) {
+		case TRANSITION(ODP_KEY_ATTR_UNSPEC, ODP_KEY_ATTR_TUN_ID):
+			*tun_id = nla_get_be64(nla);
+			break;
+
+		case TRANSITION(ODP_KEY_ATTR_UNSPEC, ODP_KEY_ATTR_IN_PORT):
+		case TRANSITION(ODP_KEY_ATTR_TUN_ID, ODP_KEY_ATTR_IN_PORT):
+			if (nla_get_u32(nla) >= DP_MAX_PORTS)
+				return -EINVAL;
+			*in_port = nla_get_u32(nla);
+			break;
+
+		default:
+			goto done;
+		}
+
+		prev_type = type;
+	}
+	if (rem)
+		return -EINVAL;
+
+done:
+	if (prev_type == ODP_KEY_ATTR_UNSPEC ||
+	    prev_type == ODP_KEY_ATTR_TUN_ID)
+		return -EINVAL;
+	return 0;
+}
+
 int flow_to_nlattrs(const struct sw_flow_key *swkey, struct sk_buff *skb)
 {
 	struct odp_key_ethernet *eth_key;

datapath/flow.h

@@ -152,6 +152,8 @@ int flow_cmp(const struct tbl_node *, void *target, int len);
 int flow_to_nlattrs(const struct sw_flow_key *, struct sk_buff *);
 int flow_from_nlattrs(struct sw_flow_key *swkey, int *key_lenp,
 		      const struct nlattr *);
+int flow_metadata_from_nlattrs(u16 *in_port, __be64 *tun_id,
+			       const struct nlattr *);
 
 static inline struct sw_flow *flow_cast(const struct tbl_node *node)
 {

lib/dpif-linux.c

@@ -34,6 +34,7 @@
 
 #include "bitmap.h"
 #include "dpif-provider.h"
+#include "dynamic-string.h"
 #include "netdev.h"
 #include "netdev-linux.h"
 #include "netdev-vport.h"
@@ -784,6 +785,7 @@ dpif_linux_flow_dump_done(const struct dpif *dpif OVS_UNUSED, void *state_)
 
 static int
 dpif_linux_execute(struct dpif *dpif_,
+                   const struct nlattr *key, size_t key_len,
                    const struct nlattr *actions, size_t actions_len,
                    const struct ofpbuf *packet)
 {
@@ -801,6 +803,7 @@ dpif_linux_execute(struct dpif *dpif_,
     execute->dp_ifindex = dpif->dp_ifindex;
 
     nl_msg_put_unspec(buf, ODP_PACKET_ATTR_PACKET, packet->data, packet->size);
+    nl_msg_put_unspec(buf, ODP_PACKET_ATTR_KEY, key, key_len);
     nl_msg_put_unspec(buf, ODP_PACKET_ATTR_ACTIONS, actions, actions_len);
 
     error = nl_sock_transact(genl_sock, buf, NULL);

lib/dpif-netdev.c

@@ -944,6 +944,7 @@ dpif_netdev_flow_dump_done(const struct dpif *dpif OVS_UNUSED, void *state_)
 
 static int
 dpif_netdev_execute(struct dpif *dpif,
+                    const struct nlattr *key_attrs, size_t key_len,
                     const struct nlattr *actions, size_t actions_len,
                     const struct ofpbuf *packet)
 {
@@ -975,7 +976,10 @@ dpif_netdev_execute(struct dpif *dpif,
          * if we don't. */
         copy = *packet;
     }
+
     flow_extract(&copy, 0, -1, &key);
+    dpif_netdev_flow_from_nlattrs(key_attrs, key_len, &key);
+
     error = dp_netdev_execute_actions(dp, &copy, &key, actions, actions_len);
     if (mutates) {
         ofpbuf_uninit(&copy);

lib/dpif-provider.h

@@ -282,9 +282,14 @@ struct dpif_class {
     int (*flow_dump_done)(const struct dpif *dpif, void *state);
 
     /* Performs the 'actions_len' bytes of actions in 'actions' on the Ethernet
-     * frame specified in 'packet'. */
+     * frame specified in 'packet' taken from the flow specified in the
-    int (*execute)(struct dpif *dpif, const struct nlattr *actions,
+     * 'key_len' bytes of 'key'.  ('key' is mostly redundant with 'packet', but
-                   size_t actions_len, const struct ofpbuf *packet);
+     * it contains some metadata that cannot be recovered from 'packet', such
+     * as tun_id and in_port.) */
+    int (*execute)(struct dpif *dpif,
+                   const struct nlattr *key, size_t key_len,
+                   const struct nlattr *actions, size_t actions_len,
+                   const struct ofpbuf *packet);
 
