/*
- * Copyright (c) 2007, 2008, 2009, 2010, 2011 Nicira Networks.
- * Distributed under the terms of the GNU GPL version 2.
+ * Copyright (c) 2007-2012 Nicira, Inc.
*
- * Significant portions of this file may be copied from parts of the Linux
- * kernel, by Linus Torvalds and others.
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of version 2 of the GNU General Public
+ * License as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA
*/
-/* Functions for managing the dp interface/device. */
-
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
#include <linux/init.h>
#include <linux/version.h>
#include <linux/ethtool.h>
#include <linux/wait.h>
-#include <asm/system.h>
#include <asm/div64.h>
-#include <asm/bug.h>
#include <linux/highmem.h>
#include <linux/netfilter_bridge.h>
#include <linux/netfilter_ipv4.h>
#include <linux/inetdevice.h>
#include <linux/list.h>
+#include <linux/openvswitch.h>
#include <linux/rculist.h>
#include <linux/dmi.h>
-#include <net/inet_ecn.h>
#include <net/genetlink.h>
+#include <net/net_namespace.h>
+#include <net/netns/generic.h>
-#include "openvswitch/datapath-protocol.h"
#include "checksum.h"
#include "datapath.h"
-#include "actions.h"
#include "flow.h"
+#include "genl_exec.h"
#include "vlan.h"
#include "tunnel.h"
#include "vport-internal_dev.h"
#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18) || \
- LINUX_VERSION_CODE >= KERNEL_VERSION(3,1,0)
-#error Kernels before 2.6.18 or after 3.0 are not supported by this version of Open vSwitch.
+ LINUX_VERSION_CODE >= KERNEL_VERSION(3,7,0)
+#error Kernels before 2.6.18 or after 3.6 are not supported by this version of Open vSwitch.
#endif
-int (*dp_ioctl_hook)(struct net_device *dev, struct ifreq *rq, int cmd);
-EXPORT_SYMBOL(dp_ioctl_hook);
+#define REHASH_FLOW_INTERVAL (10 * 60 * HZ)
+static void rehash_flow_table(struct work_struct *work);
+static DECLARE_DELAYED_WORK(rehash_flow_wq, rehash_flow_table);
+
+int ovs_net_id __read_mostly;
+
+int (*ovs_dp_ioctl_hook)(struct net_device *dev, struct ifreq *rq, int cmd);
+EXPORT_SYMBOL(ovs_dp_ioctl_hook);
/**
* DOC: Locking:
* etc.) are protected by RTNL.
*
* Writes to other state (flow table modifications, set miscellaneous datapath
- * parameters such as drop frags, etc.) are protected by genl_mutex. The RTNL
- * lock nests inside genl_mutex.
+ * parameters, etc.) are protected by genl_mutex. The RTNL lock nests inside
+ * genl_mutex.
*
* Reads are protected by RCU.
*
* each other.
*/
-/* Global list of datapaths to enable dumping them all out.
- * Protected by genl_mutex.
- */
-static LIST_HEAD(dps);
-
static struct vport *new_vport(const struct vport_parms *);
-static int queue_userspace_packets(struct datapath *, struct sk_buff *,
- const struct dp_upcall_info *);
+static int queue_gso_packets(struct net *, int dp_ifindex, struct sk_buff *,
+ const struct dp_upcall_info *);
+static int queue_userspace_packet(struct net *, int dp_ifindex,
+ struct sk_buff *,
+ const struct dp_upcall_info *);
/* Must be called with rcu_read_lock, genl_mutex, or RTNL lock. */
-struct datapath *get_dp(int dp_ifindex)
+static struct datapath *get_dp(struct net *net, int dp_ifindex)
{
struct datapath *dp = NULL;
struct net_device *dev;
rcu_read_lock();
- dev = dev_get_by_index_rcu(&init_net, dp_ifindex);
+ dev = dev_get_by_index_rcu(net, dp_ifindex);
if (dev) {
- struct vport *vport = internal_dev_get_vport(dev);
+ struct vport *vport = ovs_internal_dev_get_vport(dev);
if (vport)
dp = vport->dp;
}
return dp;
}
-EXPORT_SYMBOL_GPL(get_dp);
-
-/* Must be called with genl_mutex. */
-static struct flow_table *get_table_protected(struct datapath *dp)
-{
- return rcu_dereference_protected(dp->table, lockdep_genl_is_held());
-}
/* Must be called with rcu_read_lock or RTNL lock. */
-static struct vport *get_vport_protected(struct datapath *dp, u16 port_no)
+const char *ovs_dp_name(const struct datapath *dp)
{
- return rcu_dereference_rtnl(dp->ports[port_no]);
+ struct vport *vport = ovs_vport_rtnl_rcu(dp, OVSP_LOCAL);
+ return vport->ops->get_name(vport);
}
-/* Must be called with rcu_read_lock or RTNL lock. */
-const char *dp_name(const struct datapath *dp)
+static int get_dpifindex(struct datapath *dp)
{
- return vport_get_name(rcu_dereference_rtnl(dp->ports[OVSP_LOCAL]));
+ struct vport *local;
+ int ifindex;
+
+ rcu_read_lock();
+
+ local = ovs_vport_rcu(dp, OVSP_LOCAL);
+ if (local)
+ ifindex = local->ops->get_ifindex(local);
+ else
+ ifindex = 0;
+
+ rcu_read_unlock();
+
+ return ifindex;
}
-static inline size_t br_nlmsg_size(void)
+static size_t br_nlmsg_size(void)
{
return NLMSG_ALIGN(sizeof(struct ifinfomsg))
+ nla_total_size(IFNAMSIZ) /* IFLA_IFNAME */
int event, unsigned int flags)
{
struct datapath *dp = port->dp;
- int ifindex = vport_get_ifindex(port);
struct ifinfomsg *hdr;
struct nlmsghdr *nlh;
- if (ifindex < 0)
- return ifindex;
+ if (!port->ops->get_ifindex)
+ return -ENODEV;
nlh = nlmsg_put(skb, 0, 0, event, sizeof(*hdr), flags);
if (nlh == NULL)
hdr->ifi_family = AF_BRIDGE;
hdr->__ifi_pad = 0;
hdr->ifi_type = ARPHRD_ETHER;
- hdr->ifi_index = ifindex;
- hdr->ifi_flags = vport_get_flags(port);
+ hdr->ifi_index = port->ops->get_ifindex(port);
+ hdr->ifi_flags = port->ops->get_dev_flags(port);
hdr->ifi_change = 0;
- NLA_PUT_STRING(skb, IFLA_IFNAME, vport_get_name(port));
- NLA_PUT_U32(skb, IFLA_MASTER,
- vport_get_ifindex(get_vport_protected(dp, OVSP_LOCAL)));
- NLA_PUT_U32(skb, IFLA_MTU, vport_get_mtu(port));
+ if (nla_put_string(skb, IFLA_IFNAME, port->ops->get_name(port)) ||
+ nla_put_u32(skb, IFLA_MASTER, get_dpifindex(dp)) ||
+ nla_put_u32(skb, IFLA_MTU, port->ops->get_mtu(port)) ||
#ifdef IFLA_OPERSTATE
- NLA_PUT_U8(skb, IFLA_OPERSTATE,
- vport_is_running(port)
- ? vport_get_operstate(port)
- : IF_OPER_DOWN);
+ nla_put_u8(skb, IFLA_OPERSTATE,
+ port->ops->is_running(port) ?
+ port->ops->get_operstate(port) :
+ IF_OPER_DOWN) ||
#endif
-
- NLA_PUT(skb, IFLA_ADDRESS, ETH_ALEN, vport_get_addr(port));
+ nla_put(skb, IFLA_ADDRESS, ETH_ALEN, port->ops->get_addr(port)))
+ goto nla_put_failure;
return nlmsg_end(skb, nlh);
static void dp_ifinfo_notify(int event, struct vport *port)
{
struct sk_buff *skb;
- int err = -ENOBUFS;
+ int err;
skb = nlmsg_new(br_nlmsg_size(), GFP_KERNEL);
- if (skb == NULL)
- goto errout;
+ if (!skb) {
+ err = -ENOBUFS;
+ goto err;
+ }
err = dp_fill_ifinfo(skb, port, event, 0);
if (err < 0) {
- /* -EMSGSIZE implies BUG in br_nlmsg_size() */
- WARN_ON(err == -EMSGSIZE);
- kfree_skb(skb);
- goto errout;
+ if (err == -ENODEV) {
+ goto out;
+ } else {
+ /* -EMSGSIZE implies BUG in br_nlmsg_size() */
+ WARN_ON(err == -EMSGSIZE);
+ goto err;
+ }
}
- rtnl_notify(skb, &init_net, 0, RTNLGRP_LINK, NULL, GFP_KERNEL);
+
+ rtnl_notify(skb, ovs_dp_get_net(port->dp), 0, RTNLGRP_LINK, NULL, GFP_KERNEL);
+
return;
-errout:
- if (err < 0)
- rtnl_set_sk_err(&init_net, RTNLGRP_LINK, err);
+err:
+ rtnl_set_sk_err(ovs_dp_get_net(port->dp), RTNLGRP_LINK, err);
+out:
+ kfree_skb(skb);
}
static void release_dp(struct kobject *kobj)
{
struct datapath *dp = container_of(rcu, struct datapath, rcu);
- flow_tbl_destroy(dp->table);
+ ovs_flow_tbl_destroy((__force struct flow_table *)dp->table);
free_percpu(dp->stats_percpu);
+ release_net(ovs_dp_get_net(dp));
+ kfree(dp->ports);
kobject_put(&dp->ifobj);
}
+static struct hlist_head *vport_hash_bucket(const struct datapath *dp,
+ u16 port_no)
+{
+ return &dp->ports[port_no & (DP_VPORT_HASH_BUCKETS - 1)];
+}
+
+struct vport *ovs_lookup_vport(const struct datapath *dp, u16 port_no)
+{
+ struct vport *vport;
+ struct hlist_node *n;
+ struct hlist_head *head;
+
+ head = vport_hash_bucket(dp, port_no);
+ hlist_for_each_entry_rcu(vport, n, head, dp_hash_node) {
+ if (vport->port_no == port_no)
+ return vport;
+ }
+ return NULL;
+}
+
/* Called with RTNL lock and genl_lock. */
static struct vport *new_vport(const struct vport_parms *parms)
{
struct vport *vport;
- vport = vport_add(parms);
+ vport = ovs_vport_add(parms);
if (!IS_ERR(vport)) {
struct datapath *dp = parms->dp;
+ struct hlist_head *head = vport_hash_bucket(dp, vport->port_no);
- rcu_assign_pointer(dp->ports[parms->port_no], vport);
- list_add(&vport->node, &dp->port_list);
-
+ hlist_add_head_rcu(&vport->dp_hash_node, head);
dp_ifinfo_notify(RTM_NEWLINK, vport);
}
-
return vport;
}
/* Called with RTNL lock. */
-void dp_detach_port(struct vport *p)
+void ovs_dp_detach_port(struct vport *p)
{
ASSERT_RTNL();
if (p->port_no != OVSP_LOCAL)
- dp_sysfs_del_if(p);
+ ovs_dp_sysfs_del_if(p);
