/*
- * Copyright (c) 2008, 2009, 2010, 2011, 2012, 2013 Nicira, Inc.
+ * Copyright (c) 2008, 2009, 2010, 2011, 2012, 2013, 2014 Nicira, Inc.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* The terms written in quotes below are defined in later sections.
*
* When a datapath "port" receives a packet, it extracts the headers (the
- * "flow"). If the datapath's "flow table" contains a "flow entry" whose flow
- * is the same as the packet's, then it executes the "actions" in the flow
- * entry and increments the flow's statistics. If there is no matching flow
- * entry, the datapath instead appends the packet to an "upcall" queue.
+ * "flow"). If the datapath's "flow table" contains a "flow entry" matching
+ * the packet, then it executes the "actions" in the flow entry and increments
+ * the flow's statistics. If there is no matching flow entry, the datapath
+ * instead appends the packet to an "upcall" queue.
*
*
* Ports
* "internal" (for a simulated port used to connect to the TCP/IP stack),
* and "gre" (for a GRE tunnel).
*
- * - A Netlink PID (see "Upcall Queuing and Ordering" below).
+ * - A Netlink PID for each upcall reading thread (see "Upcall Queuing and
+ * Ordering" below).
*
* The dpif interface has functions for adding and deleting ports. When a
* datapath implements these (e.g. as the Linux and netdev datapaths do), then
* Flow Table
* ==========
*
- * The flow table is a hash table of "flow entries". Each flow entry contains:
+ * The flow table is a collection of "flow entries". Each flow entry contains:
*
* - A "flow", that is, a summary of the headers in an Ethernet packet. The
- * flow is the hash key and thus must be unique within the flow table.
- * Flows are fine-grained entities that include L2, L3, and L4 headers. A
- * single TCP connection consists of two flows, one in each direction.
+ * flow must be unique within the flow table. Flows are fine-grained
+ * entities that include L2, L3, and L4 headers. A single TCP connection
+ * consists of two flows, one in each direction.
*
* In Open vSwitch userspace, "struct flow" is the typical way to describe
* a flow, but the datapath interface uses a different data format to
* "struct ovs_key_*" in include/linux/openvswitch.h for details.
* lib/odp-util.h defines several functions for working with these flows.
*
- * (In case you are familiar with OpenFlow, datapath flows are analogous
- * to OpenFlow flow matches. The most important difference is that
- * OpenFlow allows fields to be wildcarded and prioritized, whereas a
- * datapath's flow table is a hash table so every flow must be
- * exact-match, thus without priorities.)
+ * - A "mask" that, for each bit in the flow, specifies whether the datapath
+ * should consider the corresponding flow bit when deciding whether a
+ * given packet matches the flow entry. The original datapath design did
+ * not support matching: every flow entry was exact match. With the
+ * addition of a mask, the interface supports datapaths with a spectrum of
+ * wildcard matching capabilities, from those that only support exact
+ * matches to those that support bitwise wildcarding on the entire flow
+ * key, as well as datapaths with capabilities somewhere in between.
+ *
+ * Datapaths do not provide a way to query their wildcarding capabilities,
+ * nor is it expected that the client should attempt to probe for the
+ * details of their support. Instead, a client installs flows with masks
+ * that wildcard as many bits as acceptable. The datapath then actually
+ * wildcards as many of those bits as it can and changes the wildcard bits
+ * that it does not support into exact match bits. A datapath that can
+ * wildcard any bit, for example, would install the supplied mask, an
+ * exact-match only datapath would install an exact-match mask regardless
+ * of what mask the client supplied, and a datapath in the middle of the
+ * spectrum would selectively change some wildcard bits into exact match
+ * bits.
+ *
+ * Regardless of the requested or installed mask, the datapath retains the
+ * original flow supplied by the client. (It does not, for example, "zero
+ * out" the wildcarded bits.) This allows the client to unambiguously
+ * identify the flow entry in later flow table operations.
+ *
+ * The flow table does not have priorities; that is, all flow entries have
+ * equal priority. Detecting overlapping flow entries is expensive in
+ * general, so the datapath is not required to do it. It is primarily the
+ * client's responsibility not to install flow entries whose flow and mask
+ * combinations overlap.
*
* - A list of "actions" that tell the datapath what to do with packets
* within a flow. Some examples of actions are OVS_ACTION_ATTR_OUTPUT,
* connection consists of two flows with 1-ms latency to set up each one.
*
* To receive upcalls, a client has to enable them with dpif_recv_set(). A
- * datapath should generally support multiple clients at once (e.g. so that one
- * may run "ovs-dpctl show" or "ovs-dpctl dump-flows" while "ovs-vswitchd" is
- * also running) but need not support multiple clients enabling upcalls at
- * once.
+ * datapath should generally support being opened multiple times (e.g. so that
+ * one may run "ovs-dpctl show" or "ovs-dpctl dump-flows" while "ovs-vswitchd"
+ * is also running) but need not support more than one of these clients
+ * enabling upcalls at once.
