/*
- * Copyright (c) 2008, 2009, 2010, 2011, 2012, 2013, 2014 Nicira, Inc.
+ * Copyright (c) 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016 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 <inttypes.h>
#include <net/if.h>
#include <linux/types.h>
+#include <linux/ip.h>
+#include <linux/if_tunnel.h>
#include <linux/pkt_sched.h>
+#include <linux/rtnetlink.h>
#include <poll.h>
#include <stdlib.h>
#include <strings.h>
#include "bitmap.h"
#include "dpif-provider.h"
-#include "dynamic-string.h"
+#include "openvswitch/dynamic-string.h"
#include "flow.h"
#include "fat-rwlock.h"
#include "netdev.h"
#include "netdev-linux.h"
#include "netdev-vport.h"
+#include "netlink-conntrack.h"
#include "netlink-notifier.h"
#include "netlink-socket.h"
#include "netlink.h"
#include "odp-util.h"
-#include "ofpbuf.h"
+#include "openvswitch/ofpbuf.h"
#include "packets.h"
#include "poll-loop.h"
#include "random.h"
#include "timeval.h"
#include "unaligned.h"
#include "util.h"
-#include "vlog.h"
+#include "openvswitch/vlog.h"
VLOG_DEFINE_THIS_MODULE(dpif_netlink);
#ifdef _WIN32
struct dpif_flow_stats *);
static void dpif_netlink_flow_to_dpif_flow(struct dpif *, struct dpif_flow *,
const struct dpif_netlink_flow *);
-static bool dpif_netlink_check_ufid__(struct dpif *dpif);
-static bool dpif_netlink_check_ufid(struct dpif *dpif);
/* One of the dpif channels between the kernel and userspace. */
struct dpif_channel {
/* Change notification. */
struct nl_sock *port_notifier; /* vport multicast group subscriber. */
bool refresh_channels;
-
- /* If the datapath supports indexing flows using unique identifiers, then
- * we can reduce the size of netlink messages by omitting fields like the
- * flow key during flow operations. */
- bool ufid_supported;
};
static void report_loss(struct dpif_netlink *, struct dpif_channel *,
dp->dp_ifindex, dp->dp_ifindex);
dpif->dp_ifindex = dp->dp_ifindex;
- dpif->ufid_supported = dpif_netlink_check_ufid__(&dpif->dpif);
*dpifp = &dpif->dpif;
return 0;
vport_request.cmd = OVS_VPORT_CMD_SET;
vport_request.dp_ifindex = dpif->dp_ifindex;
vport_request.port_no = u32_to_odp(i);
+ vport_request.n_upcall_pids = 1;
vport_request.upcall_pids = &upcall_pids;
dpif_netlink_vport_transact(&vport_request, NULL, NULL);
case OVS_VPORT_TYPE_GRE:
return "gre";
- case OVS_VPORT_TYPE_GRE64:
- return "gre64";
-
case OVS_VPORT_TYPE_VXLAN:
return "vxlan";
case OVS_VPORT_TYPE_LISP:
return "lisp";
+ case OVS_VPORT_TYPE_STT:
+ return "stt";
+
case OVS_VPORT_TYPE_UNSPEC:
case __OVS_VPORT_TYPE_MAX:
break;
}
static enum ovs_vport_type
-netdev_to_ovs_vport_type(const struct netdev *netdev)
+netdev_to_ovs_vport_type(const char *type)
{
- const char *type = netdev_get_type(netdev);
-
if (!strcmp(type, "tap") || !strcmp(type, "system")) {
return OVS_VPORT_TYPE_NETDEV;
} else if (!strcmp(type, "internal")) {
return OVS_VPORT_TYPE_INTERNAL;
+ } else if (strstr(type, "stt")) {
+ return OVS_VPORT_TYPE_STT;
} else if (!strcmp(type, "geneve")) {
return OVS_VPORT_TYPE_GENEVE;
- } else if (strstr(type, "gre64")) {
- return OVS_VPORT_TYPE_GRE64;
} else if (strstr(type, "gre")) {
return OVS_VPORT_TYPE_GRE;
} else if (!strcmp(type, "vxlan")) {
}
static int
-dpif_netlink_port_add__(struct dpif_netlink *dpif, struct netdev *netdev,
+dpif_netlink_port_add__(struct dpif_netlink *dpif, const char *name,
+ enum ovs_vport_type type,
+ struct ofpbuf *options,
odp_port_t *port_nop)
OVS_REQ_WRLOCK(dpif->upcall_lock)
{
- const struct netdev_tunnel_config *tnl_cfg;
- char namebuf[NETDEV_VPORT_NAME_BUFSIZE];
- const char *name = netdev_vport_get_dpif_port(netdev,
- namebuf, sizeof namebuf);
- const char *type = netdev_get_type(netdev);
struct dpif_netlink_vport request, reply;
struct ofpbuf *buf;
- uint64_t options_stub[64 / 8];
- struct ofpbuf options;
struct nl_sock **socksp = NULL;
uint32_t *upcall_pids;
int error = 0;
