/*
- * Copyright (c) 2007-2011 Nicira Networks.
+ * 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
#include <linux/kernel.h>
#include <linux/mutex.h>
#include <linux/netdevice.h>
-#include <linux/seqlock.h>
#include <linux/skbuff.h>
-#include <linux/version.h>
+#include <linux/u64_stats_sync.h>
-#include "checksum.h"
#include "compat.h"
#include "flow.h"
-#include "dp_sysfs.h"
+#include "flow_table.h"
#include "vlan.h"
+#include "vport.h"
-struct vport;
+#define DP_MAX_PORTS USHRT_MAX
+#define DP_VPORT_HASH_BUCKETS 1024
-#define DP_MAX_PORTS 1024
#define SAMPLE_ACTION_DEPTH 3
/**
* @n_lost: Number of received packets that had no matching flow in the flow
* table that could not be sent to userspace (normally due to an overflow in
* one of the datapath's queues).
+ * @n_mask_hit: Number of masks looked up for flow match.
+ * @n_mask_hit / (@n_hit + @n_missed) will be the average masks looked
+ * up per packet.
*/
struct dp_stats_percpu {
u64 n_hit;
u64 n_missed;
u64 n_lost;
- seqcount_t seqlock;
+ u64 n_mask_hit;
+ struct u64_stats_sync sync;
};
/**
* struct datapath - datapath for flow-based packet switching
* @rcu: RCU callback head for deferred destruction.
* @list_node: Element in global 'dps' list.
- * @ifobj: Represents /sys/class/net/<devname>/brif. Protected by RTNL.
- * @n_flows: Number of flows currently in flow table.
- * @table: Current flow table. Protected by genl_lock and RCU.
- * @ports: Map from port number to &struct vport. %OVSP_LOCAL port
- * always exists, other ports may be %NULL. Protected by RTNL and RCU.
- * @port_list: List of all ports in @ports in arbitrary order. RTNL required
- * to iterate or modify.
+ * @table: flow table.
+ * @ports: Hash table for ports. %OVSP_LOCAL port always exists. Protected by
+ * ovs_mutex and RCU.
* @stats_percpu: Per-CPU datapath statistics.
+ * @net: Reference to net namespace.
*
* Context: See the comment on locking at the top of datapath.c for additional
* locking information.
struct datapath {
struct rcu_head rcu;
struct list_head list_node;
- struct kobject ifobj;
/* Flow table. */
- struct flow_table __rcu *table;
+ struct flow_table table;
/* Switch ports. */
- struct vport __rcu *ports[DP_MAX_PORTS];
- struct list_head port_list;
+ struct hlist_head *ports;
/* Stats. */
struct dp_stats_percpu __percpu *stats_percpu;
+
+#ifdef CONFIG_NET_NS
+ /* Network namespace ref. */
+ struct net *net;
+#endif
+
+ u32 user_features;
};
/**
* struct ovs_skb_cb - OVS data in skb CB
* @flow: The flow associated with this packet. May be %NULL if no flow.
- * @tun_id: ID of the tunnel that encapsulated this packet. It is 0 if the
- * @ip_summed: Consistently stores L4 checksumming status across different
- * kernel versions.
- * @csum_start: Stores the offset from which to start checksumming independent
- * of the transport header on all kernel versions.
- * packet was not received on a tunnel.
- * @vlan_tci: Provides a substitute for the skb->vlan_tci field on kernels
- * before 2.6.27.
+ * @pkt_key: The flow information extracted from the packet. Must be nonnull.
+ * @tun_key: Key for the tunnel that encapsulated this packet. NULL if the
+ * @input_vport: The original vport packet came in on. This value is cached
+ * when a packet is received by OVS.
*/
struct ovs_skb_cb {
struct sw_flow *flow;
- __be64 tun_id;
-#ifdef NEED_CSUM_NORMALIZE
- enum csum_type ip_summed;
- u16 csum_start;
-#endif
-#ifdef NEED_VLAN_FIELD
- u16 vlan_tci;
-#endif
+ struct sw_flow_key *pkt_key;
+ struct ovs_key_ipv4_tunnel *tun_key;
+ struct vport *input_vport;
};
#define OVS_CB(skb) ((struct ovs_skb_cb *)(skb)->cb)
* struct dp_upcall - metadata to include with a packet to send to userspace
* @cmd: One of %OVS_PACKET_CMD_*.
* @key: Becomes %OVS_PACKET_ATTR_KEY. Must be nonnull.
