-/* Copyright (c) 2013 Nicira, Inc.
+/* Copyright (c) 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.
#include <errno.h>
-#include "ofproto/ofproto-provider.h"
#include "byte-order.h"
+#include "connectivity.h"
#include "dynamic-string.h"
#include "hash.h"
#include "hmap.h"
-#include "netdev-vport.h"
+#include "netdev.h"
#include "odp-util.h"
#include "packets.h"
+#include "seq.h"
#include "smap.h"
#include "socket-util.h"
#include "tunnel.h"
#include "vlog.h"
-/* XXX:
- *
- * Ability to generate actions on input for ECN
- * Ability to generate metadata for packet-outs
- * IPsec using skb mark.
- * VXLAN.
- * Multicast group management (possibly).
- * Disallow netdevs with names like "gre64_system" to prevent collisions. */
-
VLOG_DEFINE_THIS_MODULE(tunnel);
+/* skb mark used for IPsec tunnel packets */
+#define IPSEC_MARK 1
+
struct tnl_match {
ovs_be64 in_key;
ovs_be32 ip_src;
ovs_be32 ip_dst;
- uint32_t odp_port;
+ odp_port_t odp_port;
+ uint32_t pkt_mark;
bool in_key_flow;
+ bool ip_src_flow;
+ bool ip_dst_flow;
};
struct tnl_port {
+ struct hmap_node ofport_node;
struct hmap_node match_node;
- const struct ofport *ofport;
- unsigned int netdev_seq;
+ const struct ofport_dpif *ofport;
+ unsigned int change_seq;
+ struct netdev *netdev;
+
struct tnl_match match;
};
-static struct hmap tnl_match_map = HMAP_INITIALIZER(&tnl_match_map);
+static struct ovs_rwlock rwlock = OVS_RWLOCK_INITIALIZER;
+
+static struct hmap tnl_match_map__ = HMAP_INITIALIZER(&tnl_match_map__);
+static struct hmap *tnl_match_map OVS_GUARDED_BY(rwlock) = &tnl_match_map__;
-/* Returned to callers when their ofport will never be used to receive or send
- * tunnel traffic. Alternatively, we could ask the caller to delete their
- * ofport, but this would be unclean in the reconfguration case. For the first
- * time, an ofproto provider would have to call ofproto_port_del() on itself.*/
-static struct tnl_port void_tnl_port;
+static struct hmap ofport_map__ = HMAP_INITIALIZER(&ofport_map__);
+static struct hmap *ofport_map OVS_GUARDED_BY(rwlock) = &ofport_map__;
static struct vlog_rate_limit rl = VLOG_RATE_LIMIT_INIT(1, 5);
static struct vlog_rate_limit dbg_rl = VLOG_RATE_LIMIT_INIT(60, 60);
-static struct tnl_port *tnl_find(struct tnl_match *);
-static struct tnl_port *tnl_find_exact(struct tnl_match *);
+static struct tnl_port *tnl_find(const struct flow *) OVS_REQ_RDLOCK(rwlock);
+static struct tnl_port *tnl_find_exact(struct tnl_match *)
+ OVS_REQ_RDLOCK(rwlock);
+static struct tnl_port *tnl_find_ofport(const struct ofport_dpif *)
+ OVS_REQ_RDLOCK(rwlock);
+
static uint32_t tnl_hash(struct tnl_match *);
static void tnl_match_fmt(const struct tnl_match *, struct ds *);
-static char *tnl_port_fmt(const struct tnl_port *);
-static void tnl_port_mod_log(const struct tnl_port *, const char *action);
-static const char *tnl_port_get_name(const struct tnl_port *);
-
-static struct tnl_port *
-tnl_port_add__(const struct ofport *ofport, uint32_t odp_port,
- bool warn)
+static char *tnl_port_fmt(const struct tnl_port *) OVS_REQ_RDLOCK(rwlock);