+
dp_ifinfo_notify(RTM_DELLINK, p);
/* First drop references to device. */
- list_del(&p->node);
- rcu_assign_pointer(p->dp->ports[p->port_no], NULL);
+ hlist_del_rcu(&p->dp_hash_node);
/* Then destroy it. */
- vport_del(p);
+ ovs_vport_del(p);
}
/* Must be called with rcu_read_lock. */
-void dp_process_received_packet(struct vport *p, struct sk_buff *skb)
+void ovs_dp_process_received_packet(struct vport *p, struct sk_buff *skb)
{
struct datapath *dp = p->dp;
struct sw_flow *flow;
struct dp_stats_percpu *stats;
- int stats_counter_off;
+ u64 *stats_counter;
int error;
- OVS_CB(skb)->vport = p;
+ stats = per_cpu_ptr(dp->stats_percpu, smp_processor_id());
if (!OVS_CB(skb)->flow) {
struct sw_flow_key key;
int key_len;
- bool is_frag;
/* Extract flow from 'skb' into 'key'. */
- error = flow_extract(skb, p->port_no, &key, &key_len, &is_frag);
+ error = ovs_flow_extract(skb, p->port_no, &key, &key_len);
if (unlikely(error)) {
kfree_skb(skb);
return;
}
- if (is_frag && dp->drop_frags) {
- consume_skb(skb);
- stats_counter_off = offsetof(struct dp_stats_percpu, n_frags);
- goto out;
- }
-
/* Look up flow. */
- flow = flow_tbl_lookup(rcu_dereference(dp->table), &key, key_len);
+ flow = ovs_flow_tbl_lookup(rcu_dereference(dp->table),
+ &key, key_len);
if (unlikely(!flow)) {
struct dp_upcall_info upcall;
upcall.cmd = OVS_PACKET_CMD_MISS;
upcall.key = &key;
- upcall.userdata = 0;
- upcall.sample_pool = 0;
- upcall.actions = NULL;
- upcall.actions_len = 0;
- dp_upcall(dp, skb, &upcall);
- stats_counter_off = offsetof(struct dp_stats_percpu, n_missed);
+ upcall.userdata = NULL;
+ upcall.pid = p->upcall_pid;
+ ovs_dp_upcall(dp, skb, &upcall);
+ consume_skb(skb);
+ stats_counter = &stats->n_missed;
goto out;
}
OVS_CB(skb)->flow = flow;
}
- stats_counter_off = offsetof(struct dp_stats_percpu, n_hit);
- flow_used(OVS_CB(skb)->flow, skb);
- execute_actions(dp, skb);
+ stats_counter = &stats->n_hit;
+ ovs_flow_used(OVS_CB(skb)->flow, skb);
+ ovs_execute_actions(dp, skb);
out:
/* Update datapath statistics. */
- local_bh_disable();
- stats = per_cpu_ptr(dp->stats_percpu, smp_processor_id());
-
- write_seqcount_begin(&stats->seqlock);
- (*(u64 *)((u8 *)stats + stats_counter_off))++;
- write_seqcount_end(&stats->seqlock);
-
- local_bh_enable();
-}
-
-static void copy_and_csum_skb(struct sk_buff *skb, void *to)
-{
- u16 csum_start, csum_offset;
- __wsum csum;
-
- get_skb_csum_pointers(skb, &csum_start, &csum_offset);
- csum_start -= skb_headroom(skb);
-
- skb_copy_bits(skb, 0, to, csum_start);
-
- csum = skb_copy_and_csum_bits(skb, csum_start, to + csum_start,
- skb->len - csum_start, 0);
- *(__sum16 *)(to + csum_start + csum_offset) = csum_fold(csum);
+ u64_stats_update_begin(&stats->sync);
+ (*stats_counter)++;
+ u64_stats_update_end(&stats->sync);
}
static struct genl_family dp_packet_genl_family = {
.id = GENL_ID_GENERATE,
.hdrsize = sizeof(struct ovs_header),
.name = OVS_PACKET_FAMILY,
- .version = 1,
- .maxattr = OVS_PACKET_ATTR_MAX
+ .version = OVS_PACKET_VERSION,
+ .maxattr = OVS_PACKET_ATTR_MAX,
+ SET_NETNSOK
};
-/* Generic Netlink multicast groups for upcalls.
- *
- * We really want three unique multicast groups per datapath, but we can't even
- * get one, because genl_register_mc_group() takes genl_lock, which is also
- * held during Generic Netlink message processing, so trying to acquire
- * multicast groups during OVS_DP_NEW processing deadlocks. Instead, we
- * preallocate a few groups and use them round-robin for datapaths. Collision
- * isn't fatal--multicast listeners should check that the family is the one
- * that they want and discard others--but it wastes time and memory to receive
- * unwanted messages.
- */
-#define PACKET_N_MC_GROUPS 16
-static struct genl_multicast_group packet_mc_groups[PACKET_N_MC_GROUPS];
-
-static u32 packet_mc_group(struct datapath *dp, u8 cmd)
-{
- u32 idx;
- BUILD_BUG_ON_NOT_POWER_OF_2(PACKET_N_MC_GROUPS);
-
- idx = jhash_2words(dp->dp_ifindex, cmd, 0) & (PACKET_N_MC_GROUPS - 1);
- return packet_mc_groups[idx].id;
-}
-
-static int packet_register_mc_groups(void)
-{
- int i;
-
- for (i = 0; i < PACKET_N_MC_GROUPS; i++) {
- struct genl_multicast_group *group = &packet_mc_groups[i];
- int error;
-
- sprintf(group->name, "packet%d", i);
- error = genl_register_mc_group(&dp_packet_genl_family, group);
- if (error)
- return error;
- }
- return 0;
-}
-
-int dp_upcall(struct datapath *dp, struct sk_buff *skb, const struct dp_upcall_info *upcall_info)
+int ovs_dp_upcall(struct datapath *dp, struct sk_buff *skb,
+ const struct dp_upcall_info *upcall_info)
{
struct dp_stats_percpu *stats;
+ int dp_ifindex;
int err;
- WARN_ON_ONCE(skb_shared(skb));
-
- forward_ip_summed(skb, true);
+ if (upcall_info->pid == 0) {
+ err = -ENOTCONN;
+ goto err;
+ }
- /* Break apart GSO packets into their component pieces. Otherwise
- * userspace may try to stuff a 64kB packet into a 1500-byte MTU. */
- if (skb_is_gso(skb)) {
- struct sk_buff *nskb = skb_gso_segment(skb, NETIF_F_SG | NETIF_F_HW_CSUM);
-
- if (IS_ERR(nskb)) {
- kfree_skb(skb);
- err = PTR_ERR(nskb);
- goto err;
- }
- consume_skb(skb);
- skb = nskb;
+ dp_ifindex = get_dpifindex(dp);
+ if (!dp_ifindex) {
+ err = -ENODEV;
+ goto err;
}
- err = queue_userspace_packets(dp, skb, upcall_info);
+ forward_ip_summed(skb, true);
+
+ if (!skb_is_gso(skb))
+ err = queue_userspace_packet(ovs_dp_get_net(dp), dp_ifindex, skb, upcall_info);
+ else
+ err = queue_gso_packets(ovs_dp_get_net(dp), dp_ifindex, skb, upcall_info);
if (err)
goto err;
return 0;
err:
- local_bh_disable();
stats = per_cpu_ptr(dp->stats_percpu, smp_processor_id());
- write_seqcount_begin(&stats->seqlock);
+ u64_stats_update_begin(&stats->sync);
stats->n_lost++;
- write_seqcount_end(&stats->seqlock);
-
- local_bh_enable();
+ u64_stats_update_end(&stats->sync);
return err;
}
-/* Send each packet in the 'skb' list to userspace for 'dp' as directed by
- * 'upcall_info'. There will be only one packet unless we broke up a GSO
- * packet.
- */
-static int queue_userspace_packets(struct datapath *dp, struct sk_buff *skb,
- const struct dp_upcall_info *upcall_info)
+static int queue_gso_packets(struct net *net, int dp_ifindex,
+ struct sk_buff *skb,
+ const struct dp_upcall_info *upcall_info)
{
- u32 group = packet_mc_group(dp, upcall_info->cmd);
- struct sk_buff *nskb;
+ unsigned short gso_type = skb_shinfo(skb)->gso_type;
+ struct dp_upcall_info later_info;
+ struct sw_flow_key later_key;
+ struct sk_buff *segs, *nskb;
int err;
- do {
- struct ovs_header *upcall;
- struct sk_buff *user_skb; /* to be queued to userspace */
- struct nlattr *nla;
- unsigned int len;
-
- nskb = skb->next;
- skb->next = NULL;
-
- err = vlan_deaccel_tag(skb);
- if (unlikely(err))
- goto err_kfree_skbs;
-
- if (nla_attr_size(skb->len) > USHRT_MAX)
- goto err_kfree_skbs;
-
- len = sizeof(struct ovs_header);
- len += nla_total_size(skb->len);
- len += nla_total_size(FLOW_BUFSIZE);
- if (upcall_info->userdata)
- len += nla_total_size(8);
- if (upcall_info->sample_pool)
- len += nla_total_size(4);
- if (upcall_info->actions_len)
- len += nla_total_size(upcall_info->actions_len);
-
- user_skb = genlmsg_new(len, GFP_ATOMIC);
- if (!user_skb) {
- netlink_set_err(INIT_NET_GENL_SOCK, 0, group, -ENOBUFS);
- goto err_kfree_skbs;
- }
-
- upcall = genlmsg_put(user_skb, 0, 0, &dp_packet_genl_family, 0, upcall_info->cmd);
- upcall->dp_ifindex = dp->dp_ifindex;
+ segs = skb_gso_segment(skb, NETIF_F_SG | NETIF_F_HW_CSUM);
+ if (IS_ERR(segs))
+ return PTR_ERR(segs);
- nla = nla_nest_start(user_skb, OVS_PACKET_ATTR_KEY);
- flow_to_nlattrs(upcall_info->key, user_skb);
- nla_nest_end(user_skb, nla);
-
- if (upcall_info->userdata)
- nla_put_u64(user_skb, OVS_PACKET_ATTR_USERDATA, upcall_info->userdata);
- if (upcall_info->sample_pool)
- nla_put_u32(user_skb, OVS_PACKET_ATTR_SAMPLE_POOL, upcall_info->sample_pool);
- if (upcall_info->actions_len) {
- const struct nlattr *actions = upcall_info->actions;
- u32 actions_len = upcall_info->actions_len;
+ /* Queue all of the segments. */
+ skb = segs;
+ do {
+ err = queue_userspace_packet(net, dp_ifindex, skb, upcall_info);
+ if (err)
+ break;
- nla = nla_nest_start(user_skb, OVS_PACKET_ATTR_ACTIONS);
- memcpy(__skb_put(user_skb, actions_len), actions, actions_len);
- nla_nest_end(user_skb, nla);
+ if (skb == segs && gso_type & SKB_GSO_UDP) {
+ /* The initial flow key extracted by ovs_flow_extract()
+ * in this case is for a first fragment, so we need to
+ * properly mark later fragments.