dpif_netlink_vport_init(&request);
request.cmd = OVS_VPORT_CMD_NEW;
request.dp_ifindex = dpif->dp_ifindex;
- request.type = netdev_to_ovs_vport_type(netdev);
- if (request.type == OVS_VPORT_TYPE_UNSPEC) {
- VLOG_WARN_RL(&error_rl, "%s: cannot create port `%s' because it has "
- "unsupported type `%s'",
- dpif_name(&dpif->dpif), name, type);
- vport_del_socksp(dpif, socksp);
- return EINVAL;
- }
+ request.type = type;
request.name = name;
- if (request.type == OVS_VPORT_TYPE_NETDEV) {
-#ifdef _WIN32
- /* XXX : Map appropiate Windows handle */
-#else
- netdev_linux_ethtool_set_flag(netdev, ETH_FLAG_LRO, "LRO", false);
-#endif
- }
-
- tnl_cfg = netdev_get_tunnel_config(netdev);
- if (tnl_cfg && tnl_cfg->dst_port != 0) {
- ofpbuf_use_stack(&options, options_stub, sizeof options_stub);
- nl_msg_put_u16(&options, OVS_TUNNEL_ATTR_DST_PORT,
- ntohs(tnl_cfg->dst_port));
- request.options = ofpbuf_data(&options);
- request.options_len = ofpbuf_size(&options);
- }
-
request.port_no = *port_nop;
upcall_pids = vport_socksp_to_pids(socksp, dpif->n_handlers);
request.n_upcall_pids = socksp ? dpif->n_handlers : 1;
request.upcall_pids = upcall_pids;
+ if (options) {
+ request.options = options->data;
+ request.options_len = options->size;
+ }
+
error = dpif_netlink_vport_transact(&request, &reply, &buf);
if (!error) {
*port_nop = reply.port_no;
return error;
}
+static int
+dpif_netlink_port_add_compat(struct dpif_netlink *dpif, struct netdev *netdev,
+ odp_port_t *port_nop)
+ OVS_REQ_WRLOCK(dpif->upcall_lock)
+{
+ const struct netdev_tunnel_config *tnl_cfg;
+ char namebuf[NETDEV_VPORT_NAME_BUFSIZE];
+ const char *name = netdev_vport_get_dpif_port(netdev,
+ namebuf, sizeof namebuf);
+ const char *type = netdev_get_type(netdev);
+ uint64_t options_stub[64 / 8];
+ struct ofpbuf options;
+ enum ovs_vport_type ovs_type;
+
+ ovs_type = netdev_to_ovs_vport_type(netdev_get_type(netdev));
+ if (ovs_type == OVS_VPORT_TYPE_UNSPEC) {
+ VLOG_WARN_RL(&error_rl, "%s: cannot create port `%s' because it has "
+ "unsupported type `%s'",
+ dpif_name(&dpif->dpif), name, type);
+ return EINVAL;
+ }
+
+ if (ovs_type == OVS_VPORT_TYPE_NETDEV) {
+#ifdef _WIN32
+ /* XXX : Map appropiate Windows handle */
+#else
+ netdev_linux_ethtool_set_flag(netdev, ETH_FLAG_LRO, "LRO", false);
+#endif
+ }
+
+ tnl_cfg = netdev_get_tunnel_config(netdev);
+ if (tnl_cfg && (tnl_cfg->dst_port != 0 || tnl_cfg->exts)) {
+ ofpbuf_use_stack(&options, options_stub, sizeof options_stub);
+ if (tnl_cfg->dst_port) {
+ nl_msg_put_u16(&options, OVS_TUNNEL_ATTR_DST_PORT,
+ ntohs(tnl_cfg->dst_port));
+ }
+ if (tnl_cfg->exts) {
+ size_t ext_ofs;
+ int i;
+
+ ext_ofs = nl_msg_start_nested(&options, OVS_TUNNEL_ATTR_EXTENSION);
+ for (i = 0; i < 32; i++) {
+ if (tnl_cfg->exts & (1 << i)) {
+ nl_msg_put_flag(&options, i);
+ }
+ }
+ nl_msg_end_nested(&options, ext_ofs);
+ }
+ return dpif_netlink_port_add__(dpif, name, ovs_type, &options, port_nop);
+ } else {
+ return dpif_netlink_port_add__(dpif, name, ovs_type, NULL, port_nop);
+ }
+
+}
+
+#ifdef __linux__
+
+static int
+netdev_linux_destroy(const char *name)
+{
+ int err;
+ struct ofpbuf request, *reply;
+
+ ofpbuf_init(&request, 0);
+ nl_msg_put_nlmsghdr(&request, 0, RTM_DELLINK,
+ NLM_F_REQUEST | NLM_F_ACK);
+ ofpbuf_put_zeros(&request, sizeof(struct ifinfomsg));
+ nl_msg_put_string(&request, IFLA_IFNAME, name);
+
+ err = nl_transact(NETLINK_ROUTE, &request, &reply);
+
+ if (!err) {
+ ofpbuf_uninit(reply);
+ }
+
+ ofpbuf_uninit(&request);
+ return err;
+}
+
+static int
+netdev_vxlan_destroy(const char *name)
+{
+ return netdev_linux_destroy(name);
+}
+
+static int
+netdev_gre_destroy(const char *name)
+{
+ return netdev_linux_destroy(name);
+}
+
+static int
+netdev_geneve_destroy(const char *name)
+{
+ return netdev_linux_destroy(name);
+}
+
+/*
+ * On some older systems, these enums are not defined.
