summaryrefslogtreecommitdiff
path: root/drivers/net/vrf.c
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/net/vrf.c')
-rw-r--r--drivers/net/vrf.c959
1 files changed, 752 insertions, 207 deletions
diff --git a/drivers/net/vrf.c b/drivers/net/vrf.c
index 95909e262ba4..571847a7f86d 100644
--- a/drivers/net/vrf.c
+++ b/drivers/net/vrf.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
/*
* vrf.c: device driver to encapsulate a VRF space
*
@@ -6,13 +7,9 @@
* Copyright (c) 2015 David Ahern <dsa@cumulusnetworks.com>
*
* Based on dummy, team and ipvlan drivers
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2 of the License, or
- * (at your option) any later version.
*/
+#include <linux/ethtool.h>
#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/netdevice.h>
@@ -25,9 +22,11 @@
#include <net/rtnetlink.h>
#include <linux/u64_stats_sync.h>
#include <linux/hashtable.h>
+#include <linux/spinlock_types.h>
#include <linux/inetdevice.h>
#include <net/arp.h>
+#include <net/flow.h>
#include <net/ip.h>
#include <net/ip_fib.h>
#include <net/ip6_fib.h>
@@ -36,15 +35,82 @@
#include <net/addrconf.h>
#include <net/l3mdev.h>
#include <net/fib_rules.h>
+#include <net/netdev_lock.h>
+#include <net/sch_generic.h>
#include <net/netns/generic.h>
+#include <net/netfilter/nf_conntrack.h>
#define DRV_NAME "vrf"
-#define DRV_VERSION "1.0"
+#define DRV_VERSION "1.1"
#define FIB_RULE_PREF 1000 /* default preference for FIB rules */
+#define HT_MAP_BITS 4
+#define HASH_INITVAL ((u32)0xcafef00d)
+
+struct vrf_map {
+ DECLARE_HASHTABLE(ht, HT_MAP_BITS);
+ spinlock_t vmap_lock;
+
+ /* shared_tables:
+ * count how many distinct tables do not comply with the strict mode
+ * requirement.
+ * shared_tables value must be 0 in order to enable the strict mode.
+ *
+ * example of the evolution of shared_tables:
+ * | time
+ * add vrf0 --> table 100 shared_tables = 0 | t0
+ * add vrf1 --> table 101 shared_tables = 0 | t1
+ * add vrf2 --> table 100 shared_tables = 1 | t2
+ * add vrf3 --> table 100 shared_tables = 1 | t3
+ * add vrf4 --> table 101 shared_tables = 2 v t4
+ *
+ * shared_tables is a "step function" (or "staircase function")
+ * and it is increased by one when the second vrf is associated to a
+ * table.
+ *
+ * at t2, vrf0 and vrf2 are bound to table 100: shared_tables = 1.
+ *
+ * at t3, another dev (vrf3) is bound to the same table 100 but the
+ * value of shared_tables is still 1.
+ * This means that no matter how many new vrfs will register on the
+ * table 100, the shared_tables will not increase (considering only
+ * table 100).
+ *
+ * at t4, vrf4 is bound to table 101, and shared_tables = 2.
+ *
+ * Looking at the value of shared_tables we can immediately know if
+ * the strict_mode can or cannot be enforced. Indeed, strict_mode
+ * can be enforced iff shared_tables = 0.
+ *
+ * Conversely, shared_tables is decreased when a vrf is de-associated
+ * from a table with exactly two associated vrfs.
+ */
+ u32 shared_tables;
+
+ bool strict_mode;
+};
+
+struct vrf_map_elem {
+ struct hlist_node hnode;
+ struct list_head vrf_list; /* VRFs registered to this table */
+
+ u32 table_id;
+ int users;
+ int ifindex;
+};
+
static unsigned int vrf_net_id;
+/* per netns vrf data */
+struct netns_vrf {
+ /* protected by rtnl lock */
+ bool add_fib_rules;
+
+ struct vrf_map vmap;
+ struct ctl_table_header *ctl_hdr;
+};
+
struct net_vrf {
struct rtable __rcu *rth;
struct rt6_info __rcu *rt6;
@@ -52,59 +118,223 @@ struct net_vrf {
struct fib6_table *fib6_table;
#endif
u32 tb_id;
-};
-struct pcpu_dstats {
- u64 tx_pkts;
- u64 tx_bytes;
- u64 tx_drps;
- u64 rx_pkts;
- u64 rx_bytes;
- u64 rx_drps;
- struct u64_stats_sync syncp;
+ struct list_head me_list; /* entry in vrf_map_elem */
+ int ifindex;
};
-static void vrf_rx_stats(struct net_device *dev, int len)