+ */
+ later_key = *upcall_info->key;
+ later_key.ip.frag = OVS_FRAG_TYPE_LATER;
+
+ later_info = *upcall_info;
+ later_info.key = &later_key;
+ upcall_info = &later_info;
}
+ } while ((skb = skb->next));
- nla = __nla_reserve(user_skb, OVS_PACKET_ATTR_PACKET, skb->len);
- if (skb->ip_summed == CHECKSUM_PARTIAL)
- copy_and_csum_skb(skb, nla_data(nla));
+ /* Free all of the segments. */
+ skb = segs;
+ do {
+ nskb = skb->next;
+ if (err)
+ kfree_skb(skb);
else
- skb_copy_bits(skb, 0, nla_data(nla), skb->len);
+ consume_skb(skb);
+ } while ((skb = nskb));
+ return err;
+}
- err = genlmsg_multicast(user_skb, 0, group, GFP_ATOMIC);
+static int queue_userspace_packet(struct net *net, int dp_ifindex,
+ struct sk_buff *skb,
+ const struct dp_upcall_info *upcall_info)
+{
+ struct ovs_header *upcall;
+ struct sk_buff *nskb = NULL;
+ struct sk_buff *user_skb; /* to be queued to userspace */
+ struct nlattr *nla;
+ unsigned int len;
+ int err;
+
+ if (vlan_tx_tag_present(skb)) {
+ nskb = skb_clone(skb, GFP_ATOMIC);
+ if (!nskb)
+ return -ENOMEM;
+
+ err = vlan_deaccel_tag(nskb);
if (err)
- goto err_kfree_skbs;
+ return err;
- consume_skb(skb);
skb = nskb;
- } while (skb);
- return 0;
+ }
-err_kfree_skbs:
- kfree_skb(skb);
- while ((skb = nskb) != NULL) {
- nskb = skb->next;
- kfree_skb(skb);
+ if (nla_attr_size(skb->len) > USHRT_MAX) {
+ err = -EFBIG;
+ goto out;
+ }
+
+ len = sizeof(struct ovs_header);
+ len += nla_total_size(skb->len);
+ len += nla_total_size(FLOW_BUFSIZE);
+ if (upcall_info->cmd == OVS_PACKET_CMD_ACTION)
+ len += nla_total_size(8);
+
+ user_skb = genlmsg_new(len, GFP_ATOMIC);
+ if (!user_skb) {
+ err = -ENOMEM;
+ goto out;
}
+
+ upcall = genlmsg_put(user_skb, 0, 0, &dp_packet_genl_family,
+ 0, upcall_info->cmd);
+ upcall->dp_ifindex = dp_ifindex;
+
+ nla = nla_nest_start(user_skb, OVS_PACKET_ATTR_KEY);
+ ovs_flow_to_nlattrs(upcall_info->key, user_skb);
+ nla_nest_end(user_skb, nla);
+
+ if (upcall_info->userdata)
+ nla_put_u64(user_skb, OVS_PACKET_ATTR_USERDATA,
+ nla_get_u64(upcall_info->userdata));
+
+ nla = __nla_reserve(user_skb, OVS_PACKET_ATTR_PACKET, skb->len);
+
+ skb_copy_and_csum_dev(skb, nla_data(nla));
+
+ err = genlmsg_unicast(net, user_skb, upcall_info->pid);
+
+out:
+ kfree_skb(nskb);
return err;
}
/* Called with genl_mutex. */
-static int flush_flows(int dp_ifindex)
+static int flush_flows(struct datapath *dp)
{
struct flow_table *old_table;
struct flow_table *new_table;
- struct datapath *dp;
- dp = get_dp(dp_ifindex);
- if (!dp)
- return -ENODEV;
-
- old_table = get_table_protected(dp);
- new_table = flow_tbl_alloc(TBL_MIN_BUCKETS);
+ old_table = genl_dereference(dp->table);
+ new_table = ovs_flow_tbl_alloc(TBL_MIN_BUCKETS);
if (!new_table)
return -ENOMEM;
rcu_assign_pointer(dp->table, new_table);
- flow_tbl_deferred_destroy(old_table);
+ ovs_flow_tbl_deferred_destroy(old_table);
return 0;
}
-static int validate_actions(const struct nlattr *attr)
+static int validate_actions(const struct nlattr *attr,
+ const struct sw_flow_key *key, int depth);
+
+static int validate_sample(const struct nlattr *attr,
+ const struct sw_flow_key *key, int depth)
{
+ const struct nlattr *attrs[OVS_SAMPLE_ATTR_MAX + 1];
+ const struct nlattr *probability, *actions;
const struct nlattr *a;
int rem;
+ memset(attrs, 0, sizeof(attrs));
+ nla_for_each_nested(a, attr, rem) {
+ int type = nla_type(a);
+ if (!type || type > OVS_SAMPLE_ATTR_MAX || attrs[type])
+ return -EINVAL;
+ attrs[type] = a;
+ }
+ if (rem)
+ return -EINVAL;
+
+ probability = attrs[OVS_SAMPLE_ATTR_PROBABILITY];
+ if (!probability || nla_len(probability) != sizeof(u32))
+ return -EINVAL;
+
+ actions = attrs[OVS_SAMPLE_ATTR_ACTIONS];
+ if (!actions || (nla_len(actions) && nla_len(actions) < NLA_HDRLEN))
+ return -EINVAL;
+ return validate_actions(actions, key, depth + 1);
+}
+
+static int validate_tp_port(const struct sw_flow_key *flow_key)
+{
+ if (flow_key->eth.type == htons(ETH_P_IP)) {
+ if (flow_key->ipv4.tp.src || flow_key->ipv4.tp.dst)
+ return 0;
+ } else if (flow_key->eth.type == htons(ETH_P_IPV6)) {
+ if (flow_key->ipv6.tp.src || flow_key->ipv6.tp.dst)
+ return 0;
+ }
+
+ return -EINVAL;
+}
+
+static int validate_set(const struct nlattr *a,
+ const struct sw_flow_key *flow_key)
+{
+ const struct nlattr *ovs_key = nla_data(a);
+ int key_type = nla_type(ovs_key);
+
+ /* There can be only one key in a action */
+ if (nla_total_size(nla_len(ovs_key)) != nla_len(a))
+ return -EINVAL;
+
+ if (key_type > OVS_KEY_ATTR_MAX ||
+ nla_len(ovs_key) != ovs_key_lens[key_type])
+ return -EINVAL;
+
+ switch (key_type) {
+ const struct ovs_key_ipv4 *ipv4_key;
+ const struct ovs_key_ipv4_tunnel *tun_key;
+
+ case OVS_KEY_ATTR_PRIORITY:
+ case OVS_KEY_ATTR_TUN_ID:
+ case OVS_KEY_ATTR_ETHERNET:
+ break;
+
+ case OVS_KEY_ATTR_IPV4_TUNNEL:
+ tun_key = nla_data(ovs_key);
+ if (!tun_key->ipv4_dst)
+ return -EINVAL;
+ break;
+
+ case OVS_KEY_ATTR_IPV4:
+ if (flow_key->eth.type != htons(ETH_P_IP))
+ return -EINVAL;
+
+ if (!flow_key->ip.proto)
+ return -EINVAL;
+
+ ipv4_key = nla_data(ovs_key);
+ if (ipv4_key->ipv4_proto != flow_key->ip.proto)
+ return -EINVAL;
+
+ if (ipv4_key->ipv4_frag != flow_key->ip.frag)
+ return -EINVAL;
+
+ break;
+
+ case OVS_KEY_ATTR_TCP:
+ if (flow_key->ip.proto != IPPROTO_TCP)
+ return -EINVAL;
+
+ return validate_tp_port(flow_key);
+
+ case OVS_KEY_ATTR_UDP:
+ if (flow_key->ip.proto != IPPROTO_UDP)
+ return -EINVAL;
+
+ return validate_tp_port(flow_key);
+
+ default:
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+static int validate_userspace(const struct nlattr *attr)
+{
+ static const struct nla_policy userspace_policy[OVS_USERSPACE_ATTR_MAX + 1] = {
+ [OVS_USERSPACE_ATTR_PID] = {.type = NLA_U32 },
+ [OVS_USERSPACE_ATTR_USERDATA] = {.type = NLA_U64 },
+ };
+ struct nlattr *a[OVS_USERSPACE_ATTR_MAX + 1];
+ int error;
+
+ error = nla_parse_nested(a, OVS_USERSPACE_ATTR_MAX,
+ attr, userspace_policy);
+ if (error)
+ return error;
+
+ if (!a[OVS_USERSPACE_ATTR_PID] ||
+ !nla_get_u32(a[OVS_USERSPACE_ATTR_PID]))
+ return -EINVAL;
+
+ return 0;
+}
+
+static int validate_actions(const struct nlattr *attr,
+ const struct sw_flow_key *key, int depth)
+{
+ const struct nlattr *a;
+ int rem, err;
+
+ if (depth >= SAMPLE_ACTION_DEPTH)
+ return -EOVERFLOW;
+
nla_for_each_nested(a, attr, rem) {
+ /* Expected argument lengths, (u32)-1 for variable length. */
static const u32 action_lens[OVS_ACTION_ATTR_MAX + 1] = {
- [OVS_ACTION_ATTR_OUTPUT] = 4,
- [OVS_ACTION_ATTR_USERSPACE] = 8,
- [OVS_ACTION_ATTR_PUSH_VLAN] = 2,
+ [OVS_ACTION_ATTR_OUTPUT] = sizeof(u32),
+ [OVS_ACTION_ATTR_USERSPACE] = (u32)-1,
+ [OVS_ACTION_ATTR_PUSH_VLAN] = sizeof(struct ovs_action_push_vlan),
[OVS_ACTION_ATTR_POP_VLAN] = 0,
- [OVS_ACTION_ATTR_SET_DL_SRC] = ETH_ALEN,
- [OVS_ACTION_ATTR_SET_DL_DST] = ETH_ALEN,
- [OVS_ACTION_ATTR_SET_NW_SRC] = 4,
- [OVS_ACTION_ATTR_SET_NW_DST] = 4,
- [OVS_ACTION_ATTR_SET_NW_TOS] = 1,
- [OVS_ACTION_ATTR_SET_TP_SRC] = 2,
- [OVS_ACTION_ATTR_SET_TP_DST] = 2,
- [OVS_ACTION_ATTR_SET_TUNNEL] = 8,
- [OVS_ACTION_ATTR_SET_PRIORITY] = 4,
- [OVS_ACTION_ATTR_POP_PRIORITY] = 0,
+ [OVS_ACTION_ATTR_SET] = (u32)-1,
+ [OVS_ACTION_ATTR_SAMPLE] = (u32)-1
};
+ const struct ovs_action_push_vlan *vlan;
int type = nla_type(a);
- if (type > OVS_ACTION_ATTR_MAX || nla_len(a) != action_lens[type])
+ if (type > OVS_ACTION_ATTR_MAX ||
+ (action_lens[type] != nla_len(a) &&
+ action_lens[type] != (u32)-1))
return -EINVAL;
switch (type) {
return -EINVAL;
case OVS_ACTION_ATTR_USERSPACE:
- case OVS_ACTION_ATTR_POP_VLAN:
- case OVS_ACTION_ATTR_SET_DL_SRC:
- case OVS_ACTION_ATTR_SET_DL_DST:
- case OVS_ACTION_ATTR_SET_NW_SRC:
- case OVS_ACTION_ATTR_SET_NW_DST:
- case OVS_ACTION_ATTR_SET_TP_SRC:
- case OVS_ACTION_ATTR_SET_TP_DST:
- case OVS_ACTION_ATTR_SET_TUNNEL:
- case OVS_ACTION_ATTR_SET_PRIORITY:
- case OVS_ACTION_ATTR_POP_PRIORITY:
- /* No validation needed. */
+ err = validate_userspace(a);
+ if (err)
+ return err;
break;
case OVS_ACTION_ATTR_OUTPUT:
return -EINVAL;
break;
+
+ case OVS_ACTION_ATTR_POP_VLAN:
+ break;
+
case OVS_ACTION_ATTR_PUSH_VLAN:
- if (nla_get_be16(a) & htons(VLAN_CFI_MASK))
+ vlan = nla_data(a);
+ if (vlan->vlan_tpid != htons(ETH_P_8021Q))
+ return -EINVAL;
+ if (!(vlan->vlan_tci & htons(VLAN_TAG_PRESENT)))
return -EINVAL;
break;
- case OVS_ACTION_ATTR_SET_NW_TOS:
- if (nla_get_u8(a) & INET_ECN_MASK)
- return -EINVAL;
+ case OVS_ACTION_ATTR_SET:
+ err = validate_set(a, key);
+ if (err)
+ return err;
+ break;