+ */
+
+#ifndef IFLA_VXLAN_MAX
+#define IFLA_VXLAN_MAX 0
+#define IFLA_VXLAN_PORT 15
+#endif
+#if IFLA_VXLAN_MAX < 20
+#define IFLA_VXLAN_UDP_ZERO_CSUM6_RX 20
+#define IFLA_VXLAN_GBP 23
+#define IFLA_VXLAN_COLLECT_METADATA 25
+#endif
+
+#if IFLA_GRE_MAX < 18
+#define IFLA_GRE_COLLECT_METADATA 18
+#endif
+
+#ifndef IFLA_GENEVE_MAX
+#define IFLA_GENEVE_MAX 0
+#define IFLA_GENEVE_PORT 5
+#endif
+
+#if IFLA_GENEVE_MAX < 6
+#define IFLA_GENEVE_COLLECT_METADATA 6
+#endif
+#if IFLA_GENEVE_MAX < 10
+#define IFLA_GENEVE_UDP_ZERO_CSUM6_RX 10
+#endif
+
+static int
+netdev_vxlan_create(struct netdev *netdev)
+{
+ int err;
+ struct ofpbuf request, *reply;
+ size_t linkinfo_off, infodata_off;
+ char namebuf[NETDEV_VPORT_NAME_BUFSIZE];
+ const char *name = netdev_vport_get_dpif_port(netdev,
+ namebuf, sizeof namebuf);
+ struct ifinfomsg *ifinfo;
+ const struct netdev_tunnel_config *tnl_cfg;
+ tnl_cfg = netdev_get_tunnel_config(netdev);
+ if (!tnl_cfg) { /* or assert? */
+ return EINVAL;
+ }
+
+ ofpbuf_init(&request, 0);
+ nl_msg_put_nlmsghdr(&request, 0, RTM_NEWLINK,
+ NLM_F_REQUEST | NLM_F_ACK | NLM_F_CREATE);
+ ifinfo = ofpbuf_put_zeros(&request, sizeof(struct ifinfomsg));
+ ifinfo->ifi_change = ifinfo->ifi_flags = IFF_UP;
+ nl_msg_put_string(&request, IFLA_IFNAME, name);
+ nl_msg_put_u32(&request, IFLA_MTU, UINT16_MAX);
+ linkinfo_off = nl_msg_start_nested(&request, IFLA_LINKINFO);
+ nl_msg_put_string(&request, IFLA_INFO_KIND, "vxlan");
+ infodata_off = nl_msg_start_nested(&request, IFLA_INFO_DATA);
+ nl_msg_put_u8(&request, IFLA_VXLAN_LEARNING, 0);
+ nl_msg_put_u8(&request, IFLA_VXLAN_COLLECT_METADATA, 1);
+ nl_msg_put_u8(&request, IFLA_VXLAN_UDP_ZERO_CSUM6_RX, 1);
+ if (tnl_cfg->exts & (1 << OVS_VXLAN_EXT_GBP)) {
+ nl_msg_put_flag(&request, IFLA_VXLAN_GBP);
+ }
+ nl_msg_put_be16(&request, IFLA_VXLAN_PORT, tnl_cfg->dst_port);
+ nl_msg_end_nested(&request, infodata_off);
+ nl_msg_end_nested(&request, linkinfo_off);
+
+ err = nl_transact(NETLINK_ROUTE, &request, &reply);
+
+ if (!err) {
+ ofpbuf_uninit(reply);
+ }
+
+ /*
+ * Linux versions older than 4.3 will return EINVAL in case the VID is not
+ * set, which is sufficient to verify COLLECT_METADATA is supported.
+ */
+ if (err == EINVAL) {
+ err = EOPNOTSUPP;
+ }
+
+ ofpbuf_uninit(&request);
+ return err;
+}
+
+/*
+ * On some Linux versions, creating the device with IFLA_GRE_COLLECT_METADATA
+ * will succeed, even though that attribute is not supported. We need to verify
+ * the device has been created with that attribute. In case it has not, we
+ * destroy it and use the compat code.
+ */
+static int
+netdev_gre_verify(const char *name)
+{
+ int err;
+ struct ofpbuf request, *reply;
+ struct ifinfomsg *ifmsg;
+
+ static const struct nl_policy rtlink_policy[] = {
+ [IFLA_LINKINFO] = { .type = NL_A_NESTED },
+ };
+ static const struct nl_policy linkinfo_policy[] = {
+ [IFLA_INFO_KIND] = { .type = NL_A_STRING },
+ [IFLA_INFO_DATA] = { .type = NL_A_NESTED },
+ };
+ static const struct nl_policy gre_policy[] = {
+ [IFLA_GRE_COLLECT_METADATA] = { .type = NL_A_FLAG },
+ };
+
+ ofpbuf_init(&request, 0);
+ nl_msg_put_nlmsghdr(&request, 0, RTM_GETLINK,
+ NLM_F_REQUEST);
+ ofpbuf_put_zeros(&request, sizeof(struct ifinfomsg));
+ nl_msg_put_string(&request, IFLA_IFNAME, name);
+
+ err = nl_transact(NETLINK_ROUTE, &request, &reply);
+ if (!err) {
+ struct nlattr *rtlink[ARRAY_SIZE(rtlink_policy)];
+ struct nlattr *linkinfo[ARRAY_SIZE(linkinfo_policy)];
+ struct nlattr *gre[ARRAY_SIZE(gre_policy)];
+
+ err = EINVAL;
+ ifmsg = ofpbuf_at(reply, NLMSG_HDRLEN, sizeof *ifmsg);