+static void tnl_port_mod_log(const struct tnl_port *, const char *action)
+ OVS_REQ_RDLOCK(rwlock);
+static const char *tnl_port_get_name(const struct tnl_port *)
+ OVS_REQ_RDLOCK(rwlock);
+static void tnl_port_del__(const struct ofport_dpif *) OVS_REQ_WRLOCK(rwlock);
+
+static bool
+tnl_port_add__(const struct ofport_dpif *ofport, const struct netdev *netdev,
+ odp_port_t odp_port, bool warn)
+ OVS_REQ_WRLOCK(rwlock)
{
const struct netdev_tunnel_config *cfg;
struct tnl_port *existing_port;
struct tnl_port *tnl_port;
- cfg = netdev_get_tunnel_config(ofport->netdev);
+ cfg = netdev_get_tunnel_config(netdev);
ovs_assert(cfg);
tnl_port = xzalloc(sizeof *tnl_port);
tnl_port->ofport = ofport;
- tnl_port->netdev_seq = netdev_change_seq(tnl_port->ofport->netdev);
+ tnl_port->netdev = netdev_ref(netdev);
+ tnl_port->change_seq = seq_read(connectivity_seq_get());
tnl_port->match.in_key = cfg->in_key;
tnl_port->match.ip_src = cfg->ip_src;
tnl_port->match.ip_dst = cfg->ip_dst;
+ tnl_port->match.ip_src_flow = cfg->ip_src_flow;
+ tnl_port->match.ip_dst_flow = cfg->ip_dst_flow;
+ tnl_port->match.pkt_mark = cfg->ipsec ? IPSEC_MARK : 0;
tnl_port->match.in_key_flow = cfg->in_key_flow;
tnl_port->match.odp_port = odp_port;
ds_destroy(&ds);
free(tnl_port);
}
- return &void_tnl_port;
+ return false;
}
- hmap_insert(&tnl_match_map, &tnl_port->match_node,
+ hmap_insert(ofport_map, &tnl_port->ofport_node, hash_pointer(ofport, 0));
+ hmap_insert(tnl_match_map, &tnl_port->match_node,
tnl_hash(&tnl_port->match));
tnl_port_mod_log(tnl_port, "adding");
- return tnl_port;
+ return true;
}
/* Adds 'ofport' to the module with datapath port number 'odp_port'. 'ofport's
* must be added before they can be used by the module. 'ofport' must be a
* tunnel. */
-struct tnl_port *
-tnl_port_add(const struct ofport *ofport, uint32_t odp_port)
+void
+tnl_port_add(const struct ofport_dpif *ofport, const struct netdev *netdev,
+ odp_port_t odp_port) OVS_EXCLUDED(rwlock)
{
- return tnl_port_add__(ofport, odp_port, true);
+ ovs_rwlock_wrlock(&rwlock);
+ tnl_port_add__(ofport, netdev, odp_port, true);
+ ovs_rwlock_unlock(&rwlock);
}
-/* Checks if the tnl_port pointed to by 'tnl_portp' needs reconfiguration due
- * to changes in its netdev_tunnel_config. If it does, updates 'tnl_portp' to
- * point to a new tnl_port and returns true. Otherwise, returns false.
- * 'ofport' and 'odp_port' should be the same as would be passed to
+/* Checks if the tunnel represented by 'ofport' reconfiguration due to changes
+ * in its netdev_tunnel_config. If it does, returns true. Otherwise, returns
+ * false. 'ofport' and 'odp_port' should be the same as would be passed to
* tnl_port_add(). */
bool
-tnl_port_reconfigure(const struct ofport *ofport, uint32_t odp_port,
- struct tnl_port **tnl_portp)
+tnl_port_reconfigure(const struct ofport_dpif *ofport,
+ const struct netdev *netdev, odp_port_t odp_port)
+ OVS_EXCLUDED(rwlock)
{
- struct tnl_port *tnl_port = *tnl_portp;
+ struct tnl_port *tnl_port;
+ bool changed = false;
- if (tnl_port == &void_tnl_port) {
- *tnl_portp = tnl_port_add__(ofport, odp_port, false);