+
+ case OVS_ACTION_ATTR_SAMPLE:
+ err = validate_sample(a, key, depth);
+ if (err)
+ return err;
break;
default:
- return -EOPNOTSUPP;
+ return -EINVAL;
}
}
return 0;
}
+
static void clear_stats(struct sw_flow *flow)
{
flow->used = 0;
struct sw_flow *flow;
struct datapath *dp;
struct ethhdr *eth;
- bool is_frag;
int len;
int err;
int key_len;
nla_len(a[OVS_PACKET_ATTR_PACKET]) < ETH_HLEN)
goto err;
- err = validate_actions(a[OVS_PACKET_ATTR_ACTIONS]);
- if (err)
- goto err;
-
len = nla_len(a[OVS_PACKET_ATTR_PACKET]);
packet = __dev_alloc_skb(NET_IP_ALIGN + len, GFP_KERNEL);
err = -ENOMEM;
packet->protocol = htons(ETH_P_802_2);
/* Build an sw_flow for sending this packet. */
- flow = flow_alloc();
+ flow = ovs_flow_alloc();
err = PTR_ERR(flow);
if (IS_ERR(flow))
goto err_kfree_skb;
- err = flow_extract(packet, -1, &flow->key, &key_len, &is_frag);
+ err = ovs_flow_extract(packet, -1, &flow->key, &key_len);
if (err)
- goto err_flow_put;
+ goto err_flow_free;
- err = flow_metadata_from_nlattrs(&flow->key.eth.in_port,
- &flow->key.eth.tun_id,
- a[OVS_PACKET_ATTR_KEY]);
+ err = ovs_flow_metadata_from_nlattrs(flow, key_len, a[OVS_PACKET_ATTR_KEY]);
if (err)
- goto err_flow_put;
+ goto err_flow_free;
- flow->hash = flow_hash(&flow->key, key_len);
+ err = validate_actions(a[OVS_PACKET_ATTR_ACTIONS], &flow->key, 0);
+ if (err)
+ goto err_flow_free;
- acts = flow_actions_alloc(a[OVS_PACKET_ATTR_ACTIONS]);
+ acts = ovs_flow_actions_alloc(a[OVS_PACKET_ATTR_ACTIONS]);
err = PTR_ERR(acts);
if (IS_ERR(acts))
- goto err_flow_put;
+ goto err_flow_free;
rcu_assign_pointer(flow->sf_acts, acts);
OVS_CB(packet)->flow = flow;
+ packet->priority = flow->key.phy.priority;
rcu_read_lock();
- dp = get_dp(ovs_header->dp_ifindex);
+ dp = get_dp(sock_net(skb->sk), ovs_header->dp_ifindex);
err = -ENODEV;
if (!dp)
goto err_unlock;
- if (flow->key.eth.in_port < DP_MAX_PORTS)
- OVS_CB(packet)->vport = get_vport_protected(dp,
- flow->key.eth.in_port);
-
- err = execute_actions(dp, packet);
+ local_bh_disable();
+ err = ovs_execute_actions(dp, packet);
+ local_bh_enable();
rcu_read_unlock();
- flow_put(flow);
+ ovs_flow_free(flow);
return err;
err_unlock:
rcu_read_unlock();
-err_flow_put:
- flow_put(flow);
+err_flow_free:
+ ovs_flow_free(flow);
err_kfree_skb:
kfree_skb(packet);
err:
static void get_dp_stats(struct datapath *dp, struct ovs_dp_stats *stats)
{
int i;
- struct flow_table *table = get_table_protected(dp);
+ struct flow_table *table = genl_dereference(dp->table);
- stats->n_flows = flow_tbl_count(table);
+ stats->n_flows = ovs_flow_tbl_count(table);
- stats->n_frags = stats->n_hit = stats->n_missed = stats->n_lost = 0;
+ stats->n_hit = stats->n_missed = stats->n_lost = 0;
for_each_possible_cpu(i) {
const struct dp_stats_percpu *percpu_stats;
struct dp_stats_percpu local_stats;
- unsigned seqcount;
+ unsigned int start;
percpu_stats = per_cpu_ptr(dp->stats_percpu, i);
do {
- seqcount = read_seqcount_begin(&percpu_stats->seqlock);
+ start = u64_stats_fetch_begin_bh(&percpu_stats->sync);
local_stats = *percpu_stats;
- } while (read_seqcount_retry(&percpu_stats->seqlock, seqcount));
+ } while (u64_stats_fetch_retry_bh(&percpu_stats->sync, start));
- stats->n_frags += local_stats.n_frags;
stats->n_hit += local_stats.n_hit;
stats->n_missed += local_stats.n_missed;
stats->n_lost += local_stats.n_lost;
.id = GENL_ID_GENERATE,
.hdrsize = sizeof(struct ovs_header),
.name = OVS_FLOW_FAMILY,
- .version = 1,
- .maxattr = OVS_FLOW_ATTR_MAX
+ .version = OVS_FLOW_VERSION,
+ .maxattr = OVS_FLOW_ATTR_MAX,
+ SET_NETNSOK
};
-static struct genl_multicast_group dp_flow_multicast_group = {
+static struct genl_multicast_group ovs_dp_flow_multicast_group = {
.name = OVS_FLOW_MCGROUP
};
/* Called with genl_lock. */
static int ovs_flow_cmd_fill_info(struct sw_flow *flow, struct datapath *dp,
- struct sk_buff *skb, u32 pid, u32 seq, u32 flags, u8 cmd)
+ struct sk_buff *skb, u32 pid,
+ u32 seq, u32 flags, u8 cmd)
{
const int skb_orig_len = skb->len;
const struct sw_flow_actions *sf_acts;
if (!ovs_header)
return -EMSGSIZE;
- ovs_header->dp_ifindex = dp->dp_ifindex;
+ ovs_header->dp_ifindex = get_dpifindex(dp);
nla = nla_nest_start(skb, OVS_FLOW_ATTR_KEY);
if (!nla)
goto nla_put_failure;
- err = flow_to_nlattrs(&flow->key, skb);
+ err = ovs_flow_to_nlattrs(&flow->key, skb);
if (err)
goto error;
nla_nest_end(skb, nla);
tcp_flags = flow->tcp_flags;
spin_unlock_bh(&flow->lock);
- if (used)
- NLA_PUT_U64(skb, OVS_FLOW_ATTR_USED, flow_used_time(used));
+ if (used &&
+ nla_put_u64(skb, OVS_FLOW_ATTR_USED, ovs_flow_used_time(used)))
+ goto nla_put_failure;
- if (stats.n_packets)
- NLA_PUT(skb, OVS_FLOW_ATTR_STATS, sizeof(struct ovs_flow_stats), &stats);
+ if (stats.n_packets &&
+ nla_put(skb, OVS_FLOW_ATTR_STATS,
+ sizeof(struct ovs_flow_stats), &stats))
+ goto nla_put_failure;
- if (tcp_flags)
- NLA_PUT_U8(skb, OVS_FLOW_ATTR_TCP_FLAGS, tcp_flags);
+ if (tcp_flags &&
+ nla_put_u8(skb, OVS_FLOW_ATTR_TCP_FLAGS, tcp_flags))
+ goto nla_put_failure;
/* If OVS_FLOW_ATTR_ACTIONS doesn't fit, skip dumping the actions if
* this is the first flow to be dumped into 'skb'. This is unusual for
sf_acts = rcu_dereference_protected(flow->sf_acts,
lockdep_genl_is_held());
- len = nla_total_size(FLOW_BUFSIZE); /* OVS_FLOW_ATTR_KEY */
- len += nla_total_size(sf_acts->actions_len); /* OVS_FLOW_ATTR_ACTIONS */
- len += nla_total_size(sizeof(struct ovs_flow_stats)); /* OVS_FLOW_ATTR_STATS */
- len += nla_total_size(1); /* OVS_FLOW_ATTR_TCP_FLAGS */
- len += nla_total_size(8); /* OVS_FLOW_ATTR_USED */
- return genlmsg_new(NLMSG_ALIGN(sizeof(struct ovs_header)) + len, GFP_KERNEL);
+ /* OVS_FLOW_ATTR_KEY */
+ len = nla_total_size(FLOW_BUFSIZE);
+ /* OVS_FLOW_ATTR_ACTIONS */
+ len += nla_total_size(sf_acts->actions_len);
+ /* OVS_FLOW_ATTR_STATS */
+ len += nla_total_size(sizeof(struct ovs_flow_stats));
+ /* OVS_FLOW_ATTR_TCP_FLAGS */
+ len += nla_total_size(1);
+ /* OVS_FLOW_ATTR_USED */
+ len += nla_total_size(8);
+
+ len += NLMSG_ALIGN(sizeof(struct ovs_header));
+
+ return genlmsg_new(len, GFP_KERNEL);
}
-static struct sk_buff *ovs_flow_cmd_build_info(struct sw_flow *flow, struct datapath *dp,
+static struct sk_buff *ovs_flow_cmd_build_info(struct sw_flow *flow,
+ struct datapath *dp,
u32 pid, u32 seq, u8 cmd)
{
struct sk_buff *skb;
error = -EINVAL;
if (!a[OVS_FLOW_ATTR_KEY])
goto error;
- error = flow_from_nlattrs(&key, &key_len, a[OVS_FLOW_ATTR_KEY]);
+ error = ovs_flow_from_nlattrs(&key, &key_len, a[OVS_FLOW_ATTR_KEY]);
if (error)
goto error;
/* Validate actions. */
if (a[OVS_FLOW_ATTR_ACTIONS]) {
- error = validate_actions(a[OVS_FLOW_ATTR_ACTIONS]);
+ error = validate_actions(a[OVS_FLOW_ATTR_ACTIONS], &key, 0);
if (error)
goto error;
} else if (info->genlhdr->cmd == OVS_FLOW_CMD_NEW) {
goto error;
}
- dp = get_dp(ovs_header->dp_ifindex);
+ dp = get_dp(sock_net(skb->sk), ovs_header->dp_ifindex);
error = -ENODEV;
if (!dp)
goto error;
- table = get_table_protected(dp);
- flow = flow_tbl_lookup(table, &key, key_len);
+ table = genl_dereference(dp->table);
+ flow = ovs_flow_tbl_lookup(table, &key, key_len);
if (!flow) {
struct sw_flow_actions *acts;
goto error;
/* Expand table, if necessary, to make room. */
- if (flow_tbl_need_to_expand(table)) {
+ if (ovs_flow_tbl_need_to_expand(table)) {
struct flow_table *new_table;
- new_table = flow_tbl_expand(table);
+ new_table = ovs_flow_tbl_expand(table);
if (!IS_ERR(new_table)) {
rcu_assign_pointer(dp->table, new_table);
- flow_tbl_deferred_destroy(table);
- table = get_table_protected(dp);
+ ovs_flow_tbl_deferred_destroy(table);
+ table = genl_dereference(dp->table);
}
}
/* Allocate flow. */
- flow = flow_alloc();
+ flow = ovs_flow_alloc();
if (IS_ERR(flow)) {
error = PTR_ERR(flow);
goto error;
}
- flow->key = key;
clear_stats(flow);
/* Obtain actions. */
- acts = flow_actions_alloc(a[OVS_FLOW_ATTR_ACTIONS]);
+ acts = ovs_flow_actions_alloc(a[OVS_FLOW_ATTR_ACTIONS]);
error = PTR_ERR(acts);
if (IS_ERR(acts))
goto error_free_flow;
rcu_assign_pointer(flow->sf_acts, acts);
/* Put flow in bucket. */
- flow->hash = flow_hash(&key, key_len);
- flow_tbl_insert(table, flow);
+ ovs_flow_tbl_insert(table, flow, &key, key_len);
reply = ovs_flow_cmd_build_info(flow, dp, info->snd_pid,
- info->snd_seq, OVS_FLOW_CMD_NEW);
+ info->snd_seq,
+ OVS_FLOW_CMD_NEW);
} else {
/* We found a matching flow. */
struct sw_flow_actions *old_acts;
+ struct nlattr *acts_attrs;
/* Bail out if we're not allowed to modify an existing flow.