+static void vrf_tx_error(struct net_device *vrf_dev, struct sk_buff *skb)
{
- struct pcpu_dstats *dstats = this_cpu_ptr(dev->dstats);
+ vrf_dev->stats.tx_errors++;
+ kfree_skb(skb);
+}
+
+static struct vrf_map *netns_vrf_map(struct net *net)
+{
+ struct netns_vrf *nn_vrf = net_generic(net, vrf_net_id);
- u64_stats_update_begin(&dstats->syncp);
- dstats->rx_pkts++;
- dstats->rx_bytes += len;
- u64_stats_update_end(&dstats->syncp);
+ return &nn_vrf->vmap;
}
-static void vrf_tx_error(struct net_device *vrf_dev, struct sk_buff *skb)
+static struct vrf_map *netns_vrf_map_by_dev(struct net_device *dev)
{
- vrf_dev->stats.tx_errors++;
- kfree_skb(skb);
+ return netns_vrf_map(dev_net(dev));
+}
+
+static int vrf_map_elem_get_vrf_ifindex(struct vrf_map_elem *me)
+{
+ struct list_head *me_head = &me->vrf_list;
+ struct net_vrf *vrf;
+
+ if (list_empty(me_head))
+ return -ENODEV;
+
+ vrf = list_first_entry(me_head, struct net_vrf, me_list);
+
+ return vrf->ifindex;
+}
+
+static struct vrf_map_elem *vrf_map_elem_alloc(gfp_t flags)
+{
+ struct vrf_map_elem *me;
+
+ me = kmalloc(sizeof(*me), flags);
+ if (!me)
+ return NULL;
+
+ return me;
+}
+
+static void vrf_map_elem_free(struct vrf_map_elem *me)
+{
+ kfree(me);
+}
+
+static void vrf_map_elem_init(struct vrf_map_elem *me, int table_id,
+ int ifindex, int users)
+{
+ me->table_id = table_id;
+ me->ifindex = ifindex;
+ me->users = users;
+ INIT_LIST_HEAD(&me->vrf_list);
+}
+
+static struct vrf_map_elem *vrf_map_lookup_elem(struct vrf_map *vmap,
+ u32 table_id)
+{
+ struct vrf_map_elem *me;
+ u32 key;
+
+ key = jhash_1word(table_id, HASH_INITVAL);
+ hash_for_each_possible(vmap->ht, me, hnode, key) {
+ if (me->table_id == table_id)
+ return me;
+ }
+
+ return NULL;
+}
+
+static void vrf_map_add_elem(struct vrf_map *vmap, struct vrf_map_elem *me)
+{
+ u32 table_id = me->table_id;
+ u32 key;
+
+ key = jhash_1word(table_id, HASH_INITVAL);
+ hash_add(vmap->ht, &me->hnode, key);
+}
+
+static void vrf_map_del_elem(struct vrf_map_elem *me)
+{
+ hash_del(&me->hnode);
+}
+
+static void vrf_map_lock(struct vrf_map *vmap) __acquires(&vmap->vmap_lock)
+{
+ spin_lock(&vmap->vmap_lock);
+}
+
+static void vrf_map_unlock(struct vrf_map *vmap) __releases(&vmap->vmap_lock)
+{
+ spin_unlock(&vmap->vmap_lock);
}
-static void vrf_get_stats64(struct net_device *dev,
- struct rtnl_link_stats64 *stats)
-{
- int i;
-
- for_each_possible_cpu(i) {
- const struct pcpu_dstats *dstats;
- u64 tbytes, tpkts, tdrops, rbytes, rpkts;
- unsigned int start;
-
- dstats = per_cpu_ptr(dev->dstats, i);
- do {
- start = u64_stats_fetch_begin_irq(&dstats->syncp);
- tbytes = dstats->tx_bytes;
- tpkts = dstats->tx_pkts;
- tdrops = dstats->tx_drps;
- rbytes = dstats->rx_bytes;
- rpkts = dstats->rx_pkts;
- } while (u64_stats_fetch_retry_irq(&dstats->syncp, start));
- stats->tx_bytes += tbytes;
- stats->tx_packets += tpkts;
- stats->tx_dropped += tdrops;
- stats->rx_bytes += rbytes;
- stats->rx_packets += rpkts;
+/* called with rtnl lock held */
+static int
+vrf_map_register_dev(struct net_device *dev, struct netlink_ext_ack *extack)
+{
+ struct vrf_map *vmap = netns_vrf_map_by_dev(dev);
+ struct net_vrf *vrf = netdev_priv(dev);
+ struct vrf_map_elem *new_me, *me;
+ u32 table_id = vrf->tb_id;
+ bool free_new_me = false;
+ int users;
+ int res;
+
+ /* we pre-allocate elements used in the spin-locked section (so that we
+ * keep the spinlock as short as possible).
+ */
+ new_me = vrf_map_elem_alloc(GFP_KERNEL);
+ if (!new_me)
+ return -ENOMEM;
+
+ vrf_map_elem_init(new_me, table_id, dev->ifindex, 0);
+
+ vrf_map_lock(vmap);
+
+ me = vrf_map_lookup_elem(vmap, table_id);
+ if (!me) {
+ me = new_me;
+ vrf_map_add_elem(vmap, me);
+ goto link_vrf;
}
+
+ /* we already have an entry in the vrf_map, so it means there is (at
+ * least) a vrf registered on the specific table.