- return *tnl_portp != &void_tnl_port;
- } else if (tnl_port->ofport != ofport
+ ovs_rwlock_wrlock(&rwlock);
+ tnl_port = tnl_find_ofport(ofport);
+ if (!tnl_port) {
+ changed = tnl_port_add__(ofport, netdev, odp_port, false);
+ } else if (tnl_port->netdev != netdev
|| tnl_port->match.odp_port != odp_port
- || tnl_port->netdev_seq != netdev_change_seq(ofport->netdev)) {
+ || tnl_port->change_seq != seq_read(connectivity_seq_get())) {
VLOG_DBG("reconfiguring %s", tnl_port_get_name(tnl_port));
- tnl_port_del(tnl_port);
- *tnl_portp = tnl_port_add(ofport, odp_port);
- return true;
+ tnl_port_del__(ofport);
+ tnl_port_add__(ofport, netdev, odp_port, true);
+ changed = true;
}
- return false;
+ ovs_rwlock_unlock(&rwlock);
+ return changed;
}
-/* Removes 'tnl_port' from the module. */
-void
-tnl_port_del(struct tnl_port *tnl_port)
+static void
+tnl_port_del__(const struct ofport_dpif *ofport) OVS_REQ_WRLOCK(rwlock)
{
- if (tnl_port && tnl_port != &void_tnl_port) {
+ struct tnl_port *tnl_port;
+
+ if (!ofport) {
+ return;
+ }
+
+ tnl_port = tnl_find_ofport(ofport);
+ if (tnl_port) {
tnl_port_mod_log(tnl_port, "removing");
- hmap_remove(&tnl_match_map, &tnl_port->match_node);
+ hmap_remove(tnl_match_map, &tnl_port->match_node);
+ hmap_remove(ofport_map, &tnl_port->ofport_node);
+ netdev_close(tnl_port->netdev);
free(tnl_port);
}
}
-/* Transforms 'flow' so that it appears to have been received by a tunnel
- * OpenFlow port controlled by this module instead of the datapath port it
- * actually came in on. Sets 'flow''s in_port to the appropriate OpenFlow port
- * number. Returns the 'ofport' corresponding to the new in_port.
+/* Removes 'ofport' from the module. */
+void
+tnl_port_del(const struct ofport_dpif *ofport) OVS_EXCLUDED(rwlock)
+{
+ ovs_rwlock_wrlock(&rwlock);
+ tnl_port_del__(ofport);
+ ovs_rwlock_unlock(&rwlock);
+}
+
+/* Looks in the table of tunnels for a tunnel matching the metadata in 'flow'.
+ * Returns the 'ofport' corresponding to the new in_port, or a null pointer if
+ * none is found.
*
* Callers should verify that 'flow' needs to be received by calling
- * tnl_port_should_receive() before this function.
- *
- * Leaves 'flow' untouched and returns null if unsuccessful. */
-const struct ofport *
-tnl_port_receive(struct flow *flow)
+ * tnl_port_should_receive() before this function. */
+const struct ofport_dpif *
+tnl_port_receive(const struct flow *flow) OVS_EXCLUDED(rwlock)
{
char *pre_flow_str = NULL;
+ const struct ofport_dpif *ofport;
struct tnl_port *tnl_port;
- struct tnl_match match;
- memset(&match, 0, sizeof match);
- match.odp_port = flow->in_port;
- match.ip_src = flow->tunnel.ip_dst;
- match.ip_dst = flow->tunnel.ip_src;
- match.in_key = flow->tunnel.tun_id;
-
- tnl_port = tnl_find(&match);
+ ovs_rwlock_rdlock(&rwlock);
+ tnl_port = tnl_find(flow);
+ ofport = tnl_port ? tnl_port->ofport : NULL;
if (!tnl_port) {
- struct ds ds = DS_EMPTY_INITIALIZER;
+ char *flow_str = flow_to_string(flow);
- tnl_match_fmt(&match, &ds);
- VLOG_WARN_RL(&rl, "receive tunnel port not found (%s)", ds_cstr(&ds));