* We accept NLM_F_CREATE in place of the intended NLM_F_EXCL
/* Update actions. */
old_acts = rcu_dereference_protected(flow->sf_acts,
lockdep_genl_is_held());
- if (a[OVS_FLOW_ATTR_ACTIONS] &&
- (old_acts->actions_len != nla_len(a[OVS_FLOW_ATTR_ACTIONS]) ||
- memcmp(old_acts->actions, nla_data(a[OVS_FLOW_ATTR_ACTIONS]),
- old_acts->actions_len))) {
+ acts_attrs = a[OVS_FLOW_ATTR_ACTIONS];
+ if (acts_attrs &&
+ (old_acts->actions_len != nla_len(acts_attrs) ||
+ memcmp(old_acts->actions, nla_data(acts_attrs),
+ old_acts->actions_len))) {
struct sw_flow_actions *new_acts;
- new_acts = flow_actions_alloc(a[OVS_FLOW_ATTR_ACTIONS]);
+ new_acts = ovs_flow_actions_alloc(acts_attrs);
error = PTR_ERR(new_acts);
if (IS_ERR(new_acts))
goto error;
rcu_assign_pointer(flow->sf_acts, new_acts);
- flow_deferred_free_acts(old_acts);
+ ovs_flow_deferred_free_acts(old_acts);
}
reply = ovs_flow_cmd_build_info(flow, dp, info->snd_pid,
- info->snd_seq, OVS_FLOW_CMD_NEW);
+ info->snd_seq, OVS_FLOW_CMD_NEW);
/* Clear stats. */
if (a[OVS_FLOW_ATTR_CLEAR]) {
if (!IS_ERR(reply))
genl_notify(reply, genl_info_net(info), info->snd_pid,
- dp_flow_multicast_group.id, info->nlhdr, GFP_KERNEL);
+ ovs_dp_flow_multicast_group.id, info->nlhdr,
+ GFP_KERNEL);
else
- netlink_set_err(INIT_NET_GENL_SOCK, 0,
- dp_flow_multicast_group.id, PTR_ERR(reply));
+ netlink_set_err(GENL_SOCK(sock_net(skb->sk)), 0,
+ ovs_dp_flow_multicast_group.id, PTR_ERR(reply));
return 0;
error_free_flow:
- flow_put(flow);
+ ovs_flow_free(flow);
error:
return error;
}
if (!a[OVS_FLOW_ATTR_KEY])
return -EINVAL;
- err = flow_from_nlattrs(&key, &key_len, a[OVS_FLOW_ATTR_KEY]);
+ err = ovs_flow_from_nlattrs(&key, &key_len, a[OVS_FLOW_ATTR_KEY]);
if (err)
return err;
- dp = get_dp(ovs_header->dp_ifindex);
+ dp = get_dp(sock_net(skb->sk), ovs_header->dp_ifindex);
if (!dp)
return -ENODEV;
- table = get_table_protected(dp);
- flow = flow_tbl_lookup(table, &key, key_len);
+ table = genl_dereference(dp->table);
+ flow = ovs_flow_tbl_lookup(table, &key, key_len);
if (!flow)
return -ENOENT;
- reply = ovs_flow_cmd_build_info(flow, dp, info->snd_pid, info->snd_seq, OVS_FLOW_CMD_NEW);
+ reply = ovs_flow_cmd_build_info(flow, dp, info->snd_pid,
+ info->snd_seq, OVS_FLOW_CMD_NEW);
if (IS_ERR(reply))
return PTR_ERR(reply);
int err;
int key_len;
+ dp = get_dp(sock_net(skb->sk), ovs_header->dp_ifindex);
+ if (!dp)
+ return -ENODEV;
+
if (!a[OVS_FLOW_ATTR_KEY])
- return flush_flows(ovs_header->dp_ifindex);
- err = flow_from_nlattrs(&key, &key_len, a[OVS_FLOW_ATTR_KEY]);
+ return flush_flows(dp);
+
+ err = ovs_flow_from_nlattrs(&key, &key_len, a[OVS_FLOW_ATTR_KEY]);
if (err)
return err;
- dp = get_dp(ovs_header->dp_ifindex);
- if (!dp)
- return -ENODEV;
-
- table = get_table_protected(dp);
- flow = flow_tbl_lookup(table, &key, key_len);
+ table = genl_dereference(dp->table);
+ flow = ovs_flow_tbl_lookup(table, &key, key_len);
if (!flow)
return -ENOENT;
if (!reply)
return -ENOMEM;
- flow_tbl_remove(table, flow);
+ ovs_flow_tbl_remove(table, flow);
err = ovs_flow_cmd_fill_info(flow, dp, reply, info->snd_pid,
info->snd_seq, 0, OVS_FLOW_CMD_DEL);
BUG_ON(err < 0);
- flow_deferred_free(flow);
+ ovs_flow_deferred_free(flow);
genl_notify(reply, genl_info_net(info), info->snd_pid,
- dp_flow_multicast_group.id, info->nlhdr, GFP_KERNEL);
+ ovs_dp_flow_multicast_group.id, info->nlhdr, GFP_KERNEL);
return 0;
}
{
struct ovs_header *ovs_header = genlmsg_data(nlmsg_data(cb->nlh));
struct datapath *dp;
+ struct flow_table *table;
- dp = get_dp(ovs_header->dp_ifindex);
+ dp = get_dp(sock_net(skb->sk), ovs_header->dp_ifindex);
if (!dp)
return -ENODEV;
+ table = genl_dereference(dp->table);
+
for (;;) {
struct sw_flow *flow;
u32 bucket, obj;
bucket = cb->args[0];
obj = cb->args[1];
- flow = flow_tbl_next(get_table_protected(dp), &bucket, &obj);
+ flow = ovs_flow_tbl_next(table, &bucket, &obj);
if (!flow)
break;
- if (ovs_flow_cmd_fill_info(flow, dp, skb, NETLINK_CB(cb->skb).pid,
+ if (ovs_flow_cmd_fill_info(flow, dp, skb,
+ NETLINK_CB(cb->skb).pid,
cb->nlh->nlmsg_seq, NLM_F_MULTI,
OVS_FLOW_CMD_NEW) < 0)
break;
#ifdef HAVE_NLA_NUL_STRING
[OVS_DP_ATTR_NAME] = { .type = NLA_NUL_STRING, .len = IFNAMSIZ - 1 },
#endif
- [OVS_DP_ATTR_IPV4_FRAGS] = { .type = NLA_U32 },
- [OVS_DP_ATTR_SAMPLING] = { .type = NLA_U32 },
+ [OVS_DP_ATTR_UPCALL_PID] = { .type = NLA_U32 },
};
static struct genl_family dp_datapath_genl_family = {
.id = GENL_ID_GENERATE,
.hdrsize = sizeof(struct ovs_header),
.name = OVS_DATAPATH_FAMILY,
- .version = 1,
- .maxattr = OVS_DP_ATTR_MAX
+ .version = OVS_DATAPATH_VERSION,
+ .maxattr = OVS_DP_ATTR_MAX,
+ SET_NETNSOK
};
-static struct genl_multicast_group dp_datapath_multicast_group = {
+static struct genl_multicast_group ovs_dp_datapath_multicast_group = {
.name = OVS_DATAPATH_MCGROUP
};
u32 pid, u32 seq, u32 flags, u8 cmd)
{
struct ovs_header *ovs_header;
- struct nlattr *nla;
+ struct ovs_dp_stats dp_stats;
int err;
ovs_header = genlmsg_put(skb, pid, seq, &dp_datapath_genl_family,
if (!ovs_header)
goto error;
- ovs_header->dp_ifindex = dp->dp_ifindex;
+ ovs_header->dp_ifindex = get_dpifindex(dp);
rcu_read_lock();
- err = nla_put_string(skb, OVS_DP_ATTR_NAME, dp_name(dp));
+ err = nla_put_string(skb, OVS_DP_ATTR_NAME, ovs_dp_name(dp));
rcu_read_unlock();
if (err)
goto nla_put_failure;
- nla = nla_reserve(skb, OVS_DP_ATTR_STATS, sizeof(struct ovs_dp_stats));
- if (!nla)
+ get_dp_stats(dp, &dp_stats);
+ if (nla_put(skb, OVS_DP_ATTR_STATS, sizeof(struct ovs_dp_stats), &dp_stats))
goto nla_put_failure;
- get_dp_stats(dp, nla_data(nla));
-
- NLA_PUT_U32(skb, OVS_DP_ATTR_IPV4_FRAGS,
- dp->drop_frags ? OVS_DP_FRAG_DROP : OVS_DP_FRAG_ZERO);
-
- if (dp->sflow_probability)
- NLA_PUT_U32(skb, OVS_DP_ATTR_SAMPLING, dp->sflow_probability);
-
- nla = nla_nest_start(skb, OVS_DP_ATTR_MCGROUPS);
- if (!nla)
- goto nla_put_failure;
- NLA_PUT_U32(skb, OVS_PACKET_CMD_MISS, packet_mc_group(dp, OVS_PACKET_CMD_MISS));
- NLA_PUT_U32(skb, OVS_PACKET_CMD_ACTION, packet_mc_group(dp, OVS_PACKET_CMD_ACTION));
- NLA_PUT_U32(skb, OVS_PACKET_CMD_SAMPLE, packet_mc_group(dp, OVS_PACKET_CMD_SAMPLE));
- nla_nest_end(skb, nla);
return genlmsg_end(skb, ovs_header);
static int ovs_dp_cmd_validate(struct nlattr *a[OVS_DP_ATTR_MAX + 1])
{
- if (a[OVS_DP_ATTR_IPV4_FRAGS]) {
- u32 frags = nla_get_u32(a[OVS_DP_ATTR_IPV4_FRAGS]);
-
- if (frags != OVS_DP_FRAG_ZERO && frags != OVS_DP_FRAG_DROP)
- return -EINVAL;
- }
-
return CHECK_NUL_STRING(a[OVS_DP_ATTR_NAME], IFNAMSIZ - 1);
}
/* Called with genl_mutex and optionally with RTNL lock also. */
-static struct datapath *lookup_datapath(struct ovs_header *ovs_header, struct nlattr *a[OVS_DP_ATTR_MAX + 1])
+static struct datapath *lookup_datapath(struct net *net,
+ struct ovs_header *ovs_header,
+ struct nlattr *a[OVS_DP_ATTR_MAX + 1])
{
struct datapath *dp;
if (!a[OVS_DP_ATTR_NAME])
- dp = get_dp(ovs_header->dp_ifindex);
+ dp = get_dp(net, ovs_header->dp_ifindex);