+ if (nl_policy_parse(reply, NLMSG_HDRLEN + sizeof *ifmsg,
+ rtlink_policy, rtlink, ARRAY_SIZE(rtlink_policy))) {
+ if (nl_parse_nested(rtlink[IFLA_LINKINFO], linkinfo_policy,
+ linkinfo, ARRAY_SIZE(linkinfo_policy)) &&
+ !strcmp(nl_attr_get_string(linkinfo[IFLA_INFO_KIND]),
+ "gretap")) {
+ if (nl_parse_nested(linkinfo[IFLA_INFO_DATA], gre_policy, gre,
+ ARRAY_SIZE(gre_policy)) &&
+ nl_attr_get_flag(gre[IFLA_GRE_COLLECT_METADATA])) {
+ err = 0;
+ }
+ }
+ }
+ ofpbuf_uninit(reply);
+ }
+ ofpbuf_uninit(&request);
+ return err;
+}
+
+static int
+netdev_gre_create(struct netdev *netdev)
+{
+ int err;
+ struct ofpbuf request, *reply;
+ size_t linkinfo_off, infodata_off;
+ char namebuf[NETDEV_VPORT_NAME_BUFSIZE];
+ const char *name = netdev_vport_get_dpif_port(netdev,
+ namebuf, sizeof namebuf);
+ struct ifinfomsg *ifinfo;
+ const struct netdev_tunnel_config *tnl_cfg;
+ tnl_cfg = netdev_get_tunnel_config(netdev);
+ if (!tnl_cfg) { /* or assert? */
+ return EINVAL;
+ }
+
+ ofpbuf_init(&request, 0);
+ nl_msg_put_nlmsghdr(&request, 0, RTM_NEWLINK,
+ NLM_F_REQUEST | NLM_F_ACK | NLM_F_CREATE);
+ ifinfo = ofpbuf_put_zeros(&request, sizeof(struct ifinfomsg));
+ ifinfo->ifi_change = ifinfo->ifi_flags = IFF_UP;
+ nl_msg_put_string(&request, IFLA_IFNAME, name);
+ nl_msg_put_u32(&request, IFLA_MTU, UINT16_MAX);
+ linkinfo_off = nl_msg_start_nested(&request, IFLA_LINKINFO);
+ nl_msg_put_string(&request, IFLA_INFO_KIND, "gretap");
+ infodata_off = nl_msg_start_nested(&request, IFLA_INFO_DATA);
+ nl_msg_put_flag(&request, IFLA_GRE_COLLECT_METADATA);
+ nl_msg_end_nested(&request, infodata_off);
+ nl_msg_end_nested(&request, linkinfo_off);
+
+ err = nl_transact(NETLINK_ROUTE, &request, &reply);
+
+ if (!err) {
+ ofpbuf_uninit(reply);
+ }
+
+ if (!err && (err = netdev_gre_verify(name))) {
+ netdev_gre_destroy(name);
+ }
+
+ /*
+ * If tunnel metadata is not supported, EEXIST will be returned for zero
+ * addresses tunnel. We still need to verify metadata has been set as above.
+ */
+ if (err == EINVAL || err == EEXIST) {
+ err = EOPNOTSUPP;
+ }
+
+ ofpbuf_uninit(&request);
+ return err;
+}
+
+static int
+netdev_geneve_create(struct netdev *netdev)
+{
+ int err;
+ struct ofpbuf request, *reply;
+ size_t linkinfo_off, infodata_off;
+ char namebuf[NETDEV_VPORT_NAME_BUFSIZE];
+ const char *name = netdev_vport_get_dpif_port(netdev,
+ namebuf, sizeof namebuf);
+ struct ifinfomsg *ifinfo;
+ const struct netdev_tunnel_config *tnl_cfg;
+ tnl_cfg = netdev_get_tunnel_config(netdev);
+ if (!tnl_cfg) { /* or assert? */
+ return EINVAL;
+ }
+
+ ofpbuf_init(&request, 0);
+ nl_msg_put_nlmsghdr(&request, 0, RTM_NEWLINK,
+ NLM_F_REQUEST | NLM_F_ACK | NLM_F_CREATE);
+ ifinfo = ofpbuf_put_zeros(&request, sizeof(struct ifinfomsg));
+ ifinfo->ifi_change = ifinfo->ifi_flags = IFF_UP;
+ nl_msg_put_string(&request, IFLA_IFNAME, name);
+ nl_msg_put_u32(&request, IFLA_MTU, UINT16_MAX);
+ linkinfo_off = nl_msg_start_nested(&request, IFLA_LINKINFO);
+ nl_msg_put_string(&request, IFLA_INFO_KIND, "geneve");
+ infodata_off = nl_msg_start_nested(&request, IFLA_INFO_DATA);
+ nl_msg_put_flag(&request, IFLA_GENEVE_COLLECT_METADATA);
+ nl_msg_put_u8(&request, IFLA_GENEVE_UDP_ZERO_CSUM6_RX, 1);
+ nl_msg_put_be16(&request, IFLA_GENEVE_PORT, tnl_cfg->dst_port);
+ nl_msg_end_nested(&request, infodata_off);
+ nl_msg_end_nested(&request, linkinfo_off);
+
+ err = nl_transact(NETLINK_ROUTE, &request, &reply);
+
+ if (!err) {
+ ofpbuf_uninit(reply);
+ }
+
+ /*
+ * Linux versions older than 4.3 will return EINVAL in case the GENEVE_ID is
+ * not set, which is sufficient to verify COLLECT_METADATA is supported.