- ds_destroy(&ds);
- return NULL;
- }
-
- if (is_ip_any(flow)
- && ((flow->tunnel.ip_tos & IP_ECN_MASK) == IP_ECN_CE)
- && (flow->nw_tos & IP_ECN_MASK) == IP_ECN_NOT_ECT) {
- VLOG_WARN_RL(&rl, "dropping tunnel packet marked ECN CE but is not ECN"
- " capable");
- return NULL;
+ VLOG_WARN_RL(&rl, "receive tunnel port not found (%s)", flow_str);
+ free(flow_str);
+ goto out;
}
if (!VLOG_DROP_DBG(&dbg_rl)) {
pre_flow_str = flow_to_string(flow);
}
- flow->in_port = tnl_port->ofport->ofp_port;
- memset(&flow->tunnel, 0, sizeof flow->tunnel);
- flow->tunnel.tun_id = match.in_key;
-
if (pre_flow_str) {
char *post_flow_str = flow_to_string(flow);
char *tnl_str = tnl_port_fmt(tnl_port);
free(pre_flow_str);
free(post_flow_str);
}
- return tnl_port->ofport;
+
+out:
+ ovs_rwlock_unlock(&rwlock);
+ return ofport;
+}
+
+static bool
+tnl_ecn_ok(const struct flow *base_flow, struct flow *flow)
+{
+ if (is_ip_any(base_flow)
+ && (flow->tunnel.ip_tos & IP_ECN_MASK) == IP_ECN_CE) {
+ if ((base_flow->nw_tos & IP_ECN_MASK) == IP_ECN_NOT_ECT) {
+ VLOG_WARN_RL(&rl, "dropping tunnel packet marked ECN CE"
+ " but is not ECN capable");
+ return false;
+ } else {
+ /* Set the ECN CE value in the tunneled packet. */
+ flow->nw_tos |= IP_ECN_CE;
+ }
+ }
+
+ return true;
+}
+
+/* Should be called at the beginning of action translation to initialize
+ * wildcards and perform any actions based on receiving on tunnel port.
+ *
+ * Returns false if the packet must be dropped. */
+bool
+tnl_xlate_init(const struct flow *base_flow, struct flow *flow,
+ struct flow_wildcards *wc)
+{
+ if (tnl_port_should_receive(flow)) {
+ memset(&wc->masks.tunnel, 0xff, sizeof wc->masks.tunnel);
+ wc->masks.tunnel.flags = (FLOW_TNL_F_DONT_FRAGMENT |
+ FLOW_TNL_F_CSUM |
+ FLOW_TNL_F_KEY);
+ memset(&wc->masks.pkt_mark, 0xff, sizeof wc->masks.pkt_mark);
+
+ if (!tnl_ecn_ok(base_flow, flow)) {
+ return false;
+ }
+
+ flow->pkt_mark &= ~IPSEC_MARK;
+ }
+
+ return true;
}
/* Given that 'flow' should be output to the ofport corresponding to
* 'tnl_port', updates 'flow''s tunnel headers and returns the actual datapath
- * port that the output should happen on. May return OVSP_NONE if the output
+ * port that the output should happen on. May return ODPP_NONE if the output
* shouldn't occur. */
-uint32_t
-tnl_port_send(const struct tnl_port *tnl_port, struct flow *flow)
+odp_port_t
+tnl_port_send(const struct ofport_dpif *ofport, struct flow *flow,
+ struct flow_wildcards *wc) OVS_EXCLUDED(rwlock)
{
const struct netdev_tunnel_config *cfg;
+ struct tnl_port *tnl_port;
char *pre_flow_str = NULL;
+ odp_port_t out_port;
- if (tnl_port == &void_tnl_port) {
- return OVSP_NONE;
+ ovs_rwlock_rdlock(&rwlock);
+ tnl_port = tnl_find_ofport(ofport);
+ out_port = tnl_port ? tnl_port->match.odp_port : ODPP_NONE;
+ if (!tnl_port) {
+ goto out;
}
- cfg = netdev_get_tunnel_config(tnl_port->ofport->netdev);
+ cfg = netdev_get_tunnel_config(tnl_port->netdev);
ovs_assert(cfg);
if (!VLOG_DROP_DBG(&dbg_rl)) {
pre_flow_str = flow_to_string(flow);
}
- flow->tunnel.ip_src = tnl_port->match.ip_src;