else {
struct vport *vport;
rcu_read_lock();
- vport = vport_locate(nla_data(a[OVS_DP_ATTR_NAME]));
+ vport = ovs_vport_locate(net, nla_data(a[OVS_DP_ATTR_NAME]));
dp = vport && vport->port_no == OVSP_LOCAL ? vport->dp : NULL;
rcu_read_unlock();
}
return dp ? dp : ERR_PTR(-ENODEV);
}
-/* Called with genl_mutex. */
-static void change_datapath(struct datapath *dp, struct nlattr *a[OVS_DP_ATTR_MAX + 1])
-{
- if (a[OVS_DP_ATTR_IPV4_FRAGS])
- dp->drop_frags = nla_get_u32(a[OVS_DP_ATTR_IPV4_FRAGS]) == OVS_DP_FRAG_DROP;
- if (a[OVS_DP_ATTR_SAMPLING])
- dp->sflow_probability = nla_get_u32(a[OVS_DP_ATTR_SAMPLING]);
-}
-
static int ovs_dp_cmd_new(struct sk_buff *skb, struct genl_info *info)
{
struct nlattr **a = info->attrs;
struct sk_buff *reply;
struct datapath *dp;
struct vport *vport;
- int err;
+ struct ovs_net *ovs_net;
+ int err, i;
err = -EINVAL;
- if (!a[OVS_DP_ATTR_NAME])
+ if (!a[OVS_DP_ATTR_NAME] || !a[OVS_DP_ATTR_UPCALL_PID])
goto err;
err = ovs_dp_cmd_validate(a);
goto err;
rtnl_lock();
- err = -ENODEV;
- if (!try_module_get(THIS_MODULE))
- goto err_unlock_rtnl;
err = -ENOMEM;
dp = kzalloc(sizeof(*dp), GFP_KERNEL);
if (dp == NULL)
- goto err_put_module;
- INIT_LIST_HEAD(&dp->port_list);
+ goto err_unlock_rtnl;
/* Initialize kobject for bridge. This will be added as
* /sys/class/net/<devname>/brif later, if sysfs is enabled. */
dp->ifobj.kset = NULL;
kobject_init(&dp->ifobj, &dp_ktype);
+ ovs_dp_set_net(dp, hold_net(sock_net(skb->sk)));
+
/* Allocate table. */
err = -ENOMEM;
- rcu_assign_pointer(dp->table, flow_tbl_alloc(TBL_MIN_BUCKETS));
+ rcu_assign_pointer(dp->table, ovs_flow_tbl_alloc(TBL_MIN_BUCKETS));
if (!dp->table)
goto err_free_dp;
+ dp->stats_percpu = alloc_percpu(struct dp_stats_percpu);
+ if (!dp->stats_percpu) {
+ err = -ENOMEM;
+ goto err_destroy_table;
+ }
+
+ dp->ports = kmalloc(DP_VPORT_HASH_BUCKETS * sizeof(struct hlist_head),
+ GFP_KERNEL);
+ if (!dp->ports) {
+ err = -ENOMEM;
+ goto err_destroy_percpu;
+ }
+
+ for (i = 0; i < DP_VPORT_HASH_BUCKETS; i++)
+ INIT_HLIST_HEAD(&dp->ports[i]);
+
/* Set up our datapath device. */
parms.name = nla_data(a[OVS_DP_ATTR_NAME]);
parms.type = OVS_VPORT_TYPE_INTERNAL;
parms.options = NULL;
parms.dp = dp;
parms.port_no = OVSP_LOCAL;
+ parms.upcall_pid = nla_get_u32(a[OVS_DP_ATTR_UPCALL_PID]);
+
vport = new_vport(&parms);
if (IS_ERR(vport)) {
err = PTR_ERR(vport);
if (err == -EBUSY)
err = -EEXIST;
- goto err_destroy_table;
- }
- dp->dp_ifindex = vport_get_ifindex(vport);
-
- dp->drop_frags = 0;
- dp->stats_percpu = alloc_percpu(struct dp_stats_percpu);
- if (!dp->stats_percpu) {
- err = -ENOMEM;
- goto err_destroy_local_port;
+ goto err_destroy_ports_array;
}
- change_datapath(dp, a);
-
- reply = ovs_dp_cmd_build_info(dp, info->snd_pid, info->snd_seq, OVS_DP_CMD_NEW);
+ reply = ovs_dp_cmd_build_info(dp, info->snd_pid,
+ info->snd_seq, OVS_DP_CMD_NEW);
err = PTR_ERR(reply);
if (IS_ERR(reply))
goto err_destroy_local_port;
- list_add_tail(&dp->list_node, &dps);
- dp_sysfs_add_dp(dp);
+ ovs_net = net_generic(ovs_dp_get_net(dp), ovs_net_id);
+ list_add_tail(&dp->list_node, &ovs_net->dps);
+ ovs_dp_sysfs_add_dp(dp);
rtnl_unlock();
genl_notify(reply, genl_info_net(info), info->snd_pid,
- dp_datapath_multicast_group.id, info->nlhdr, GFP_KERNEL);
+ ovs_dp_datapath_multicast_group.id, info->nlhdr,
+ GFP_KERNEL);
return 0;
err_destroy_local_port:
- dp_detach_port(get_vport_protected(dp, OVSP_LOCAL));
+ ovs_dp_detach_port(ovs_vport_rtnl(dp, OVSP_LOCAL));
+err_destroy_ports_array:
+ kfree(dp->ports);
+err_destroy_percpu:
+ free_percpu(dp->stats_percpu);
err_destroy_table:
- flow_tbl_destroy(get_table_protected(dp));
+ ovs_flow_tbl_destroy(genl_dereference(dp->table));
err_free_dp:
+ release_net(ovs_dp_get_net(dp));
kfree(dp);
-err_put_module:
- module_put(THIS_MODULE);
err_unlock_rtnl:
rtnl_unlock();
err:
return err;
}
-static int ovs_dp_cmd_del(struct sk_buff *skb, struct genl_info *info)
+/* Called with genl_mutex. */
+static void __dp_destroy(struct datapath *dp)
{
- struct vport *vport, *next_vport;
- struct sk_buff *reply;
- struct datapath *dp;
- int err;
-
- err = ovs_dp_cmd_validate(info->attrs);
- if (err)
- goto exit;
+ int i;
rtnl_lock();
- dp = lookup_datapath(info->userhdr, info->attrs);
- err = PTR_ERR(dp);
- if (IS_ERR(dp))
- goto exit_unlock;
- reply = ovs_dp_cmd_build_info(dp, info->snd_pid, info->snd_seq, OVS_DP_CMD_DEL);
- err = PTR_ERR(reply);
- if (IS_ERR(reply))
- goto exit_unlock;
+ for (i = 0; i < DP_VPORT_HASH_BUCKETS; i++) {
+ struct vport *vport;
+ struct hlist_node *node, *n;
- list_for_each_entry_safe (vport, next_vport, &dp->port_list, node)
- if (vport->port_no != OVSP_LOCAL)
- dp_detach_port(vport);
+ hlist_for_each_entry_safe(vport, node, n, &dp->ports[i], dp_hash_node)
+ if (vport->port_no != OVSP_LOCAL)
+ ovs_dp_detach_port(vport);
+ }
- dp_sysfs_del_dp(dp);
+ ovs_dp_sysfs_del_dp(dp);
list_del(&dp->list_node);
- dp_detach_port(get_vport_protected(dp, OVSP_LOCAL));
+ ovs_dp_detach_port(ovs_vport_rtnl(dp, OVSP_LOCAL));
/* rtnl_unlock() will wait until all the references to devices that
* are pending unregistration have been dropped. We do it here to
rtnl_unlock();
call_rcu(&dp->rcu, destroy_dp_rcu);
- module_put(THIS_MODULE);
+}
+
+static int ovs_dp_cmd_del(struct sk_buff *skb, struct genl_info *info)
+{
+ struct sk_buff *reply;
+ struct datapath *dp;
+ int err;
+
+ err = ovs_dp_cmd_validate(info->attrs);
+ if (err)
+ return err;
+
+ dp = lookup_datapath(sock_net(skb->sk), info->userhdr, info->attrs);
+ err = PTR_ERR(dp);
+ if (IS_ERR(dp))
+ return err;
+
+ reply = ovs_dp_cmd_build_info(dp, info->snd_pid,
+ info->snd_seq, OVS_DP_CMD_DEL);
+ err = PTR_ERR(reply);
+ if (IS_ERR(reply))
+ return err;
+
+ __dp_destroy(dp);
genl_notify(reply, genl_info_net(info), info->snd_pid,
- dp_datapath_multicast_group.id, info->nlhdr, GFP_KERNEL);
+ ovs_dp_datapath_multicast_group.id, info->nlhdr,
+ GFP_KERNEL);
return 0;
-
-exit_unlock:
- rtnl_unlock();
-exit:
- return err;
}
static int ovs_dp_cmd_set(struct sk_buff *skb, struct genl_info *info)
if (err)
return err;
- dp = lookup_datapath(info->userhdr, info->attrs);
+ dp = lookup_datapath(sock_net(skb->sk), info->userhdr, info->attrs);
if (IS_ERR(dp))
return PTR_ERR(dp);
- change_datapath(dp, info->attrs);
-
- reply = ovs_dp_cmd_build_info(dp, info->snd_pid, info->snd_seq, OVS_DP_CMD_NEW);
+ reply = ovs_dp_cmd_build_info(dp, info->snd_pid,
+ info->snd_seq, OVS_DP_CMD_NEW);
if (IS_ERR(reply)) {
err = PTR_ERR(reply);
- netlink_set_err(INIT_NET_GENL_SOCK, 0,
- dp_datapath_multicast_group.id, err);
+ netlink_set_err(GENL_SOCK(sock_net(skb->sk)), 0,
+ ovs_dp_datapath_multicast_group.id, err);
return 0;
}
genl_notify(reply, genl_info_net(info), info->snd_pid,
- dp_datapath_multicast_group.id, info->nlhdr, GFP_KERNEL);
+ ovs_dp_datapath_multicast_group.id, info->nlhdr,
+ GFP_KERNEL);
+
return 0;
}
if (err)
return err;
- dp = lookup_datapath(info->userhdr, info->attrs);