+ */
+ free_new_me = true;
+ if (vmap->strict_mode) {
+ /* vrfs cannot share the same table */
+ NL_SET_ERR_MSG(extack, "Table is used by another VRF");
+ res = -EBUSY;
+ goto unlock;
+ }
+
+link_vrf:
+ users = ++me->users;
+ if (users == 2)
+ ++vmap->shared_tables;
+
+ list_add(&vrf->me_list, &me->vrf_list);
+
+ res = 0;
+
+unlock:
+ vrf_map_unlock(vmap);
+
+ /* clean-up, if needed */
+ if (free_new_me)
+ vrf_map_elem_free(new_me);
+
+ return res;
+}
+
+/* called with rtnl lock held */
+static void vrf_map_unregister_dev(struct net_device *dev)
+{
+ struct vrf_map *vmap = netns_vrf_map_by_dev(dev);
+ struct net_vrf *vrf = netdev_priv(dev);
+ u32 table_id = vrf->tb_id;
+ struct vrf_map_elem *me;
+ int users;
+
+ vrf_map_lock(vmap);
+
+ me = vrf_map_lookup_elem(vmap, table_id);
+ if (!me)
+ goto unlock;
+
+ list_del(&vrf->me_list);
+
+ users = --me->users;
+ if (users == 1) {
+ --vmap->shared_tables;
+ } else if (users == 0) {
+ vrf_map_del_elem(me);
+
+ /* no one will refer to this element anymore */
+ vrf_map_elem_free(me);
+ }
+
+unlock:
+ vrf_map_unlock(vmap);
+}
+
+/* return the vrf device index associated with the table_id */
+static int vrf_ifindex_lookup_by_table_id(struct net *net, u32 table_id)
+{
+ struct vrf_map *vmap = netns_vrf_map(net);
+ struct vrf_map_elem *me;
+ int ifindex;
+
+ vrf_map_lock(vmap);
+
+ if (!vmap->strict_mode) {
+ ifindex = -EPERM;
+ goto unlock;
+ }
+
+ me = vrf_map_lookup_elem(vmap, table_id);
+ if (!me) {
+ ifindex = -ENODEV;
+ goto unlock;
+ }
+
+ ifindex = vrf_map_elem_get_vrf_ifindex(me);
+
+unlock:
+ vrf_map_unlock(vmap);
+
+ return ifindex;
}
/* by default VRF devices do not have a qdisc and are expected
@@ -113,15 +343,13 @@ static void vrf_get_stats64(struct net_device *dev,
static bool qdisc_tx_is_default(const struct net_device *dev)
{
struct netdev_queue *txq;
- struct Qdisc *qdisc;
if (dev->num_tx_queues > 1)
return false;
txq = netdev_get_tx_queue(dev, 0);
- qdisc = rcu_access_pointer(txq->qdisc);
- return !qdisc->enqueue;
+ return qdisc_txq_has_no_queue(txq);
}
/* Local traffic destined to local address. Reinsert the packet to rx
@@ -130,7 +358,7 @@ static bool qdisc_tx_is_default(const struct net_device *dev)
static int vrf_local_xmit(struct sk_buff *skb, struct net_device *dev,
struct dst_entry *dst)
{
- int len = skb->len;
+ unsigned int len = skb->len;
skb_orphan(skb);
@@ -143,20 +371,34 @@ static int vrf_local_xmit(struct sk_buff *skb, struct net_device *dev,
skb->protocol = eth_type_trans(skb, dev);
- if (likely(netif_rx(skb) == NET_RX_SUCCESS))
- vrf_rx_stats(dev, len);
+ if (likely(__netif_rx(skb) == NET_RX_SUCCESS))
+ dev_dstats_rx_add(dev, len);
else
- this_cpu_inc(dev->dstats->rx_drps);
+ dev_dstats_rx_dropped(dev);
return NETDEV_TX_OK;
}
+static void vrf_nf_set_untracked(struct sk_buff *skb)
+{
+ if (skb_get_nfct(skb) == 0)
+ nf_ct_set(skb, NULL, IP_CT_UNTRACKED);
+}
+
+static void vrf_nf_reset_ct(struct sk_buff *skb)
+{
+ if (skb_get_nfct(skb) == IP_CT_UNTRACKED)
+ nf_reset_ct(skb);
+}
+
#if IS_ENABLED(CONFIG_IPV6)
static int vrf_ip6_local_out(struct net *net, struct sock *sk,
struct sk_buff *skb)
{
int err;
+ vrf_nf_reset_ct(skb);
+
err = nf_hook(NFPROTO_IPV6, NF_INET_LOCAL_OUT, net,
sk, skb, NULL, skb_dst(skb)->dev, dst_output);
@@ -169,32 +411,36 @@ static int vrf_ip6_local_out(struct net *net, struct sock *sk,
static netdev_tx_t vrf_process_v6_outbound(struct sk_buff *skb,
struct net_device *dev)
{
- const struct ipv6hdr *iph = ipv6_hdr(skb);
+ const struct ipv6hdr *iph;
struct net *net = dev_net(skb->dev);
- struct flowi6 fl6 = {
- /* needed to match OIF rule */
- .flowi6_oif = dev->ifindex,
- .flowi6_iif = LOOPBACK_IFINDEX,
- .daddr = iph->daddr,
- .saddr = iph->saddr,
- .flowlabel = ip6_flowinfo(iph),
- .flowi6_mark = skb->mark,
- .flowi6_proto = iph->nexthdr,
- .flowi6_flags = FLOWI_FLAG_SKIP_NH_OIF,
- };
+ struct flowi6 fl6;
int ret = NET_XMIT_DROP;
struct dst_entry *dst;
struct dst_entry *dst_null = &net->ipv6.ip6_null_entry->dst;
- dst = ip6_route_output(net, NULL, &fl6);
- if (dst == dst_null)
+ if (!pskb_may_pull(skb, ETH_HLEN + sizeof(struct ipv6hdr)))
+ goto err;
+
+ iph = ipv6_hdr(skb);
+
+ memset(&fl6, 0, sizeof(fl6));
+ /* needed to match OIF rule */
+ fl6.flowi6_l3mdev = dev->ifindex;
+ fl6.flowi6_iif = LOOPBACK_IFINDEX;
+ fl6.daddr = iph->daddr;
+ fl6.saddr = iph->saddr;
+ fl6.flowlabel = ip6_flowinfo(iph);
+ fl6.flowi6_mark = skb->mark;
+ fl6.flowi6_proto = iph->nexthdr;
+
+ dst = ip6_dst_lookup_flow(net, NULL, &fl6, NULL);
+ if (IS_ERR(dst) || dst == dst_null)
goto err;
skb_dst_drop(skb);
- /* if dst.dev is loopback or the VRF device again this is locally
- * originated traffic destined to a local address. Short circuit
- * to Rx path
+ /* if dst.dev is the VRF device again this is locally originated traffic
+ * destined to a local address. Short circuit to Rx path.
