Recirculation implementation for Linux kernel data path.
Signed-off-by: Andy Zhou <az...@nicira.com>
---
datapath/actions.c | 55 ++++++++++++++++++++++++++++++++++++++++++++++-
datapath/datapath.c | 35 ++++++++++++++++++------------
datapath/datapath.h | 3 ++-
datapath/flow.h | 4 ++++
datapath/flow_netlink.c | 55 ++++++++++++++++++++++++++++++++++++++++++++++-
5 files changed, 135 insertions(+), 17 deletions(-)
diff --git a/datapath/actions.c b/datapath/actions.c
index 0b66e7c..81ba67e 100644
--- a/datapath/actions.c
+++ b/datapath/actions.c
@@ -1,5 +1,5 @@
/*
- * Copyright (c) 2007-2013 Nicira, Inc.
+ * Copyright (c) 2007-2014 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
@@ -506,6 +506,37 @@ static int execute_set_action(struct sk_buff *skb,
return err;
}
+static int execute_recirc(struct datapath *dp, struct sk_buff *skb,
+ const struct nlattr *recirc_attr)
+{
+ const struct ovs_action_recirc *act_recirc;
+ struct sw_flow_key *key = OVS_CB(skb)->pkt_key;
+ struct vport *p;
+ u32 dp_hash = 0;
+
+ act_recirc = (struct ovs_action_recirc *) nla_data(recirc_attr);
+
+ key->recirc_id = act_recirc->recirc_id;
+ if (act_recirc->hash_alg == OVS_RECIRC_HASH_ALG_L4) {
+ dp_hash = skb_get_rxhash(skb);
+ /* skb_get_rxhash() can return 0 when there
+ * is packet error. */
+ if (!dp_hash)
+ dp_hash = 0x1;
+ }
+ key->dp_hash = dp_hash;
+
+ p = ovs_vport_rcu(dp, key->phy.in_port);
+ if (unlikely(!p)) {
+ kfree_skb(skb);
+ return -ENODEV;
+ }
+
+ ovs_dp_process_packet_with_key(p, skb);
+
+ return 0;
+}
+
/* 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)
@@ -546,12 +577,34 @@ static int do_execute_actions(struct datapath *dp, struct
sk_buff *skb,
err = pop_vlan(skb);
break;
+ case OVS_ACTION_ATTR_RECIRC: {
+ struct sk_buff *recirc_skb;
+ const bool last_action = (a->nla_len == rem);
+
+ if (last_action || keep_skb)
+ /* Clone skb if not the last action. */
+ recirc_skb = skb_clone(skb, GFP_ATOMIC);
+ else
+ recirc_skb = skb;
+
+ err = execute_recirc(dp, recirc_skb, a);
+
+ /* Return directly if recirc is the last action
+ * or err. */
+ if (last_action || err)
+ return err;
+
+ 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);
+ if (unlikely(err)) /* skb already freed. */
+ return err;
break;
}
diff --git a/datapath/datapath.c b/datapath/datapath.c
index c6d42db..e2a2693 100644
--- a/datapath/datapath.c
+++ b/datapath/datapath.c
@@ -240,33 +240,24 @@ void ovs_dp_detach_port(struct vport *p)
ovs_vport_del(p);
}
-/* Must be called with rcu_read_lock. */
-void ovs_dp_process_received_packet(struct vport *p, struct sk_buff *skb)
+void ovs_dp_process_packet_with_key(const struct vport *p, struct sk_buff *skb)
{
struct datapath *dp = p->dp;
struct sw_flow *flow;
struct dp_stats_percpu *stats;
- struct sw_flow_key key;
+ struct sw_flow_key *key = OVS_CB(skb)->pkt_key;
u64 *stats_counter;
u32 n_mask_hit;
- int error;
stats = this_cpu_ptr(dp->stats_percpu);
- /* Extract flow from 'skb' into 'key'. */
- error = ovs_flow_extract(skb, p->port_no, &key);
- if (unlikely(error)) {
- kfree_skb(skb);
- return;
- }
-
/* Look up flow. */
- flow = ovs_flow_tbl_lookup_stats(&dp->table, &key, &n_mask_hit);
+ flow = ovs_flow_tbl_lookup_stats(&dp->table, key, &n_mask_hit);
if (unlikely(!flow)) {
struct dp_upcall_info upcall;
upcall.cmd = OVS_PACKET_CMD_MISS;
- upcall.key = &key;
+ upcall.key = key;
upcall.userdata = NULL;
upcall.portid = p->upcall_portid;
ovs_dp_upcall(dp, skb, &upcall);
@@ -276,7 +267,6 @@ void ovs_dp_process_received_packet(struct vport *p, struct
sk_buff *skb)
}
OVS_CB(skb)->flow = flow;
- OVS_CB(skb)->pkt_key = &key;
ovs_flow_stats_update(OVS_CB(skb)->flow, skb);
ovs_execute_actions(dp, skb);
@@ -290,6 +280,23 @@ out:
u64_stats_update_end(&stats->sync);
}
+/* Must be called with rcu_read_lock. */
+void ovs_dp_process_received_packet(struct vport *p, struct sk_buff *skb)
+{
+ int error;
+ struct sw_flow_key key;
+
+ /* Extract flow from 'skb' into 'key'. */
+ error = ovs_flow_extract(skb, p->port_no, &key);
+ if (unlikely(error)) {
+ kfree_skb(skb);
+ return;
+ }
+
+ OVS_CB(skb)->pkt_key = &key;
+ ovs_dp_process_packet_with_key(p, skb);
+}
+
int ovs_dp_upcall(struct datapath *dp, struct sk_buff *skb,
const struct dp_upcall_info *upcall_info)
{
diff --git a/datapath/datapath.h b/datapath/datapath.h
index 40e0f90..d3b06ee 100644
--- a/datapath/datapath.h
+++ b/datapath/datapath.h
@@ -1,5 +1,5 @@
/*
- * Copyright (c) 2007-2012 Nicira, Inc.