     /* Retrieves 'dpif''s "listen mask" into '*listen_mask'.  A 1-bit of value
      * 2**X set in '*listen_mask' indicates that 'dpif' will receive messages

lib/dpif.c

@@ -914,11 +914,15 @@ dpif_flow_dump_done(struct dpif_flow_dump *dump)
 }
 
 /* Causes 'dpif' to perform the 'actions_len' bytes of actions in 'actions' on
- * the Ethernet frame specified in 'packet'.
+ * the Ethernet frame specified in 'packet' taken from the flow specified in
+ * the 'key_len' bytes of 'key'.  ('key' is mostly redundant with 'packet', but
+ * it contains some metadata that cannot be recovered from 'packet', such as
+ * tun_id and in_port.)
  *
  * Returns 0 if successful, otherwise a positive errno value. */
 int
 dpif_execute(struct dpif *dpif,
+             const struct nlattr *key, size_t key_len,
              const struct nlattr *actions, size_t actions_len,
              const struct ofpbuf *buf)
 {
@@ -926,7 +930,8 @@ dpif_execute(struct dpif *dpif,
 
     COVERAGE_INC(dpif_execute);
     if (actions_len > 0) {
-        error = dpif->dpif_class->execute(dpif, actions, actions_len, buf);
+        error = dpif->dpif_class->execute(dpif, key, key_len,
+                                          actions, actions_len, buf);
     } else {
         error = 0;
     }

lib/dpif.h

@@ -148,8 +148,10 @@ bool dpif_flow_dump_next(struct dpif_flow_dump *,
                          const struct dpif_flow_stats **);
 int dpif_flow_dump_done(struct dpif_flow_dump *);
 
-int dpif_execute(struct dpif *, const struct nlattr *actions,
+int dpif_execute(struct dpif *,
-                 size_t actions_len, const struct ofpbuf *);
+                 const struct nlattr *key, size_t key_len,
+                 const struct nlattr *actions, size_t actions_len,
+                 const struct ofpbuf *);
 
 enum dpif_upcall_type {
     DPIF_UC_MISS,               /* Miss in flow table. */

ofproto/ofproto-dpif.c

@@ -2007,9 +2007,16 @@ execute_odp_actions(struct ofproto_dpif *ofproto, const struct flow *flow,
 
         return true;
     } else {
+        struct odputil_keybuf keybuf;
+        struct ofpbuf key;
         int error;
 
-        error = dpif_execute(ofproto->dpif, odp_actions, actions_len, packet);
+        ofpbuf_use_stack(&key, &keybuf, sizeof keybuf);
+        odp_flow_key_from_flow(&key, flow);
+
+        error = dpif_execute(ofproto->dpif, key.data, key.size,
+                             odp_actions, actions_len, packet);
+
         ofpbuf_delete(packet);
         return !error;
     }
@@ -2671,12 +2678,20 @@ static int
 send_packet(struct ofproto_dpif *ofproto, uint32_t odp_port,
             const struct ofpbuf *packet)
 {
-    struct ofpbuf odp_actions;
+    struct ofpbuf key, odp_actions;
+    struct odputil_keybuf keybuf;
+    struct flow flow;
     int error;
 
+    flow_extract((struct ofpbuf *) packet, 0, 0, &flow);
+    ofpbuf_use_stack(&key, &keybuf, sizeof keybuf);
+    odp_flow_key_from_flow(&key, &flow);
+
     ofpbuf_init(&odp_actions, 32);
     nl_msg_put_u32(&odp_actions, ODP_ACTION_ATTR_OUTPUT, odp_port);
-    error = dpif_execute(ofproto->dpif, odp_actions.data, odp_actions.size,
+    error = dpif_execute(ofproto->dpif,
+                         key.data, key.size,
+                         odp_actions.data, odp_actions.size,
                          packet);
     ofpbuf_uninit(&odp_actions);
 
@@ -3721,13 +3736,18 @@ packet_out(struct ofproto *ofproto_, struct ofpbuf *packet,
     error = validate_actions(ofp_actions, n_ofp_actions, flow,
                              ofproto->max_ports);
     if (!error) {
+        struct odputil_keybuf keybuf;
         struct action_xlate_ctx ctx;
         struct ofpbuf *odp_actions;
+        struct ofpbuf key;
+
+        ofpbuf_use_stack(&key, &keybuf, sizeof keybuf);
+        odp_flow_key_from_flow(&key, flow);
 
         action_xlate_ctx_init(&ctx, ofproto, flow, packet);
         odp_actions = xlate_actions(&ctx, ofp_actions, n_ofp_actions);
-        dpif_execute(ofproto->dpif, odp_actions->data, odp_actions->size,
+        dpif_execute(ofproto->dpif, key.data, key.size,
-                     packet);
+                     odp_actions->data, odp_actions->size, packet);
         ofpbuf_delete(odp_actions);
     }
     return error;