*/
if (dst->dev == dev)
return vrf_local_xmit(skb, dev, dst);
@@ -204,6 +450,7 @@ static netdev_tx_t vrf_process_v6_outbound(struct sk_buff *skb,
/* strip the ethernet header added for pass through VRF device */
__skb_pull(skb, skb_network_offset(skb));
+ memset(IP6CB(skb), 0, sizeof(*IP6CB(skb)));
ret = vrf_ip6_local_out(net, skb->sk, skb);
if (unlikely(net_xmit_eval(ret)))
dev->stats.tx_errors++;
@@ -230,6 +477,8 @@ static int vrf_ip_local_out(struct net *net, struct sock *sk,
{
int err;
+ vrf_nf_reset_ct(skb);
+
err = nf_hook(NFPROTO_IPV4, NF_INET_LOCAL_OUT, net, sk,
skb, NULL, skb_dst(skb)->dev, dst_output);
if (likely(err == 1))
@@ -241,30 +490,35 @@ static int vrf_ip_local_out(struct net *net, struct sock *sk,
static netdev_tx_t vrf_process_v4_outbound(struct sk_buff *skb,
struct net_device *vrf_dev)
{
- struct iphdr *ip4h = ip_hdr(skb);
+ struct iphdr *ip4h;
int ret = NET_XMIT_DROP;
- struct flowi4 fl4 = {
- /* needed to match OIF rule */
- .flowi4_oif = vrf_dev->ifindex,
- .flowi4_iif = LOOPBACK_IFINDEX,
- .flowi4_tos = RT_TOS(ip4h->tos),
- .flowi4_flags = FLOWI_FLAG_ANYSRC | FLOWI_FLAG_SKIP_NH_OIF,
- .flowi4_proto = ip4h->protocol,
- .daddr = ip4h->daddr,
- .saddr = ip4h->saddr,
- };
+ struct flowi4 fl4;
struct net *net = dev_net(vrf_dev);
struct rtable *rt;
+ if (!pskb_may_pull(skb, ETH_HLEN + sizeof(struct iphdr)))
+ goto err;
+
+ ip4h = ip_hdr(skb);
+
+ memset(&fl4, 0, sizeof(fl4));
+ /* needed to match OIF rule */
+ fl4.flowi4_l3mdev = vrf_dev->ifindex;
+ fl4.flowi4_iif = LOOPBACK_IFINDEX;
+ fl4.flowi4_dscp = ip4h_dscp(ip4h);
+ fl4.flowi4_flags = FLOWI_FLAG_ANYSRC;
+ fl4.flowi4_proto = ip4h->protocol;
+ fl4.daddr = ip4h->daddr;
+ fl4.saddr = ip4h->saddr;
+
rt = ip_route_output_flow(net, &fl4, NULL);
if (IS_ERR(rt))
goto err;
skb_dst_drop(skb);
- /* if dst.dev is loopback or the VRF device again this is locally
- * originated traffic destined to a local address. Short circuit
- * to Rx path
+ /* if dst.dev is the VRF device again this is locally originated traffic
+ * destined to a local address. Short circuit to Rx path.