*
*
* Upcall Queuing and Ordering
* PID in "action" upcalls is that dpif_port_get_pid() returns a constant value
* and all upcalls are appended to a single queue.
*
- * The ideal behavior is:
+ * The preferred behavior is:
*
* - Each port has a PID that identifies the queue used for "miss" upcalls
* on that port. (Thus, if each port has its own queue for "miss"
*
* - Upcalls that specify the "special" Netlink PID are queued separately.
*
+ * Multiple threads may want to read upcalls simultaneously from a single
+ * datapath. To support multiple threads well, one extends the above preferred
+ * behavior:
+ *
+ * - Each port has multiple PIDs. The datapath distributes "miss" upcalls
+ * across the PIDs, ensuring that a given flow is mapped in a stable way
+ * to a single PID.
+ *
+ * - For "action" upcalls, the thread can specify its own Netlink PID or
+ * other threads' Netlink PID of the same port for offloading purpose
+ * (e.g. in a "round robin" manner).
+ *
*
* Packet Format
* =============
* location.
*
* - Adding and removing ports to achieve a new configuration.
+ *
+ *
+ * Thread-safety
+ * =============
+ *
+ * Most of the dpif functions are fully thread-safe: they may be called from
+ * any number of threads on the same or different dpif objects. The exceptions
+ * are:
+ *
+ * - dpif_port_poll() and dpif_port_poll_wait() are conditionally
+ * thread-safe: they may be called from different threads only on
+ * different dpif objects.
+ *
+ * - dpif_flow_dump_next() is conditionally thread-safe: It may be called
+ * from different threads with the same 'struct dpif_flow_dump', but all
+ * other parameters must be different for each thread.
+ *
+ * - dpif_flow_dump_done() is conditionally thread-safe: All threads that
+ * share the same 'struct dpif_flow_dump' must have finished using it.
+ * This function must then be called exactly once for a particular
+ * dpif_flow_dump to finish the corresponding flow dump operation.
+ *
+ * - Functions that operate on 'struct dpif_port_dump' are conditionally
+ * thread-safe with respect to those objects. That is, one may dump ports
+ * from any number of threads at once, but each thread must use its own
+ * struct dpif_port_dump.
*/
#ifndef DPIF_H
#define DPIF_H 1
#include <stdbool.h>
#include <stddef.h>
#include <stdint.h>
-#include <linux/openvswitch.h>
-#include "openflow/openflow.h"
#include "netdev.h"
+#include "ofpbuf.h"
+#include "openflow/openflow.h"
+#include "packets.h"
#include "util.h"
#ifdef __cplusplus
struct ds;
struct flow;
struct nlattr;
-struct ofpbuf;
struct sset;
struct dpif_class;
uint64_t n_missed; /* Number of flow table misses. */
uint64_t n_lost; /* Number of misses not sent to userspace. */
uint64_t n_flows; /* Number of flows present. */
+ uint64_t n_mask_hit; /* Number of mega flow masks visited for
+ flow table matches. */
+ uint32_t n_masks; /* Number of mega flow masks. */
};
int dpif_get_dp_stats(const struct dpif *, struct dpif_dp_stats *);
const char *dpif_port_open_type(const char *datapath_type,
const char *port_type);
-int dpif_port_add(struct dpif *, struct netdev *, uint32_t *port_nop);
-int dpif_port_del(struct dpif *, uint32_t port_no);
+int dpif_port_add(struct dpif *, struct netdev *, odp_port_t *port_nop);
+int dpif_port_del(struct dpif *, odp_port_t port_no);
/* A port within a datapath.