+ dp = lookup_datapath(sock_net(skb->sk), info->userhdr, info->attrs);
if (IS_ERR(dp))
return PTR_ERR(dp);
- reply = ovs_dp_cmd_build_info(dp, info->snd_pid, info->snd_seq, OVS_DP_CMD_NEW);
+ reply = ovs_dp_cmd_build_info(dp, info->snd_pid,
+ info->snd_seq, OVS_DP_CMD_NEW);
if (IS_ERR(reply))
return PTR_ERR(reply);
static int ovs_dp_cmd_dump(struct sk_buff *skb, struct netlink_callback *cb)
{
+ struct ovs_net *ovs_net = net_generic(sock_net(skb->sk), ovs_net_id);
struct datapath *dp;
int skip = cb->args[0];
int i = 0;
- list_for_each_entry (dp, &dps, list_node) {
- if (i < skip)
- continue;
- if (ovs_dp_cmd_fill_info(dp, skb, NETLINK_CB(cb->skb).pid,
+ list_for_each_entry(dp, &ovs_net->dps, list_node) {
+ if (i >= skip &&
+ ovs_dp_cmd_fill_info(dp, skb, NETLINK_CB(cb->skb).pid,
cb->nlh->nlmsg_seq, NLM_F_MULTI,
OVS_DP_CMD_NEW) < 0)
break;
static const struct nla_policy vport_policy[OVS_VPORT_ATTR_MAX + 1] = {
#ifdef HAVE_NLA_NUL_STRING
[OVS_VPORT_ATTR_NAME] = { .type = NLA_NUL_STRING, .len = IFNAMSIZ - 1 },
- [OVS_VPORT_ATTR_PORT_NO] = { .type = NLA_U32 },
- [OVS_VPORT_ATTR_TYPE] = { .type = NLA_U32 },
[OVS_VPORT_ATTR_STATS] = { .len = sizeof(struct ovs_vport_stats) },
[OVS_VPORT_ATTR_ADDRESS] = { .len = ETH_ALEN },
#else
[OVS_VPORT_ATTR_STATS] = { .minlen = sizeof(struct ovs_vport_stats) },
[OVS_VPORT_ATTR_ADDRESS] = { .minlen = ETH_ALEN },
#endif
+ [OVS_VPORT_ATTR_PORT_NO] = { .type = NLA_U32 },
+ [OVS_VPORT_ATTR_TYPE] = { .type = NLA_U32 },
+ [OVS_VPORT_ATTR_UPCALL_PID] = { .type = NLA_U32 },
[OVS_VPORT_ATTR_OPTIONS] = { .type = NLA_NESTED },
};
.id = GENL_ID_GENERATE,
.hdrsize = sizeof(struct ovs_header),
.name = OVS_VPORT_FAMILY,
- .version = 1,
- .maxattr = OVS_VPORT_ATTR_MAX
+ .version = OVS_VPORT_VERSION,
+ .maxattr = OVS_VPORT_ATTR_MAX,
+ SET_NETNSOK
};
-struct genl_multicast_group dp_vport_multicast_group = {
+struct genl_multicast_group ovs_dp_vport_multicast_group = {
.name = OVS_VPORT_MCGROUP
};
u32 pid, u32 seq, u32 flags, u8 cmd)
{
struct ovs_header *ovs_header;
- struct nlattr *nla;
- int ifindex;
+ struct ovs_vport_stats vport_stats;
int err;
ovs_header = genlmsg_put(skb, pid, seq, &dp_vport_genl_family,
if (!ovs_header)
return -EMSGSIZE;
- ovs_header->dp_ifindex = vport->dp->dp_ifindex;
+ ovs_header->dp_ifindex = get_dpifindex(vport->dp);
- NLA_PUT_U32(skb, OVS_VPORT_ATTR_PORT_NO, vport->port_no);
- NLA_PUT_U32(skb, OVS_VPORT_ATTR_TYPE, vport_get_type(vport));
- NLA_PUT_STRING(skb, OVS_VPORT_ATTR_NAME, vport_get_name(vport));
-
- nla = nla_reserve(skb, OVS_VPORT_ATTR_STATS, sizeof(struct ovs_vport_stats));
- if (!nla)
+ if (nla_put_u32(skb, OVS_VPORT_ATTR_PORT_NO, vport->port_no) ||
+ nla_put_u32(skb, OVS_VPORT_ATTR_TYPE, vport->ops->type) ||
+ nla_put_string(skb, OVS_VPORT_ATTR_NAME, vport->ops->get_name(vport)) ||
+ nla_put_u32(skb, OVS_VPORT_ATTR_UPCALL_PID, vport->upcall_pid))
goto nla_put_failure;
- vport_get_stats(vport, nla_data(nla));
+ ovs_vport_get_stats(vport, &vport_stats);
+ if (nla_put(skb, OVS_VPORT_ATTR_STATS, sizeof(struct ovs_vport_stats),
+ &vport_stats))
+ goto nla_put_failure;
- NLA_PUT(skb, OVS_VPORT_ATTR_ADDRESS, ETH_ALEN, vport_get_addr(vport));
+ if (nla_put(skb, OVS_VPORT_ATTR_ADDRESS, ETH_ALEN,
+ vport->ops->get_addr(vport)))
+ goto nla_put_failure;
- err = vport_get_options(vport, skb);
+ err = ovs_vport_get_options(vport, skb);
if (err == -EMSGSIZE)
goto error;
- ifindex = vport_get_ifindex(vport);
- if (ifindex > 0)
- NLA_PUT_U32(skb, OVS_VPORT_ATTR_IFINDEX, ifindex);
-
return genlmsg_end(skb, ovs_header);
nla_put_failure:
}
/* Called with RTNL lock or RCU read lock. */
-static struct vport *lookup_vport(struct ovs_header *ovs_header,
+static struct vport *lookup_vport(struct net *net,
+ struct ovs_header *ovs_header,
struct nlattr *a[OVS_VPORT_ATTR_MAX + 1])
{
struct datapath *dp;
struct vport *vport;
if (a[OVS_VPORT_ATTR_NAME]) {
- vport = vport_locate(nla_data(a[OVS_VPORT_ATTR_NAME]));
+ vport = ovs_vport_locate(net, nla_data(a[OVS_VPORT_ATTR_NAME]));
if (!vport)
return ERR_PTR(-ENODEV);
+ if (ovs_header->dp_ifindex &&
+ ovs_header->dp_ifindex != get_dpifindex(vport->dp))
+ return ERR_PTR(-ENODEV);
return vport;
} else if (a[OVS_VPORT_ATTR_PORT_NO]) {
u32 port_no = nla_get_u32(a[OVS_VPORT_ATTR_PORT_NO]);
if (port_no >= DP_MAX_PORTS)
return ERR_PTR(-EFBIG);
- dp = get_dp(ovs_header->dp_ifindex);
+ dp = get_dp(net, ovs_header->dp_ifindex);
if (!dp)
return ERR_PTR(-ENODEV);
- vport = get_vport_protected(dp, port_no);
+ vport = ovs_vport_rtnl_rcu(dp, port_no);
if (!vport)
return ERR_PTR(-ENOENT);
return vport;
}
/* Called with RTNL lock. */
-static int change_vport(struct vport *vport, struct nlattr *a[OVS_VPORT_ATTR_MAX + 1])
+static int change_vport(struct vport *vport,
+ struct nlattr *a[OVS_VPORT_ATTR_MAX + 1])
{
int err = 0;
if (a[OVS_VPORT_ATTR_STATS])
- vport_set_stats(vport, nla_data(a[OVS_VPORT_ATTR_STATS]));
+ ovs_vport_set_stats(vport, nla_data(a[OVS_VPORT_ATTR_STATS]));
if (a[OVS_VPORT_ATTR_ADDRESS])
- err = vport_set_addr(vport, nla_data(a[OVS_VPORT_ATTR_ADDRESS]));
+ err = ovs_vport_set_addr(vport, nla_data(a[OVS_VPORT_ATTR_ADDRESS]));
return err;
}
int err;
err = -EINVAL;
- if (!a[OVS_VPORT_ATTR_NAME] || !a[OVS_VPORT_ATTR_TYPE])
+ if (!a[OVS_VPORT_ATTR_NAME] || !a[OVS_VPORT_ATTR_TYPE] ||
+ !a[OVS_VPORT_ATTR_UPCALL_PID])
goto exit;
err = ovs_vport_cmd_validate(a);
goto exit;
rtnl_lock();
- dp = get_dp(ovs_header->dp_ifindex);
+ dp = get_dp(sock_net(skb->sk), ovs_header->dp_ifindex);
err = -ENODEV;
if (!dp)
goto exit_unlock;
if (port_no >= DP_MAX_PORTS)
goto exit_unlock;
- vport = get_vport_protected(dp, port_no);
+ vport = ovs_vport_rtnl(dp, port_no);
err = -EBUSY;
if (vport)
goto exit_unlock;
err = -EFBIG;
goto exit_unlock;
}
- vport = get_vport_protected(dp, port_no);
+ vport = ovs_vport_rtnl(dp, port_no);
if (!vport)
break;
}
parms.options = a[OVS_VPORT_ATTR_OPTIONS];
parms.dp = dp;
parms.port_no = port_no;
+ parms.upcall_pid = nla_get_u32(a[OVS_VPORT_ATTR_UPCALL_PID]);
vport = new_vport(&parms);
err = PTR_ERR(vport);
if (IS_ERR(vport))
goto exit_unlock;
- dp_sysfs_add_if(vport);
+ ovs_dp_sysfs_add_if(vport);
err = change_vport(vport, a);
if (!err) {
reply = ovs_vport_cmd_build_info(vport, info->snd_pid,
- info->snd_seq, OVS_VPORT_CMD_NEW);
+ info->snd_seq,
+ OVS_VPORT_CMD_NEW);
if (IS_ERR(reply))
err = PTR_ERR(reply);
}
if (err) {
- dp_detach_port(vport);
+ ovs_dp_detach_port(vport);
goto exit_unlock;
}
genl_notify(reply, genl_info_net(info), info->snd_pid,
- dp_vport_multicast_group.id, info->nlhdr, GFP_KERNEL);
-
+ ovs_dp_vport_multicast_group.id, info->nlhdr, GFP_KERNEL);
exit_unlock:
rtnl_unlock();
goto exit;
rtnl_lock();
- vport = lookup_vport(info->userhdr, a);
+ vport = lookup_vport(sock_net(skb->sk), info->userhdr, a);
err = PTR_ERR(vport);
if (IS_ERR(vport))
goto exit_unlock;
err = 0;
- if (a[OVS_VPORT_ATTR_OPTIONS])
- err = vport_set_options(vport, a[OVS_VPORT_ATTR_OPTIONS]);