*/
if (rt->dst.dev == vrf_dev)
return vrf_local_xmit(skb, vrf_dev, &rt->dst);
@@ -279,6 +533,7 @@ static netdev_tx_t vrf_process_v4_outbound(struct sk_buff *skb,
RT_SCOPE_LINK);
}
+ memset(IPCB(skb), 0, sizeof(*IPCB(skb)));
ret = vrf_ip_local_out(dev_net(skb_dst(skb)->dev), skb->sk, skb);
if (unlikely(net_xmit_eval(ret)))
vrf_dev->stats.tx_errors++;
@@ -307,25 +562,19 @@ static netdev_tx_t is_ip_tx_frame(struct sk_buff *skb, struct net_device *dev)
static netdev_tx_t vrf_xmit(struct sk_buff *skb, struct net_device *dev)
{
- int len = skb->len;
- netdev_tx_t ret = is_ip_tx_frame(skb, dev);
-
- if (likely(ret == NET_XMIT_SUCCESS || ret == NET_XMIT_CN)) {
- struct pcpu_dstats *dstats = this_cpu_ptr(dev->dstats);
+ unsigned int len = skb->len;
+ netdev_tx_t ret;
- u64_stats_update_begin(&dstats->syncp);
- dstats->tx_pkts++;
- dstats->tx_bytes += len;
- u64_stats_update_end(&dstats->syncp);
- } else {
- this_cpu_inc(dev->dstats->tx_drps);
- }
+ ret = is_ip_tx_frame(skb, dev);
+ if (likely(ret == NET_XMIT_SUCCESS || ret == NET_XMIT_CN))
+ dev_dstats_tx_add(dev, len);
+ else
+ dev_dstats_tx_dropped(dev);
return ret;
}
-static int vrf_finish_direct(struct net *net, struct sock *sk,
- struct sk_buff *skb)
+static void vrf_finish_direct(struct sk_buff *skb)
{
struct net_device *vrf_dev = skb->dev;
@@ -344,7 +593,7 @@ static int vrf_finish_direct(struct net *net, struct sock *sk,
skb_pull(skb, ETH_HLEN);
}
- return 1;
+ vrf_nf_reset_ct(skb);
}
#if IS_ENABLED(CONFIG_IPV6)
@@ -354,27 +603,27 @@ static int vrf_finish_output6(struct net *net, struct sock *sk,
{
struct dst_entry *dst = skb_dst(skb);
struct net_device *dev = dst->dev;
+ const struct in6_addr *nexthop;
struct neighbour *neigh;
- struct in6_addr *nexthop;
int ret;
- nf_reset(skb);
+ vrf_nf_reset_ct(skb);
skb->protocol = htons(ETH_P_IPV6);
skb->dev = dev;
- rcu_read_lock_bh();
- nexthop = rt6_nexthop((struct rt6_info *)dst, &ipv6_hdr(skb)->daddr);
+ rcu_read_lock();
+ nexthop = rt6_nexthop(dst_rt6_info(dst), &ipv6_hdr(skb)->daddr);
neigh = __ipv6_neigh_lookup_noref(dst->dev, nexthop);
if (unlikely(!neigh))
neigh = __neigh_create(&nd_tbl, nexthop, dst->dev, false);
if (!IS_ERR(neigh)) {
sock_confirm_neigh(skb, neigh);
- ret = neigh_output(neigh, skb);
- rcu_read_unlock_bh();
+ ret = neigh_output(neigh, skb, false);
+ rcu_read_unlock();
return ret;
}
- rcu_read_unlock_bh();
+ rcu_read_unlock();
IP6_INC_STATS(dev_net(dst->dev),
ip6_dst_idev(dst), IPSTATS_MIB_OUTNOROUTES);
@@ -423,15 +672,41 @@ static struct sk_buff *vrf_ip6_out_redirect(struct net_device *vrf_dev,
return skb;
}
+static int vrf_output6_direct_finish(struct net *net, struct sock *sk,
+ struct sk_buff *skb)
+{
+ vrf_finish_direct(skb);
+
+ return vrf_ip6_local_out(net, sk, skb);
+}
+
static int vrf_output6_direct(struct net *net, struct sock *sk,
struct sk_buff *skb)
{
+ int err = 1;
+
skb->protocol = htons(ETH_P_IPV6);
- return NF_HOOK_COND(NFPROTO_IPV6, NF_INET_POST_ROUTING,
- net, sk, skb, NULL, skb->dev,
- vrf_finish_direct,
- !(IPCB(skb)->flags & IPSKB_REROUTED));
+ if (!(IPCB(skb)->flags & IPSKB_REROUTED))
+ err = nf_hook(NFPROTO_IPV6, NF_INET_POST_ROUTING, net, sk, skb,
+ NULL, skb->dev, vrf_output6_direct_finish);
+
+ if (likely(err == 1))
+ vrf_finish_direct(skb);
+
+ return err;
+}
+
+static int vrf_ip6_out_direct_finish(struct net *net, struct sock *sk,
+ struct sk_buff *skb)
+{
+ int err;
+
+ err = vrf_output6_direct(net, sk, skb);
+ if (likely(err == 1))
+ err = vrf_ip6_local_out(net, sk, skb);
+
+ return err;
}
static struct sk_buff *vrf_ip6_out_direct(struct net_device *vrf_dev,
@@ -444,18 +719,15 @@ static struct sk_buff *vrf_ip6_out_direct(struct net_device *vrf_dev,
skb->dev = vrf_dev;
err = nf_hook(NFPROTO_IPV6, NF_INET_LOCAL_OUT, net, sk,
- skb, NULL, vrf_dev, vrf_output6_direct);
+ skb, NULL, vrf_dev, vrf_ip6_out_direct_finish);
if (likely(err == 1))
err = vrf_output6_direct(net, sk, skb);
- /* reset skb device */
if (likely(err == 1))
- nf_reset(skb);
- else
- skb = NULL;
+ return skb;
- return skb;