+ */
+ if (err == EINVAL) {
+ err = EOPNOTSUPP;
+ }
+
+ ofpbuf_uninit(&request);
+ return err;
+}
+
+#else
+
+static int
+netdev_vxlan_create(struct netdev *netdev OVS_UNUSED)
+{
+ return EOPNOTSUPP;
+}
+
+static int
+netdev_gre_create(struct netdev *netdev OVS_UNUSED)
+{
+ return EOPNOTSUPP;
+}
+
+static int
+netdev_geneve_create(struct netdev *netdev OVS_UNUSED)
+{
+ return EOPNOTSUPP;
+}
+
+static int
+netdev_vxlan_destroy(const char *name OVS_UNUSED)
+{
+ return EOPNOTSUPP;
+}
+
+static int
+netdev_gre_destroy(const char *name OVS_UNUSED)
+{
+ return EOPNOTSUPP;
+}
+
+static int
+netdev_geneve_destroy(const char *name OVS_UNUSED)
+{
+ return EOPNOTSUPP;
+}
+
+#endif
+
+static int
+dpif_netlink_port_query__(const struct dpif_netlink *dpif, odp_port_t port_no,
+ const char *port_name, struct dpif_port *dpif_port);
+
+static int
+dpif_netlink_port_create(struct netdev *netdev)
+{
+ switch (netdev_to_ovs_vport_type(netdev_get_type(netdev))) {
+ case OVS_VPORT_TYPE_VXLAN:
+ return netdev_vxlan_create(netdev);
+ case OVS_VPORT_TYPE_GRE:
+ return netdev_gre_create(netdev);
+ case OVS_VPORT_TYPE_GENEVE:
+ return netdev_geneve_create(netdev);
+ case OVS_VPORT_TYPE_NETDEV:
+ case OVS_VPORT_TYPE_INTERNAL:
+ case OVS_VPORT_TYPE_LISP:
+ case OVS_VPORT_TYPE_STT:
+ case OVS_VPORT_TYPE_UNSPEC:
+ case __OVS_VPORT_TYPE_MAX:
+ default:
+ return EOPNOTSUPP;
+ }
+ return 0;
+}
+
+static int
+dpif_netlink_port_destroy(const char *name, const char *type)
+{
+ switch (netdev_to_ovs_vport_type(type)) {
+ case OVS_VPORT_TYPE_VXLAN:
+ return netdev_vxlan_destroy(name);
+ case OVS_VPORT_TYPE_GRE:
+ return netdev_gre_destroy(name);
+ case OVS_VPORT_TYPE_GENEVE:
+ return netdev_geneve_destroy(name);
+ case OVS_VPORT_TYPE_NETDEV:
+ case OVS_VPORT_TYPE_INTERNAL:
+ case OVS_VPORT_TYPE_LISP:
+ case OVS_VPORT_TYPE_STT:
+ case OVS_VPORT_TYPE_UNSPEC:
+ case __OVS_VPORT_TYPE_MAX:
+ default:
+ return EOPNOTSUPP;
+ }
+ return 0;
+}
+
+static int
+dpif_netlink_port_create_and_add(struct dpif_netlink *dpif, struct netdev *netdev,
+ odp_port_t *port_nop)
+ OVS_REQ_WRLOCK(dpif->upcall_lock)
+{
+ int error;
+ char namebuf[NETDEV_VPORT_NAME_BUFSIZE];
+ const char *name = netdev_vport_get_dpif_port(netdev,
+ namebuf, sizeof namebuf);
+
+ error = dpif_netlink_port_create(netdev);
+ if (error) {
+ return error;
+ }
+
+ error = dpif_netlink_port_add__(dpif, name, OVS_VPORT_TYPE_NETDEV, NULL, port_nop);
+ if (error) {
+ VLOG_DBG("failed to add port, destroying: %d", error);
+ dpif_netlink_port_destroy(name, netdev_get_type(netdev));
+ }
+ return error;
+}
+
static int
dpif_netlink_port_add(struct dpif *dpif_, struct netdev *netdev,
odp_port_t *port_nop)
int error;
fat_rwlock_wrlock(&dpif->upcall_lock);
- error = dpif_netlink_port_add__(dpif, netdev, port_nop);
+ error = dpif_netlink_port_create_and_add(dpif, netdev, port_nop);
+ if (error == EOPNOTSUPP) {
+ error = dpif_netlink_port_add_compat(dpif, netdev, port_nop);
+ }
fat_rwlock_unlock(&dpif->upcall_lock);
return error;
{
struct dpif_netlink_vport vport;
int error;
+ struct dpif_port dpif_port;
+
+ error = dpif_netlink_port_query__(dpif, port_no, NULL, &dpif_port);
+ if (error) {
+ return error;
+ }
dpif_netlink_vport_init(&vport);
vport.cmd = OVS_VPORT_CMD_DEL;
vport_del_channels(dpif, port_no);
+ dpif_netlink_port_destroy(dpif_port.name, dpif_port.type);
+ dpif_port_destroy(&dpif_port);
+
return error;
}
const struct dpif_flow_del *del,
struct dpif_netlink_flow *request)
{
- return dpif_netlink_init_flow_del__(dpif, del->key, del->key_len,
- del->ufid, dpif->ufid_supported,
- request);
+ dpif_netlink_init_flow_del__(dpif, del->key, del->key_len,
+ del->ufid, del->terse, request);
}
struct dpif_netlink_flow_dump {
dpif_flow->actions = datapath_flow->actions;
dpif_flow->actions_len = datapath_flow->actions_len;
dpif_flow->ufid_present = datapath_flow->ufid_present;
+ dpif_flow->pmd_id = PMD_ID_NULL;
if (datapath_flow->ufid_present) {
dpif_flow->ufid = datapath_flow->ufid;
} else {
n_flows = 0;
while (!n_flows
- || (n_flows < max_flows && ofpbuf_size(&thread->nl_flows))) {
+ || (n_flows < max_flows && thread->nl_flows.size)) {
struct dpif_netlink_flow datapath_flow;
struct ofpbuf nl_flow;
int error;
size_t key_ofs;
ofpbuf_prealloc_tailroom(buf, (64
- + ofpbuf_size(d_exec->packet)
+ + dp_packet_size(d_exec->packet)
+ ODP_KEY_METADATA_SIZE
+ d_exec->actions_len));
k_exec->dp_ifindex = dp_ifindex;
nl_msg_put_unspec(buf, OVS_PACKET_ATTR_PACKET,
- ofpbuf_data(d_exec->packet),
- ofpbuf_size(d_exec->packet));
+ dp_packet_data(d_exec->packet),
+ dp_packet_size(d_exec->packet));
key_ofs = nl_msg_start_nested(buf, OVS_PACKET_ATTR_KEY);
- odp_key_from_pkt_metadata(buf, &d_exec->md);
+ odp_key_from_pkt_metadata(buf, &d_exec->packet->md);
nl_msg_end_nested(buf, key_ofs);
nl_msg_put_unspec(buf, OVS_PACKET_ATTR_ACTIONS,
d_exec->actions, d_exec->actions_len);
if (d_exec->probe) {
- nl_msg_put_flag(buf, OVS_FLOW_ATTR_PROBE);
+ nl_msg_put_flag(buf, OVS_PACKET_ATTR_PROBE);
+ }
+ if (d_exec->mtu) {
+ nl_msg_put_u16(buf, OVS_PACKET_ATTR_MRU, d_exec->mtu);
}
}
-#define MAX_OPS 50
-
-static void
+/* Executes, against 'dpif', up to the first 'n_ops' operations in 'ops'.