- flow->tunnel.ip_dst = tnl_port->match.ip_dst;
+ if (!cfg->ip_src_flow) {
+ flow->tunnel.ip_src = tnl_port->match.ip_src;
+ }
+ if (!cfg->ip_dst_flow) {
+ flow->tunnel.ip_dst = tnl_port->match.ip_dst;
+ }
+ flow->pkt_mark = tnl_port->match.pkt_mark;
if (!cfg->out_key_flow) {
flow->tunnel.tun_id = cfg->out_key;
}
if (cfg->ttl_inherit && is_ip_any(flow)) {
+ wc->masks.nw_ttl = 0xff;
flow->tunnel.ip_ttl = flow->nw_ttl;
} else {
flow->tunnel.ip_ttl = cfg->ttl;
}
if (cfg->tos_inherit && is_ip_any(flow)) {
+ wc->masks.nw_tos = 0xff;
flow->tunnel.ip_tos = flow->nw_tos & IP_DSCP_MASK;
} else {
flow->tunnel.ip_tos = cfg->tos;
}
+ /* ECN fields are always inherited. */
+ if (is_ip_any(flow)) {
+ wc->masks.nw_tos |= IP_ECN_MASK;
+ }
+
if ((flow->nw_tos & IP_ECN_MASK) == IP_ECN_CE) {
flow->tunnel.ip_tos |= IP_ECN_ECT_0;
} else {
free(post_flow_str);
}
- return tnl_port->match.odp_port;
+out:
+ ovs_rwlock_unlock(&rwlock);
+ return out_port;
}
static uint32_t
}
static struct tnl_port *
-tnl_find_exact(struct tnl_match *match)
+tnl_find_ofport(const struct ofport_dpif *ofport) OVS_REQ_RDLOCK(rwlock)
{
struct tnl_port *tnl_port;
- HMAP_FOR_EACH_WITH_HASH (tnl_port, match_node, tnl_hash(match),
- &tnl_match_map) {
- if (!memcmp(match, &tnl_port->match, sizeof *match)) {
+ HMAP_FOR_EACH_IN_BUCKET (tnl_port, ofport_node, hash_pointer(ofport, 0),
+ ofport_map) {
+ if (tnl_port->ofport == ofport) {
return tnl_port;
}
}
}
static struct tnl_port *
-tnl_find(struct tnl_match *match_)
+tnl_find_exact(struct tnl_match *match) OVS_REQ_RDLOCK(rwlock)
{
- struct tnl_match match = *match_;
- bool is_multicast = ip_is_multicast(match.ip_src);
struct tnl_port *tnl_port;
- /* remote_ip, local_ip, in_key */
- if (!is_multicast) {
- tnl_port = tnl_find_exact(&match);
- if (tnl_port) {
+ HMAP_FOR_EACH_WITH_HASH (tnl_port, match_node, tnl_hash(match),
+ tnl_match_map) {
+ if (!memcmp(match, &tnl_port->match, sizeof *match)) {
return tnl_port;
}
}
+ return NULL;
+}
- /* remote_ip, in_key */
- match.ip_src = 0;
- tnl_port = tnl_find_exact(&match);
- if (tnl_port) {
- return tnl_port;
- }
- match.ip_src = match_->ip_src;
+/* Returns the tnl_port that is the best match for the tunnel data in 'flow',
+ * or NULL if no tnl_port matches 'flow'. */
+static struct tnl_port *
+tnl_find(const struct flow *flow) OVS_REQ_RDLOCK(rwlock)
+{
+ enum ip_src_type {
+ IP_SRC_CFG, /* ip_src must equal configured address. */
+ IP_SRC_ANY, /* Any ip_src is acceptable. */
+ IP_SRC_FLOW /* ip_src is handled in flow table. */
+ };
+
+ struct tnl_match_pattern {
+ bool in_key_flow;
+ bool ip_dst_flow;
+ enum ip_src_type ip_src;
+ };
+
+ static const struct tnl_match_pattern patterns[] = {
+ { false, false, IP_SRC_CFG }, /* remote_ip, local_ip, in_key. */
+ { false, false, IP_SRC_ANY }, /* remote_ip, in_key. */
+ { true, false, IP_SRC_CFG }, /* remote_ip, local_ip. */
+ { true, false, IP_SRC_ANY }, /* remote_ip. */
+ { true, true, IP_SRC_ANY }, /* Flow-based remote. */
+ { true, true, IP_SRC_FLOW }, /* Flow-based everything. */
+ };
+
+ const struct tnl_match_pattern *p;