+ return NULL;
}
static struct sk_buff *vrf_ip6_out(struct net_device *vrf_dev,
@@ -466,7 +738,10 @@ static struct sk_buff *vrf_ip6_out(struct net_device *vrf_dev,
if (rt6_need_strict(&ipv6_hdr(skb)->daddr))
return skb;
- if (qdisc_tx_is_default(vrf_dev))
+ vrf_nf_set_untracked(skb);
+
+ if (qdisc_tx_is_default(vrf_dev) ||
+ IP6CB(skb)->flags & IP6SKB_XFRM_TRANSFORMED)
return vrf_ip6_out_direct(vrf_dev, sk, skb);
return vrf_ip6_out_redirect(vrf_dev, skb);
@@ -487,16 +762,16 @@ static void vrf_rt6_release(struct net_device *dev, struct net_vrf *vrf)
*/
if (rt6) {
dst = &rt6->dst;
- dev_put(dst->dev);
+ netdev_ref_replace(dst->dev, net->loopback_dev,
+ &dst->dev_tracker, GFP_KERNEL);
dst->dev = net->loopback_dev;
- dev_hold(dst->dev);
dst_release(dst);
}
}
static int vrf_rt6_create(struct net_device *dev)
{
- int flags = DST_HOST | DST_NOPOLICY | DST_NOXFRM;
+ int flags = DST_NOPOLICY | DST_NOXFRM;
struct net_vrf *vrf = netdev_priv(dev);
struct net *net = dev_net(dev);
struct rt6_info *rt6;
@@ -545,48 +820,39 @@ static int vrf_rt6_create(struct net_device *dev)
static int vrf_finish_output(struct net *net, struct sock *sk, struct sk_buff *skb)
{
struct dst_entry *dst = skb_dst(skb);
- struct rtable *rt = (struct rtable *)dst;
+ struct rtable *rt = dst_rtable(dst);
struct net_device *dev = dst->dev;
unsigned int hh_len = LL_RESERVED_SPACE(dev);
struct neighbour *neigh;
- u32 nexthop;
- int ret = -EINVAL;
+ bool is_v6gw = false;
- nf_reset(skb);
+ vrf_nf_reset_ct(skb);
/* Be paranoid, rather than too clever. */
if (unlikely(skb_headroom(skb) < hh_len && dev->header_ops)) {
- struct sk_buff *skb2;
-
- skb2 = skb_realloc_headroom(skb, LL_RESERVED_SPACE(dev));
- if (!skb2) {
- ret = -ENOMEM;
- goto err;
+ skb = skb_expand_head(skb, hh_len);
+ if (!skb) {
+ dev->stats.tx_errors++;
+ return -ENOMEM;
}
- if (skb->sk)
- skb_set_owner_w(skb2, skb->sk);
-
- consume_skb(skb);
- skb = skb2;
}
- rcu_read_lock_bh();
+ rcu_read_lock();
- nexthop = (__force u32)rt_nexthop(rt, ip_hdr(skb)->daddr);
- neigh = __ipv4_neigh_lookup_noref(dev, nexthop);
- if (unlikely(!neigh))
- neigh = __neigh_create(&arp_tbl, &nexthop, dev, false);
+ neigh = ip_neigh_for_gw(rt, skb, &is_v6gw);
if (!IS_ERR(neigh)) {
+ int ret;
+
sock_confirm_neigh(skb, neigh);
- ret = neigh_output(neigh, skb);
- rcu_read_unlock_bh();
+ /* if crossing protocols, can not use the cached header */
+ ret = neigh_output(neigh, skb, is_v6gw);
+ rcu_read_unlock();
return ret;
}
- rcu_read_unlock_bh();
-err:
+ rcu_read_unlock();
vrf_tx_error(skb->dev, skb);
- return ret;
+ return -EINVAL;
}
static int vrf_output(struct net *net, struct sock *sk, struct sk_buff *skb)
@@ -636,15 +902,41 @@ static struct sk_buff *vrf_ip_out_redirect(struct net_device *vrf_dev,
return skb;
}
+static int vrf_output_direct_finish(struct net *net, struct sock *sk,
+ struct sk_buff *skb)
+{
+ vrf_finish_direct(skb);
+
+ return vrf_ip_local_out(net, sk, skb);
+}
+
static int vrf_output_direct(struct net *net, struct sock *sk,
struct sk_buff *skb)
{
+ int err = 1;
+
skb->protocol = htons(ETH_P_IP);
- return NF_HOOK_COND(NFPROTO_IPV4, NF_INET_POST_ROUTING,
- net, sk, skb, NULL, skb->dev,
- vrf_finish_direct,
- !(IPCB(skb)->flags & IPSKB_REROUTED));
+ if (!(IPCB(skb)->flags & IPSKB_REROUTED))
+ err = nf_hook(NFPROTO_IPV4, NF_INET_POST_ROUTING, net, sk, skb,
+ NULL, skb->dev, vrf_output_direct_finish);
+
+ if (likely(err == 1))
+ vrf_finish_direct(skb);
+
+ return err;
+}
+
+static int vrf_ip_out_direct_finish(struct net *net, struct sock *sk,
+ struct sk_buff *skb)
+{
+ int err;
+
+ err = vrf_output_direct(net, sk, skb);
+ if (likely(err == 1))
+ err = vrf_ip_local_out(net, sk, skb);
+
+ return err;
}
static struct sk_buff *vrf_ip_out_direct(struct net_device *vrf_dev,
@@ -657,18 +949,15 @@ static struct sk_buff *vrf_ip_out_direct(struct net_device *vrf_dev,
skb->dev = vrf_dev;
err = nf_hook(NFPROTO_IPV4, NF_INET_LOCAL_OUT, net, sk,
- skb, NULL, vrf_dev, vrf_output_direct);
+ skb, NULL, vrf_dev, vrf_ip_out_direct_finish);