+ * Copyright (c) 2007-2014 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
@@ -188,6 +188,7 @@ extern struct genl_family dp_vport_genl_family;
extern struct genl_multicast_group ovs_dp_vport_multicast_group;
void ovs_dp_process_received_packet(struct vport *, struct sk_buff *);
+void ovs_dp_process_packet_with_key(const struct vport *, struct sk_buff *);
void ovs_dp_detach_port(struct vport *);
int ovs_dp_upcall(struct datapath *, struct sk_buff *,
const struct dp_upcall_info *);
diff --git a/datapath/flow.h b/datapath/flow.h
index 5043a6e..3ccaec5 100644
--- a/datapath/flow.h
+++ b/datapath/flow.h
@@ -68,6 +68,8 @@ static inline void ovs_flow_tun_key_init(struct
ovs_key_ipv4_tunnel *tun_key,
}
struct sw_flow_key {
+ u32 dp_hash; /* Datapath computed hash value */
+ u32 recirc_id; /* Recirculation ID */
struct ovs_key_ipv4_tunnel tun_key; /* Encapsulating tunnel key. */
struct {
u32 priority; /* Packet QoS priority. */
@@ -190,4 +192,6 @@ u64 ovs_flow_used_time(unsigned long flow_jiffies);
int ovs_flow_extract(struct sk_buff *, u16 in_port, struct sw_flow_key *);
+u32 ovs_flow_hash_symmetric_l4(struct sw_flow_key *);
+
#endif /* flow.h */
diff --git a/datapath/flow_netlink.c b/datapath/flow_netlink.c
index 5c32cd0..207bdeb 100644
--- a/datapath/flow_netlink.c
+++ b/datapath/flow_netlink.c
@@ -123,14 +123,20 @@ static bool match_validate(const struct sw_flow_match
*match,
| (1ULL << OVS_KEY_ATTR_ICMP)
| (1ULL << OVS_KEY_ATTR_ICMPV6)
| (1ULL << OVS_KEY_ATTR_ARP)
+ | (1ULL << OVS_KEY_ATTR_DP_HASH)
| (1ULL << OVS_KEY_ATTR_ND));
/* Always allowed mask fields. */
mask_allowed |= ((1ULL << OVS_KEY_ATTR_TUNNEL)
| (1ULL << OVS_KEY_ATTR_IN_PORT)
- | (1ULL << OVS_KEY_ATTR_ETHERTYPE));
+ | (1ULL << OVS_KEY_ATTR_ETHERTYPE)
+ | (1ULL << OVS_KEY_ATTR_RECIRC_ID));
/* Check key attributes. */
+ if (match->key->dp_hash) {
+ mask_allowed |= (1ULL << OVS_KEY_ATTR_DP_HASH);
+ }
+
if (match->key->eth.type == htons(ETH_P_ARP)
|| match->key->eth.type == htons(ETH_P_RARP)) {
key_expected |= 1ULL << OVS_KEY_ATTR_ARP;
@@ -252,6 +258,8 @@ static const int ovs_key_lens[OVS_KEY_ATTR_MAX + 1] = {
[OVS_KEY_ATTR_ICMPV6] = sizeof(struct ovs_key_icmpv6),
[OVS_KEY_ATTR_ARP] = sizeof(struct ovs_key_arp),
[OVS_KEY_ATTR_ND] = sizeof(struct ovs_key_nd),
+ [OVS_KEY_ATTR_DP_HASH] = sizeof(u32),
+ [OVS_KEY_ATTR_RECIRC_ID] = sizeof(u32),
[OVS_KEY_ATTR_TUNNEL] = -1,
};
@@ -455,6 +463,35 @@ static int ipv4_tun_to_nlattr(struct sk_buff *skb,
static int metadata_from_nlattrs(struct sw_flow_match *match, u64 *attrs,
const struct nlattr **a, bool is_mask)
{
+ if (*attrs & (1ULL << OVS_KEY_ATTR_DP_HASH)) {