+ * Returns the number actually executed (at least 1, if 'n_ops' is
+ * positive). */
+static size_t
dpif_netlink_operate__(struct dpif_netlink *dpif,
struct dpif_op **ops, size_t n_ops)
{
+ enum { MAX_OPS = 50 };
+
struct op_auxdata {
struct nl_transaction txn;
struct nl_transaction *txnsp[MAX_OPS];
size_t i;
- ovs_assert(n_ops <= MAX_OPS);
+ n_ops = MIN(n_ops, MAX_OPS);
for (i = 0; i < n_ops; i++) {
struct op_auxdata *aux = &auxes[i];
struct dpif_op *op = ops[i];
struct dpif_flow_put *put;
struct dpif_flow_del *del;
- struct dpif_execute *execute;
struct dpif_flow_get *get;
struct dpif_netlink_flow flow;
break;
case DPIF_OP_EXECUTE:
- execute = &op->u.execute;
- dpif_netlink_encode_execute(dpif->dp_ifindex, execute,
- &aux->request);
+ /* Can't execute a packet that won't fit in a Netlink attribute. */
+ if (OVS_UNLIKELY(nl_attr_oversized(
+ dp_packet_size(op->u.execute.packet)))) {
+ /* Report an error immediately if this is the first operation.
+ * Otherwise the easiest thing to do is to postpone to the next
+ * call (when this will be the first operation). */
+ if (i == 0) {
+ VLOG_ERR_RL(&error_rl,
+ "dropping oversized %"PRIu32"-byte packet",
+ dp_packet_size(op->u.execute.packet));
+ op->error = ENOBUFS;
+ return 1;
+ }
+ n_ops = i;
+ } else {
+ dpif_netlink_encode_execute(dpif->dp_ifindex, &op->u.execute,
+ &aux->request);
+ }
break;
case DPIF_OP_FLOW_GET:
ofpbuf_uninit(&aux->request);
ofpbuf_uninit(&aux->reply);
}
+
+ return n_ops;
}
static void
struct dpif_netlink *dpif = dpif_netlink_cast(dpif_);
while (n_ops > 0) {
- size_t chunk = MIN(n_ops, MAX_OPS);
- dpif_netlink_operate__(dpif, ops, chunk);
+ size_t chunk = dpif_netlink_operate__(dpif, ops, n_ops);
ops += chunk;
n_ops -= chunk;
}
}
#endif
-/* Checks support for unique flow identifiers. */
-static bool
-dpif_netlink_check_ufid__(struct dpif *dpif_)
-{
- struct flow flow;
- struct odputil_keybuf keybuf;
- struct ofpbuf key;
- ovs_u128 ufid;
-
- memset(&flow, 0, sizeof flow);
- flow.dl_type = htons(0x1234);
-
- ofpbuf_use_stack(&key, &keybuf, sizeof keybuf);
- odp_flow_key_from_flow(&key, &flow, NULL, 0, true);
- dpif_flow_hash(dpif_, ofpbuf_data(&key), ofpbuf_size(&key), &ufid);
-
- return dpif_probe_feature(dpif_, "UFID", &key, &ufid);
-}
-
-static bool
-dpif_netlink_check_ufid(struct dpif *dpif_)
-{
- const struct dpif_netlink *dpif = dpif_netlink_cast(dpif_);
-
- if (dpif->ufid_supported) {
- VLOG_INFO("%s: Datapath supports userspace flow ids",
- dpif_name(dpif_));
- } else {
- VLOG_INFO("%s: Datapath does not support userspace flow ids",
- dpif_name(dpif_));
- }
- return dpif->ufid_supported;
-}
-
/* Synchronizes 'channels' in 'dpif->handlers' with the set of vports
* currently in 'dpif' in the kernel, by adding a new set of channels for
* any kernel vport that lacks one and deleting any channels that have no
/* OVS_PACKET_CMD_ACTION only. */
[OVS_PACKET_ATTR_USERDATA] = { .type = NL_A_UNSPEC, .optional = true },
[OVS_PACKET_ATTR_EGRESS_TUN_KEY] = { .type = NL_A_NESTED, .optional = true },
+ [OVS_PACKET_ATTR_ACTIONS] = { .type = NL_A_NESTED, .optional = true },
+ [OVS_PACKET_ATTR_MRU] = { .type = NL_A_U16, .optional = true }
};
- struct ovs_header *ovs_header;
- struct nlattr *a[ARRAY_SIZE(ovs_packet_policy)];
- struct nlmsghdr *nlmsg;
- struct genlmsghdr *genl;
- struct ofpbuf b;
- int type;
-
- ofpbuf_use_const(&b, ofpbuf_data(buf), ofpbuf_size(buf));
+ struct ofpbuf b = ofpbuf_const_initializer(buf->data, buf->size);