if (likely(err == 1))
err = vrf_output_direct(net, sk, skb);
- /* reset skb device */
if (likely(err == 1))
- nf_reset(skb);
- else
- skb = NULL;
+ return skb;
- return skb;
+ return NULL;
}
static struct sk_buff *vrf_ip_out(struct net_device *vrf_dev,
@@ -680,7 +969,10 @@ static struct sk_buff *vrf_ip_out(struct net_device *vrf_dev,
ipv4_is_lbcast(ip_hdr(skb)->daddr))
return skb;
- if (qdisc_tx_is_default(vrf_dev))
+ vrf_nf_set_untracked(skb);
+
+ if (qdisc_tx_is_default(vrf_dev) ||
+ IPCB(skb)->flags & IPSKB_XFRM_TRANSFORMED)
return vrf_ip_out_direct(vrf_dev, sk, skb);
return vrf_ip_out_redirect(vrf_dev, skb);
@@ -717,9 +1009,9 @@ static void vrf_rtable_release(struct net_device *dev, struct net_vrf *vrf)
*/
if (rth) {
dst = &rth->dst;
- dev_put(dst->dev);
+ netdev_ref_replace(dst->dev, net->loopback_dev,
+ &dst->dev_tracker, GFP_KERNEL);
dst->dev = net->loopback_dev;
- dev_hold(dst->dev);
dst_release(dst);
}
}
@@ -733,7 +1025,7 @@ static int vrf_rtable_create(struct net_device *dev)
return -ENOMEM;
/* create a dst for routing packets out through a VRF device */
- rth = rt_dst_alloc(dev, 0, RTN_UNICAST, 1, 1, 0);
+ rth = rt_dst_alloc(dev, 0, RTN_UNICAST, 1);
if (!rth)
return -ENOMEM;
@@ -832,31 +1124,21 @@ static void vrf_dev_uninit(struct net_device *dev)
vrf_rtable_release(dev, vrf);
vrf_rt6_release(dev, vrf);
-
- free_percpu(dev->dstats);
- dev->dstats = NULL;
}
static int vrf_dev_init(struct net_device *dev)
{
struct net_vrf *vrf = netdev_priv(dev);
- dev->dstats = netdev_alloc_pcpu_stats(struct pcpu_dstats);
- if (!dev->dstats)
- goto out_nomem;
-
/* create the default dst which points back to us */
if (vrf_rtable_create(dev) != 0)
- goto out_stats;
+ goto out_nomem;
if (vrf_rt6_create(dev) != 0)
goto out_rth;
dev->flags = IFF_MASTER | IFF_NOARP;
- /* MTU is irrelevant for VRF device; set to 64k similar to lo */
- dev->mtu = 64 * 1024;
-
/* similarly, oper state is irrelevant; set to up to avoid confusion */
dev->operstate = IF_OPER_UP;
netdev_lockdep_set_classes(dev);
@@ -864,9 +1146,6 @@ static int vrf_dev_init(struct net_device *dev)
out_rth:
vrf_rtable_release(dev, vrf);
-out_stats:
- free_percpu(dev->dstats);
- dev->dstats = NULL;
out_nomem:
return -ENOMEM;
}
@@ -875,7 +1154,7 @@ static const struct net_device_ops vrf_netdev_ops = {
.ndo_init = vrf_dev_init,
.ndo_uninit = vrf_dev_uninit,
.ndo_start_xmit = vrf_xmit,
- .ndo_get_stats64 = vrf_get_stats64,
+ .ndo_set_mac_address = eth_mac_addr,
.ndo_add_slave = vrf_add_slave,
.ndo_del_slave = vrf_del_slave,
};
@@ -905,6 +1184,62 @@ static struct sk_buff *vrf_rcv_nfhook(u8 pf, unsigned int hook,
return skb;
}
+static int vrf_prepare_mac_header(struct sk_buff *skb,
+ struct net_device *vrf_dev, u16 proto)
+{
+ struct ethhdr *eth;
+ int err;
+
+ /* in general, we do not know if there is enough space in the head of
+ * the packet for hosting the mac header.
+ */
+ err = skb_cow_head(skb, LL_RESERVED_SPACE(vrf_dev));
+ if (unlikely(err))
+ /* no space in the skb head */
+ return -ENOBUFS;
+
+ __skb_push(skb, ETH_HLEN);
+ eth = (struct ethhdr *)skb->data;
+
+ skb_reset_mac_header(skb);
+ skb_reset_mac_len(skb);
+
+ /* we set the ethernet destination and the source addresses to the
+ * address of the VRF device.
+ */
+ ether_addr_copy(eth->h_dest, vrf_dev->dev_addr);
+ ether_addr_copy(eth->h_source, vrf_dev->dev_addr);
+ eth->h_proto = htons(proto);
+
+ /* the destination address of the Ethernet frame corresponds to the
+ * address set on the VRF interface; therefore, the packet is intended
+ * to be processed locally.
+ */
+ skb->protocol = eth->h_proto;
+ skb->pkt_type = PACKET_HOST;
+
+ skb_postpush_rcsum(skb, skb->data, ETH_HLEN);
+
+ skb_pull_inline(skb, ETH_HLEN);
+
+ return 0;
+}
+
+/* prepare and add the mac header to the packet if it was not set previously.
+ * In this way, packet sniffers such as tcpdump can parse the packet correctly.
+ * If the mac header was already set, the original mac header is left
+ * untouched and the function returns immediately.