+ u32 hash_val = nla_get_u32(a[OVS_KEY_ATTR_DP_HASH]);
+
+ if (!is_mask && !hash_val) {
+ OVS_NLERR("Hash value can not be zero\n");
+ return -EINVAL;
+ }
+
+ SW_FLOW_KEY_PUT(match, dp_hash, hash_val, is_mask);
+ *attrs &= ~(1ULL << OVS_KEY_ATTR_DP_HASH);
+ }
+
+ if (*attrs & (1ULL << OVS_KEY_ATTR_RECIRC_ID)) {
+ u32 recirc_id = nla_get_u32(a[OVS_KEY_ATTR_RECIRC_ID]);
+
+ if (is_mask && (recirc_id > 0 && recirc_id < UINT_MAX)) {
+ OVS_NLERR("Reicrc_id mask is neither wildcard, nor
exact match \n");
+ return -EINVAL;
+ }
+
+ SW_FLOW_KEY_PUT(match, recirc_id, recirc_id, is_mask);
+ *attrs &= ~(1ULL << OVS_KEY_ATTR_RECIRC_ID);
+ }
+
+ if (is_mask) {
+ /* Always exact match recirc_id. */
+ SW_FLOW_KEY_PUT(match, recirc_id, UINT_MAX, is_mask);
+ }
+
if (*attrs & (1ULL << OVS_KEY_ATTR_PRIORITY)) {
SW_FLOW_KEY_PUT(match, phy.priority,
nla_get_u32(a[OVS_KEY_ATTR_PRIORITY]), is_mask);
@@ -859,6 +896,8 @@ int ovs_nla_get_flow_metadata(struct sw_flow *flow,
flow->key.phy.in_port = DP_MAX_PORTS;
flow->key.phy.priority = 0;
flow->key.phy.skb_mark = 0;
+ flow->key.dp_hash = 0;
+ flow->key.recirc_id = 0;
memset(tun_key, 0, sizeof(flow->key.tun_key));
err = parse_flow_nlattrs(attr, a, &attrs);
@@ -882,6 +921,16 @@ int ovs_nla_put_flow(const struct sw_flow_key *swkey,
struct nlattr *nla, *encap;
bool is_mask = (swkey != output);
+ /* Only output dp_hash and recric_id if their key values are none zero*/
+ /* Ommitting those key attributes implies zero on the other size */
+ if (swkey->dp_hash)
+ if (nla_put_u32(skb, OVS_KEY_ATTR_DP_HASH, output->dp_hash))
+ goto nla_put_failure;
+
+ if (swkey->recirc_id)
+ if (nla_put_u32(skb, OVS_KEY_ATTR_RECIRC_ID, output->recirc_id))
+ goto nla_put_failure;
+
if (nla_put_u32(skb, OVS_KEY_ATTR_PRIORITY, output->phy.priority))
goto nla_put_failure;
@@ -1418,6 +1467,7 @@ int ovs_nla_copy_actions(const struct nlattr *attr,
/* Expected argument lengths, (u32)-1 for variable length. */
static const u32 action_lens[OVS_ACTION_ATTR_MAX + 1] = {
[OVS_ACTION_ATTR_OUTPUT] = sizeof(u32),
+ [OVS_ACTION_ATTR_RECIRC] = sizeof(struct
ovs_action_recirc),
[OVS_ACTION_ATTR_USERSPACE] = (u32)-1,
[OVS_ACTION_ATTR_PUSH_VLAN] = sizeof(struct
ovs_action_push_vlan),
[OVS_ACTION_ATTR_POP_VLAN] = 0,
@@ -1461,6 +1511,9 @@ int ovs_nla_copy_actions(const struct nlattr *attr,
return -EINVAL;
break;
+ case OVS_ACTION_ATTR_RECIRC:
+ break;
+
case OVS_ACTION_ATTR_SET:
err = validate_set(a, key, sfa, &skip_copy);
if (err)
--
1.7.9.5
_______________________________________________
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