+ struct nlmsghdr *nlmsg = ofpbuf_try_pull(&b, sizeof *nlmsg);
+ struct genlmsghdr *genl = ofpbuf_try_pull(&b, sizeof *genl);
+ struct ovs_header *ovs_header = ofpbuf_try_pull(&b, sizeof *ovs_header);
- nlmsg = ofpbuf_try_pull(&b, sizeof *nlmsg);
- genl = ofpbuf_try_pull(&b, sizeof *genl);
- ovs_header = ofpbuf_try_pull(&b, sizeof *ovs_header);
+ struct nlattr *a[ARRAY_SIZE(ovs_packet_policy)];
if (!nlmsg || !genl || !ovs_header
|| nlmsg->nlmsg_type != ovs_packet_family
|| !nl_policy_parse(&b, 0, ovs_packet_policy, a,
return EINVAL;
}
- type = (genl->cmd == OVS_PACKET_CMD_MISS ? DPIF_UC_MISS
- : genl->cmd == OVS_PACKET_CMD_ACTION ? DPIF_UC_ACTION
- : -1);
+ int type = (genl->cmd == OVS_PACKET_CMD_MISS ? DPIF_UC_MISS
+ : genl->cmd == OVS_PACKET_CMD_ACTION ? DPIF_UC_ACTION
+ : -1);
if (type < 0) {
return EINVAL;
}
dpif_flow_hash(&dpif->dpif, upcall->key, upcall->key_len, &upcall->ufid);
upcall->userdata = a[OVS_PACKET_ATTR_USERDATA];
upcall->out_tun_key = a[OVS_PACKET_ATTR_EGRESS_TUN_KEY];
+ upcall->actions = a[OVS_PACKET_ATTR_ACTIONS];
+ upcall->mru = a[OVS_PACKET_ATTR_MRU];
/* Allow overwriting the netlink attribute header without reallocating. */
- ofpbuf_use_stub(&upcall->packet,
+ dp_packet_use_stub(&upcall->packet,
CONST_CAST(struct nlattr *,
nl_attr_get(a[OVS_PACKET_ATTR_PACKET])) - 1,
nl_attr_get_size(a[OVS_PACKET_ATTR_PACKET]) +
sizeof(struct nlattr));
- ofpbuf_set_data(&upcall->packet,
- (char *)ofpbuf_data(&upcall->packet) + sizeof(struct nlattr));
- ofpbuf_set_size(&upcall->packet, nl_attr_get_size(a[OVS_PACKET_ATTR_PACKET]));
+ dp_packet_set_data(&upcall->packet,
+ (char *)dp_packet_data(&upcall->packet) + sizeof(struct nlattr));
+ dp_packet_set_size(&upcall->packet, nl_attr_get_size(a[OVS_PACKET_ATTR_PACKET]));
*dp_ifindex = ovs_header->dp_ifindex;
return version_str;
}
+#ifdef __linux__
+struct dpif_netlink_ct_dump_state {
+ struct ct_dpif_dump_state up;
+ struct nl_ct_dump_state *nl_ct_dump;
+};
+
+static int
+dpif_netlink_ct_dump_start(struct dpif *dpif OVS_UNUSED,
+ struct ct_dpif_dump_state **dump_,
+ const uint16_t *zone)
+{
+ struct dpif_netlink_ct_dump_state *dump;
+ int err;
+
+ dump = xzalloc(sizeof *dump);
+ err = nl_ct_dump_start(&dump->nl_ct_dump, zone);
+ if (err) {
+ free(dump);
+ return err;
+ }
+
+ *dump_ = &dump->up;
+
+ return 0;
+}
+
+static int
+dpif_netlink_ct_dump_next(struct dpif *dpif OVS_UNUSED,
+ struct ct_dpif_dump_state *dump_,
+ struct ct_dpif_entry *entry)
+{
+ struct dpif_netlink_ct_dump_state *dump;
+
+ INIT_CONTAINER(dump, dump_, up);
+
+ return nl_ct_dump_next(dump->nl_ct_dump, entry);
+}
+
+static int
+dpif_netlink_ct_dump_done(struct dpif *dpif OVS_UNUSED,
+ struct ct_dpif_dump_state *dump_)
+{
+ struct dpif_netlink_ct_dump_state *dump;
+ int err;
+
+ INIT_CONTAINER(dump, dump_, up);
+
+ err = nl_ct_dump_done(dump->nl_ct_dump);
+ free(dump);
+ return err;
+}
+
+static int
+dpif_netlink_ct_flush(struct dpif *dpif OVS_UNUSED, const uint16_t *zone)
+{
+ if (zone) {
+ return nl_ct_flush_zone(*zone);
+ } else {
+ return nl_ct_flush();
+ }
+}
+#endif
+
const struct dpif_class dpif_netlink_class = {
"system",
+ NULL, /* init */
dpif_netlink_enumerate,
NULL,
dpif_netlink_open,
dpif_netlink_recv,
dpif_netlink_recv_wait,
dpif_netlink_recv_purge,
+ NULL, /* register_dp_purge_cb */
NULL, /* register_upcall_cb */
NULL, /* enable_upcall */
NULL, /* disable_upcall */
dpif_netlink_get_datapath_version, /* get_datapath_version */
- dpif_netlink_check_ufid,