+ if (a[OVS_VPORT_ATTR_TYPE] &&
+ nla_get_u32(a[OVS_VPORT_ATTR_TYPE]) != vport->ops->type)
+ err = -EINVAL;
+
+ if (!err && a[OVS_VPORT_ATTR_OPTIONS])
+ err = ovs_vport_set_options(vport, a[OVS_VPORT_ATTR_OPTIONS]);
if (!err)
err = change_vport(vport, a);
+ else
+ goto exit_unlock;
+ if (!err && a[OVS_VPORT_ATTR_UPCALL_PID])
+ vport->upcall_pid = nla_get_u32(a[OVS_VPORT_ATTR_UPCALL_PID]);
reply = ovs_vport_cmd_build_info(vport, info->snd_pid, info->snd_seq,
OVS_VPORT_CMD_NEW);
if (IS_ERR(reply)) {
- err = PTR_ERR(reply);
- netlink_set_err(INIT_NET_GENL_SOCK, 0,
- dp_vport_multicast_group.id, err);
- return 0;
+ netlink_set_err(GENL_SOCK(sock_net(skb->sk)), 0,
+ ovs_dp_vport_multicast_group.id, PTR_ERR(reply));
+ goto exit_unlock;
}
genl_notify(reply, genl_info_net(info), info->snd_pid,
- dp_vport_multicast_group.id, info->nlhdr, GFP_KERNEL);
+ ovs_dp_vport_multicast_group.id, info->nlhdr, GFP_KERNEL);
exit_unlock:
rtnl_unlock();
goto exit;
rtnl_lock();
- vport = lookup_vport(info->userhdr, a);
+ vport = lookup_vport(sock_net(skb->sk), info->userhdr, a);
err = PTR_ERR(vport);
if (IS_ERR(vport))
goto exit_unlock;
if (IS_ERR(reply))
goto exit_unlock;
- dp_detach_port(vport);
+ ovs_dp_detach_port(vport);
genl_notify(reply, genl_info_net(info), info->snd_pid,
- dp_vport_multicast_group.id, info->nlhdr, GFP_KERNEL);
+ ovs_dp_vport_multicast_group.id, info->nlhdr, GFP_KERNEL);
exit_unlock:
rtnl_unlock();
goto exit;
rcu_read_lock();
- vport = lookup_vport(ovs_header, a);
+ vport = lookup_vport(sock_net(skb->sk), ovs_header, a);
err = PTR_ERR(vport);
if (IS_ERR(vport))
goto exit_unlock;
{
struct ovs_header *ovs_header = genlmsg_data(nlmsg_data(cb->nlh));
struct datapath *dp;
- u32 port_no;
- int retval;
+ int bucket = cb->args[0], skip = cb->args[1];
+ int i, j = 0;
- dp = get_dp(ovs_header->dp_ifindex);
+ dp = get_dp(sock_net(skb->sk), ovs_header->dp_ifindex);
if (!dp)
return -ENODEV;
rcu_read_lock();
- for (port_no = cb->args[0]; port_no < DP_MAX_PORTS; port_no++) {
+ for (i = bucket; i < DP_VPORT_HASH_BUCKETS; i++) {
struct vport *vport;
-
- vport = get_vport_protected(dp, port_no);
- if (!vport)
- continue;
-
- if (ovs_vport_cmd_fill_info(vport, skb, NETLINK_CB(cb->skb).pid,
- cb->nlh->nlmsg_seq, NLM_F_MULTI,
- OVS_VPORT_CMD_NEW) < 0)
- break;
+ struct hlist_node *n;
+
+ j = 0;
+ hlist_for_each_entry_rcu(vport, n, &dp->ports[i], dp_hash_node) {
+ if (j >= skip &&
+ ovs_vport_cmd_fill_info(vport, skb,
+ NETLINK_CB(cb->skb).pid,
+ cb->nlh->nlmsg_seq,
+ NLM_F_MULTI,
+ OVS_VPORT_CMD_NEW) < 0)
+ goto out;
+
+ j++;
+ }
+ skip = 0;
}
+out:
rcu_read_unlock();
- cb->args[0] = port_no;
- retval = skb->len;
+ cb->args[0] = i;
+ cb->args[1] = j;
- return retval;
+ return skb->len;
}
static struct genl_ops dp_vport_genl_ops[] = {
static const struct genl_family_and_ops dp_genl_families[] = {
{ &dp_datapath_genl_family,
dp_datapath_genl_ops, ARRAY_SIZE(dp_datapath_genl_ops),
- &dp_datapath_multicast_group },
+ &ovs_dp_datapath_multicast_group },
{ &dp_vport_genl_family,
dp_vport_genl_ops, ARRAY_SIZE(dp_vport_genl_ops),
- &dp_vport_multicast_group },
+ &ovs_dp_vport_multicast_group },
{ &dp_flow_genl_family,
dp_flow_genl_ops, ARRAY_SIZE(dp_flow_genl_ops),
- &dp_flow_multicast_group },
+ &ovs_dp_flow_multicast_group },
{ &dp_packet_genl_family,
dp_packet_genl_ops, ARRAY_SIZE(dp_packet_genl_ops),
NULL },
}
}
- err = packet_register_mc_groups();
- if (err)
- goto error;
return 0;
error:
return err;
}
+static int __rehash_flow_table(void *dummy)
+{
+ struct datapath *dp;
+ struct net *net;
+
+ rtnl_lock();
+ for_each_net(net) {
+ struct ovs_net *ovs_net = net_generic(net, ovs_net_id);
+
+ list_for_each_entry(dp, &ovs_net->dps, list_node) {
+ struct flow_table *old_table = genl_dereference(dp->table);
+ struct flow_table *new_table;
+
+ new_table = ovs_flow_tbl_rehash(old_table);
+ if (!IS_ERR(new_table)) {
+ rcu_assign_pointer(dp->table, new_table);
+ ovs_flow_tbl_deferred_destroy(old_table);
+ }
+ }
+ }
+ rtnl_unlock();
+ return 0;
+}
+
+static void rehash_flow_table(struct work_struct *work)
+{
+ genl_exec(__rehash_flow_table, NULL);
+ schedule_delayed_work(&rehash_flow_wq, REHASH_FLOW_INTERVAL);
+}
+
+static int dp_destroy_all(void *data)
+{
+ struct datapath *dp, *dp_next;
+ struct ovs_net *ovs_net = data;
+
+ list_for_each_entry_safe(dp, dp_next, &ovs_net->dps, list_node)
+ __dp_destroy(dp);
+
+ return 0;
+}
+
+static int __net_init ovs_init_net(struct net *net)
+{
+ struct ovs_net *ovs_net = net_generic(net, ovs_net_id);
+
+ INIT_LIST_HEAD(&ovs_net->dps);
+ return 0;
+}
+
+static void __net_exit ovs_exit_net(struct net *net)
+{
+ struct ovs_net *ovs_net = net_generic(net, ovs_net_id);
+
+ genl_exec(dp_destroy_all, ovs_net);
+}
+
+static struct pernet_operations ovs_net_ops = {
+ .init = ovs_init_net,
+ .exit = ovs_exit_net,
+ .id = &ovs_net_id,
+ .size = sizeof(struct ovs_net),
+};
+
static int __init dp_init(void)
{
struct sk_buff *dummy_skb;
BUILD_BUG_ON(sizeof(struct ovs_skb_cb) > sizeof(dummy_skb->cb));
- printk("Open vSwitch %s, built "__DATE__" "__TIME__"\n", VERSION BUILDNR);
+ pr_info("Open vSwitch switching datapath %s, built "__DATE__" "__TIME__"\n",
+ VERSION);
- err = tnl_init();
+ err = genl_exec_init();
if (err)
goto error;
- err = flow_init();
+ err = ovs_workqueues_init();
+ if (err)
+ goto error_genl_exec;
+
+ err = ovs_tnl_init();
+ if (err)
+ goto error_wq;
+
+ err = ovs_flow_init();
if (err)
goto error_tnl_exit;
- err = vport_init();
+ err = ovs_vport_init();
if (err)
goto error_flow_exit;
- err = register_netdevice_notifier(&dp_device_notifier);
+ err = register_pernet_device(&ovs_net_ops);
if (err)
goto error_vport_exit;
+ err = register_netdevice_notifier(&ovs_dp_device_notifier);
+ if (err)
+ goto error_netns_exit;
+
err = dp_register_genl();
if (err < 0)
goto error_unreg_notifier;
+ schedule_delayed_work(&rehash_flow_wq, REHASH_FLOW_INTERVAL);
+
return 0;
error_unreg_notifier:
- unregister_netdevice_notifier(&dp_device_notifier);
+ unregister_netdevice_notifier(&ovs_dp_device_notifier);
+error_netns_exit:
+ unregister_pernet_device(&ovs_net_ops);
error_vport_exit:
- vport_exit();
+ ovs_vport_exit();
error_flow_exit:
- flow_exit();
+ ovs_flow_exit();
error_tnl_exit:
- tnl_exit();
+ ovs_tnl_exit();
+error_wq:
+ ovs_workqueues_exit();
+error_genl_exec:
+ genl_exec_exit();
error:
return err;
}
static void dp_cleanup(void)
{
- rcu_barrier();
+ cancel_delayed_work_sync(&rehash_flow_wq);
dp_unregister_genl(ARRAY_SIZE(dp_genl_families));
- unregister_netdevice_notifier(&dp_device_notifier);
- vport_exit();
- flow_exit();
- tnl_exit();
+ unregister_netdevice_notifier(&ovs_dp_device_notifier);
+ unregister_pernet_device(&ovs_net_ops);
+ rcu_barrier();
+ ovs_vport_exit();
+ ovs_flow_exit();
+ ovs_tnl_exit();
+ ovs_workqueues_exit();
+ genl_exec_exit();
}
module_init(dp_init);
MODULE_DESCRIPTION("Open vSwitch switching datapath");
MODULE_LICENSE("GPL");
+MODULE_VERSION(VERSION);