*
struct dpif_port {
char *name; /* Network device name, e.g. "eth0". */
char *type; /* Network device type, e.g. "system". */
- uint32_t port_no; /* Port number within datapath. */
+ odp_port_t port_no; /* Port number within datapath. */
};
void dpif_port_clone(struct dpif_port *, const struct dpif_port *);
void dpif_port_destroy(struct dpif_port *);
bool dpif_port_exists(const struct dpif *dpif, const char *devname);
-int dpif_port_query_by_number(const struct dpif *, uint32_t port_no,
+int dpif_port_query_by_number(const struct dpif *, odp_port_t port_no,
struct dpif_port *);
int dpif_port_query_by_name(const struct dpif *, const char *devname,
struct dpif_port *);
-int dpif_port_get_name(struct dpif *, uint32_t port_no,
+int dpif_port_get_name(struct dpif *, odp_port_t port_no,
char *name, size_t name_size);
-int dpif_get_max_ports(const struct dpif *);
-uint32_t dpif_port_get_pid(const struct dpif *, uint32_t port_no);
+uint32_t dpif_port_get_pid(const struct dpif *, odp_port_t port_no,
+ uint32_t hash);
struct dpif_port_dump {
const struct dpif *dpif;
uint64_t n_packets;
uint64_t n_bytes;
long long int used;
- uint8_t tcp_flags;
+ uint16_t tcp_flags;
};
void dpif_flow_stats_extract(const struct flow *, const struct ofpbuf *packet,
int dpif_flow_flush(struct dpif *);
int dpif_flow_put(struct dpif *, enum dpif_flow_put_flags,
const struct nlattr *key, size_t key_len,
+ const struct nlattr *mask, size_t mask_len,
const struct nlattr *actions, size_t actions_len,
struct dpif_flow_stats *);
int dpif_flow_del(struct dpif *,
struct dpif_flow_dump {
const struct dpif *dpif;
- int error;
- void *state;
+ void *iter;
};
-void dpif_flow_dump_start(struct dpif_flow_dump *, const struct dpif *);
-bool dpif_flow_dump_next(struct dpif_flow_dump *,
+void dpif_flow_dump_state_init(const struct dpif *, void **statep);
+int dpif_flow_dump_start(struct dpif_flow_dump *, const struct dpif *);
+bool dpif_flow_dump_next(struct dpif_flow_dump *, void *state,
const struct nlattr **key, size_t *key_len,
+ const struct nlattr **mask, size_t *mask_len,
const struct nlattr **actions, size_t *actions_len,
const struct dpif_flow_stats **);
+bool dpif_flow_dump_next_may_destroy_keys(struct dpif_flow_dump *dump,
+ void *state);
int dpif_flow_dump_done(struct dpif_flow_dump *);
-\f
-/* Packet operations. */
-
-int dpif_execute(struct dpif *,
- const struct nlattr *key, size_t key_len,
- const struct nlattr *actions, size_t actions_len,
- const struct ofpbuf *);
+void dpif_flow_dump_state_uninit(const struct dpif *, void *state);
\f
/* Operation batching interface.
*
enum dpif_flow_put_flags flags; /* DPIF_FP_*. */
const struct nlattr *key; /* Flow to put. */
size_t key_len; /* Length of 'key' in bytes. */
+ const struct nlattr *mask; /* Mask to put. */
+ size_t mask_len; /* Length of 'mask' in bytes. */
const struct nlattr *actions; /* Actions to perform on flow. */
size_t actions_len; /* Length of 'actions' in bytes. */
};
struct dpif_execute {
- const struct nlattr *key; /* Partial flow key (only for metadata). */
- size_t key_len; /* Length of 'key' in bytes. */
+ /* Raw support for execute passed along to the provider. */
const struct nlattr *actions; /* Actions to execute on packet. */
size_t actions_len; /* Length of 'actions' in bytes. */
- const struct ofpbuf *packet; /* Packet to execute. */
+ struct ofpbuf *packet; /* Packet to execute. */
+ struct pkt_metadata md; /* Packet metadata. */
+
+ /* Some dpif providers do not implement every action. The Linux kernel
+ * datapath, in particular, does not implement ARP field modification.
+ *
+ * If this member is set to true, the dpif layer executes in userspace all
+ * of the actions that it can, and for OVS_ACTION_ATTR_OUTPUT and
+ * OVS_ACTION_ATTR_USERSPACE actions it passes the packet through to the
+ * dpif implementation. */
+ bool needs_help;
};
+int dpif_execute(struct dpif *, struct dpif_execute *);
+
struct dpif_op {
enum dpif_op_type type;
int error;
/* A packet passed up from the datapath to userspace.
*
- * If 'key' or 'actions' is nonnull, then it points into data owned by
- * 'packet', so their memory cannot be freed separately. (This is hardly a
- * great way to do things but it works out OK for the dpif providers and
- * clients that exist so far.)
+ * The 'packet', 'key' and 'userdata' may point into data in a buffer
+ * provided by the caller, so the buffer should be released only after the
+ * upcall processing has been finished.
+ *
+ * While being processed, the 'packet' may be reallocated, so the packet must
+ * be separately released with ofpbuf_uninit().
*/
struct dpif_upcall {
/* All types. */
enum dpif_upcall_type type;
- struct ofpbuf *packet; /* Packet data. */
+ struct ofpbuf packet; /* Packet data. */
struct nlattr *key; /* Flow key. */
size_t key_len; /* Length of 'key' in bytes. */
/* DPIF_UC_ACTION only. */
- uint64_t userdata; /* Argument to OVS_ACTION_ATTR_USERSPACE. */
+ struct nlattr *userdata; /* Argument to OVS_ACTION_ATTR_USERSPACE. */
};
int dpif_recv_set(struct dpif *, bool enable);
-int dpif_recv(struct dpif *, struct dpif_upcall *, struct ofpbuf *);
+int dpif_handlers_set(struct dpif *, uint32_t n_handlers);
+int dpif_recv(struct dpif *, uint32_t handler_id, struct dpif_upcall *,
+ struct ofpbuf *);
void dpif_recv_purge(struct dpif *);
-void dpif_recv_wait(struct dpif *);
+void dpif_recv_wait(struct dpif *, uint32_t handler_id);
\f
/* Miscellaneous. */
git://git.onelab.eu / sliver-openvswitch.git / blobdiff