+#ifdef __linux__
+ dpif_netlink_ct_dump_start,
+ dpif_netlink_ct_dump_next,
+ dpif_netlink_ct_dump_done,
+ dpif_netlink_ct_flush,
+#else
+ NULL, /* ct_dump_start */
+ NULL, /* ct_dump_next */
+ NULL, /* ct_dump_done */
+ NULL, /* ct_flush */
+#endif
};
static int
error = nl_lookup_genl_family(OVS_DATAPATH_FAMILY,
&ovs_datapath_family);
if (error) {
- VLOG_ERR("Generic Netlink family '%s' does not exist. "
- "The Open vSwitch kernel module is probably not loaded.",
- OVS_DATAPATH_FAMILY);
+ VLOG_WARN("Generic Netlink family '%s' does not exist. "
+ "The Open vSwitch kernel module is probably not loaded.",
+ OVS_DATAPATH_FAMILY);
}
if (!error) {
error = nl_lookup_genl_family(OVS_VPORT_FAMILY, &ovs_vport_family);
[OVS_VPORT_ATTR_OPTIONS] = { .type = NL_A_NESTED, .optional = true },
};
- struct nlattr *a[ARRAY_SIZE(ovs_vport_policy)];
- struct ovs_header *ovs_header;
- struct nlmsghdr *nlmsg;
- struct genlmsghdr *genl;
- struct ofpbuf b;
-
dpif_netlink_vport_init(vport);
- ofpbuf_use_const(&b, ofpbuf_data(buf), ofpbuf_size(buf));
- nlmsg = ofpbuf_try_pull(&b, sizeof *nlmsg);
- genl = ofpbuf_try_pull(&b, sizeof *genl);
- ovs_header = ofpbuf_try_pull(&b, sizeof *ovs_header);
+ struct ofpbuf b = ofpbuf_const_initializer(buf->data, buf->size);
+ struct nlmsghdr *nlmsg = ofpbuf_try_pull(&b, sizeof *nlmsg);
+ struct genlmsghdr *genl = ofpbuf_try_pull(&b, sizeof *genl);
+ struct ovs_header *ovs_header = ofpbuf_try_pull(&b, sizeof *ovs_header);
+
+ struct nlattr *a[ARRAY_SIZE(ovs_vport_policy)];
if (!nlmsg || !genl || !ovs_header
|| nlmsg->nlmsg_type != ovs_vport_family
|| !nl_policy_parse(&b, 0, ovs_vport_policy, a,
.optional = true },
};
- struct nlattr *a[ARRAY_SIZE(ovs_datapath_policy)];
- struct ovs_header *ovs_header;
- struct nlmsghdr *nlmsg;
- struct genlmsghdr *genl;
- struct ofpbuf b;
-
dpif_netlink_dp_init(dp);
- ofpbuf_use_const(&b, ofpbuf_data(buf), ofpbuf_size(buf));
- nlmsg = ofpbuf_try_pull(&b, sizeof *nlmsg);
- genl = ofpbuf_try_pull(&b, sizeof *genl);
- ovs_header = ofpbuf_try_pull(&b, sizeof *ovs_header);
+ struct ofpbuf b = ofpbuf_const_initializer(buf->data, buf->size);
+ struct nlmsghdr *nlmsg = ofpbuf_try_pull(&b, sizeof *nlmsg);
+ struct genlmsghdr *genl = ofpbuf_try_pull(&b, sizeof *genl);
+ struct ovs_header *ovs_header = ofpbuf_try_pull(&b, sizeof *ovs_header);
+
+ struct nlattr *a[ARRAY_SIZE(ovs_datapath_policy)];
if (!nlmsg || !genl || !ovs_header
|| nlmsg->nlmsg_type != ovs_datapath_family
|| !nl_policy_parse(&b, 0, ovs_datapath_policy, a,
/* The kernel never uses OVS_FLOW_ATTR_UFID_FLAGS. */
};
- struct nlattr *a[ARRAY_SIZE(ovs_flow_policy)];
- struct ovs_header *ovs_header;
- struct nlmsghdr *nlmsg;
- struct genlmsghdr *genl;
- struct ofpbuf b;
-
dpif_netlink_flow_init(flow);
- ofpbuf_use_const(&b, ofpbuf_data(buf), ofpbuf_size(buf));
- nlmsg = ofpbuf_try_pull(&b, sizeof *nlmsg);
- genl = ofpbuf_try_pull(&b, sizeof *genl);
- ovs_header = ofpbuf_try_pull(&b, sizeof *ovs_header);
+ struct ofpbuf b = ofpbuf_const_initializer(buf->data, buf->size);
+ struct nlmsghdr *nlmsg = ofpbuf_try_pull(&b, sizeof *nlmsg);
+ struct genlmsghdr *genl = ofpbuf_try_pull(&b, sizeof *genl);
+ struct ovs_header *ovs_header = ofpbuf_try_pull(&b, sizeof *ovs_header);
+
+ struct nlattr *a[ARRAY_SIZE(ovs_flow_policy)];
if (!nlmsg || !genl || !ovs_header
|| nlmsg->nlmsg_type != ovs_flow_family
|| !nl_policy_parse(&b, 0, ovs_flow_policy, a,
projects / cascardo / ovs.git / blobdiff