- * @userdata: If nonnull, its u64 value is extracted and passed to userspace as
+ * @userdata: If nonnull, its variable-length value is passed to userspace as
* %OVS_PACKET_ATTR_USERDATA.
- * @pid: Netlink PID to which packet should be sent. If @pid is 0 then no
+ * @portid: Netlink PID to which packet should be sent. If @portid is 0 then no
* packet is sent and the packet is accounted in the datapath's @n_lost
* counter.
*/
u8 cmd;
const struct sw_flow_key *key;
const struct nlattr *userdata;
- u32 pid;
+ u32 portid;
};
-extern struct notifier_block dp_device_notifier;
-extern struct genl_multicast_group dp_vport_multicast_group;
-extern int (*dp_ioctl_hook)(struct net_device *dev, struct ifreq *rq, int cmd);
+/**
+ * struct ovs_net - Per net-namespace data for ovs.
+ * @dps: List of datapaths to enable dumping them all out.
+ * Protected by genl_mutex.
+ * @vport_net: Per network namespace data for vport.
+ */
+struct ovs_net {
+ struct list_head dps;
+ struct vport_net vport_net;
+ struct work_struct dp_notify_work;
+};
-void dp_process_received_packet(struct vport *, struct sk_buff *);
-void dp_detach_port(struct vport *);
-int dp_upcall(struct datapath *, struct sk_buff *,
- const struct dp_upcall_info *);
+extern int ovs_net_id;
+void ovs_lock(void);
+void ovs_unlock(void);
-struct datapath *get_dp(int dp_idx);
-const char *dp_name(const struct datapath *dp);
-struct sk_buff *ovs_vport_cmd_build_info(struct vport *, u32 pid, u32 seq,
+#ifdef CONFIG_LOCKDEP
+int lockdep_ovsl_is_held(void);
+#else
+#define lockdep_ovsl_is_held() 1
+#endif
+
+#define ASSERT_OVSL() WARN_ON(unlikely(!lockdep_ovsl_is_held()))
+#define ovsl_dereference(p) \
+ rcu_dereference_protected(p, lockdep_ovsl_is_held())
+#define rcu_dereference_ovsl(p) \
+ rcu_dereference_check(p, lockdep_ovsl_is_held())
+
+static inline struct net *ovs_dp_get_net(struct datapath *dp)
+{
+ return read_pnet(&dp->net);
+}
+
+static inline void ovs_dp_set_net(struct datapath *dp, struct net *net)
+{
+ write_pnet(&dp->net, net);
+}
+
+struct vport *ovs_lookup_vport(const struct datapath *dp, u16 port_no);
+
+static inline struct vport *ovs_vport_rcu(const struct datapath *dp, int port_no)
+{
+ WARN_ON_ONCE(!rcu_read_lock_held());
+ return ovs_lookup_vport(dp, port_no);
+}
+
+static inline struct vport *ovs_vport_ovsl_rcu(const struct datapath *dp, int port_no)
+{
+ WARN_ON_ONCE(!rcu_read_lock_held() && !lockdep_ovsl_is_held());
+ return ovs_lookup_vport(dp, port_no);
+}
+
+static inline struct vport *ovs_vport_ovsl(const struct datapath *dp, int port_no)
+{
+ ASSERT_OVSL();
+ return ovs_lookup_vport(dp, port_no);
+}
+
+extern struct notifier_block ovs_dp_device_notifier;
+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(struct sk_buff *,
+ struct sw_flow_key *pkt_key, bool recirc);
+void ovs_dp_detach_port(struct vport *);
+int ovs_dp_upcall(struct datapath *, struct sk_buff *,
+ const struct dp_upcall_info *);
+
+const char *ovs_dp_name(const struct datapath *dp);
+struct sk_buff *ovs_vport_cmd_build_info(struct vport *, u32 portid, u32 seq,
u8 cmd);
-int execute_actions(struct datapath *dp, struct sk_buff *skb);
+int ovs_execute_actions(struct datapath *dp, struct sk_buff *skb, bool recirc);
+void ovs_dp_notify_wq(struct work_struct *work);
+
+#define OVS_NLERR(fmt, ...) \
+do { \
+ if (net_ratelimit()) \
+ pr_info("netlink: " fmt, ##__VA_ARGS__); \
+} while (0)
#endif /* datapath.h */
git://git.onelab.eu / sliver-openvswitch.git / blobdiff