+ struct tnl_match match;
- /* remote_ip, local_ip */
- if (!is_multicast) {
- match.in_key = 0;
- match.in_key_flow = true;
- tnl_port = tnl_find_exact(&match);
- if (tnl_port) {
- return tnl_port;
- }
- match.in_key = match_->in_key;
- match.in_key_flow = false;
- }
+ memset(&match, 0, sizeof match);
+ match.odp_port = flow->in_port.odp_port;
+ match.pkt_mark = flow->pkt_mark;
- /* remote_ip */
- match.ip_src = 0;
- match.in_key = 0;
- match.in_key_flow = true;
- tnl_port = tnl_find_exact(&match);
- if (tnl_port) {
- return tnl_port;
- }
- match.ip_src = match_->ip_src;
- match.in_key = match_->in_key;
- match.in_key_flow = false;
+ for (p = patterns; p < &patterns[ARRAY_SIZE(patterns)]; p++) {
+ struct tnl_port *tnl_port;
- if (is_multicast) {
- match.ip_src = 0;
- match.ip_dst = match_->ip_src;
+ match.in_key_flow = p->in_key_flow;
+ match.in_key = p->in_key_flow ? 0 : flow->tunnel.tun_id;
- /* multicast remote_ip, in_key */
- tnl_port = tnl_find_exact(&match);
- if (tnl_port) {
- return tnl_port;
- }
+ match.ip_dst_flow = p->ip_dst_flow;
+ match.ip_dst = p->ip_dst_flow ? 0 : flow->tunnel.ip_src;
+
+ match.ip_src_flow = p->ip_src == IP_SRC_FLOW;
+ match.ip_src = p->ip_src == IP_SRC_CFG ? flow->tunnel.ip_dst : 0;
- /* multicast remote_ip */
- match.in_key = 0;
- match.in_key_flow = true;
tnl_port = tnl_find_exact(&match);
if (tnl_port) {
return tnl_port;
}
}
+
return NULL;
}
static void
tnl_match_fmt(const struct tnl_match *match, struct ds *ds)
+ OVS_REQ_RDLOCK(rwlock)
{
- ds_put_format(ds, IP_FMT"->"IP_FMT, IP_ARGS(match->ip_src),
- IP_ARGS(match->ip_dst));
+ if (!match->ip_dst_flow) {
+ ds_put_format(ds, IP_FMT"->"IP_FMT, IP_ARGS(match->ip_src),
+ IP_ARGS(match->ip_dst));
+ } else if (!match->ip_src_flow) {
+ ds_put_format(ds, IP_FMT"->flow", IP_ARGS(match->ip_src));
+ } else {
+ ds_put_cstr(ds, "flow->flow");
+ }
if (match->in_key_flow) {
ds_put_cstr(ds, ", key=flow");
}
ds_put_format(ds, ", dp port=%"PRIu32, match->odp_port);
+ ds_put_format(ds, ", pkt mark=%"PRIu32, match->pkt_mark);
}
static void
tnl_port_mod_log(const struct tnl_port *tnl_port, const char *action)
+ OVS_REQ_RDLOCK(rwlock)
{
if (VLOG_IS_DBG_ENABLED()) {
struct ds ds = DS_EMPTY_INITIALIZER;
}
static char *
-tnl_port_fmt(const struct tnl_port *tnl_port)
+tnl_port_fmt(const struct tnl_port *tnl_port) OVS_REQ_RDLOCK(rwlock)
{
const struct netdev_tunnel_config *cfg =
- netdev_get_tunnel_config(tnl_port->ofport->netdev);
+ netdev_get_tunnel_config(tnl_port->netdev);
struct ds ds = DS_EMPTY_INITIALIZER;
ds_put_format(&ds, "port %"PRIu32": %s (%s: ", tnl_port->match.odp_port,
tnl_port_get_name(tnl_port),
- netdev_get_type(tnl_port->ofport->netdev));
+ netdev_get_type(tnl_port->netdev));
tnl_match_fmt(&tnl_port->match, &ds);
if (cfg->out_key != cfg->in_key ||
}
static const char *
-tnl_port_get_name(const struct tnl_port *tnl_port)
+tnl_port_get_name(const struct tnl_port *tnl_port) OVS_REQ_RDLOCK(rwlock)
{
- return netdev_get_name(tnl_port->ofport->netdev);
+ return netdev_get_name(tnl_port->netdev);
}
projects / cascardo / ovs.git / blobdiff