+ */
+static int vrf_add_mac_header_if_unset(struct sk_buff *skb,
+ struct net_device *vrf_dev,
+ u16 proto, struct net_device *orig_dev)
+{
+ if (skb_mac_header_was_set(skb) && dev_has_header(orig_dev))
+ return 0;
+
+ return vrf_prepare_mac_header(skb, vrf_dev, proto);
+}
+
#if IS_ENABLED(CONFIG_IPV6)
/* neighbor handling is done with actual device; do not want
* to flip skb->dev for those ndisc packets. This really fails
@@ -967,6 +1302,8 @@ static void vrf_ip6_input_dst(struct sk_buff *skb, struct net_device *vrf_dev,
struct net *net = dev_net(vrf_dev);
struct rt6_info *rt6;
+ skb_dst_drop(skb);
+
rt6 = vrf_ip6_route_lookup(net, vrf_dev, &fl6, ifindex, skb,
RT6_LOOKUP_F_HAS_SADDR | RT6_LOOKUP_F_IFACE);
if (unlikely(!rt6))
@@ -986,27 +1323,42 @@ static struct sk_buff *vrf_ip6_rcv(struct net_device *vrf_dev,
bool is_ndisc = ipv6_ndisc_frame(skb);
/* loopback, multicast & non-ND link-local traffic; do not push through
- * packet taps again. Reset pkt_type for upper layers to process skb
+ * packet taps again. Reset pkt_type for upper layers to process skb.
+ * For non-loopback strict packets, determine the dst using the original
+ * ifindex.
*/
if (skb->pkt_type == PACKET_LOOPBACK || (need_strict && !is_ndisc)) {
skb->dev = vrf_dev;
skb->skb_iif = vrf_dev->ifindex;
IP6CB(skb)->flags |= IP6SKB_L3SLAVE;
+
if (skb->pkt_type == PACKET_LOOPBACK)
skb->pkt_type = PACKET_HOST;
+ else
+ vrf_ip6_input_dst(skb, vrf_dev, orig_iif);
+
goto out;
}
/* if packet is NDISC then keep the ingress interface */
if (!is_ndisc) {
- vrf_rx_stats(vrf_dev, skb->len);
+ struct net_device *orig_dev = skb->dev;
+
+ dev_dstats_rx_add(vrf_dev, skb->len);
skb->dev = vrf_dev;
skb->skb_iif = vrf_dev->ifindex;
if (!list_empty(&vrf_dev->ptype_all)) {
- skb_push(skb, skb->mac_len);
- dev_queue_xmit_nit(skb, vrf_dev);
- skb_pull(skb, skb->mac_len);
+ int err;
+
+ err = vrf_add_mac_header_if_unset(skb, vrf_dev,
+ ETH_P_IPV6,
+ orig_dev);
+ if (likely(!err)) {
+ skb_push(skb, skb->mac_len);
+ dev_queue_xmit_nit(skb, vrf_dev);
+ skb_pull(skb, skb->mac_len);
+ }
}
IP6CB(skb)->flags |= IP6SKB_L3SLAVE;
@@ -1031,6 +1383,8 @@ static struct sk_buff *vrf_ip6_rcv(struct net_device *vrf_dev,
static struct sk_buff *vrf_ip_rcv(struct net_device *vrf_dev,
struct sk_buff *skb)
{
+ struct net_device *orig_dev = skb->dev;
+
skb->dev = vrf_dev;
skb->skb_iif = vrf_dev->ifindex;
IPCB(skb)->flags |= IPSKB_L3SLAVE;
@@ -1046,12 +1400,18 @@ static struct sk_buff *vrf_ip_rcv(struct net_device *vrf_dev,
goto out;
}
- vrf_rx_stats(vrf_dev, skb->len);
+ dev_dstats_rx_add(vrf_dev, skb->len);
if (!list_empty(&vrf_dev->ptype_all)) {
- skb_push(skb, skb->mac_len);
- dev_queue_xmit_nit(skb, vrf_dev);
- skb_pull(skb, skb->mac_len);
+ int err;
+
+ err = vrf_add_mac_header_if_unset(skb, vrf_dev, ETH_P_IP,
+ orig_dev);
+ if (likely(!err)) {
+ skb_push(skb, skb->mac_len);
+ dev_queue_xmit_nit(skb, vrf_dev);
+ skb_pull(skb, skb->mac_len);
+ }
}
skb = vrf_rcv_nfhook(NFPROTO_IPV4, NF_INET_PRE_ROUTING, skb, vrf_dev);
@@ -1077,12 +1437,14 @@ static struct sk_buff *vrf_l3_rcv(struct net_device *vrf_dev,
#if IS_ENABLED(CONFIG_IPV6)
/* send to link-local or multicast address via interface enslaved to
* VRF device. Force lookup to VRF table without changing flow struct
+ * Note: Caller to this function must hold rcu_read_lock() and no refcnt
+ * is taken on the dst by this function.
*/
static struct dst_entry *vrf_link_scope_lookup(const struct net_device *dev,
struct flowi6 *fl6)
{
struct net *net = dev_net(dev);
- int flags = RT6_LOOKUP_F_IFACE;
+ int flags = RT6_LOOKUP_F_IFACE | RT6_LOOKUP_F_DST_NOREF;
struct dst_entry *dst = NULL;
struct rt6_info *rt;
@@ -1092,7 +1454,6 @@ static struct dst_entry *vrf_link_scope_lookup(const struct net_device *dev,
*/
if (fl6->flowi6_oif == dev->ifindex) {
dst = &net->ipv6.ip6_null_entry->dst;
- dst_hold(dst);
return dst;
}
@@ -1119,8 +1480,8 @@ static const struct l3mdev_ops vrf_l3mdev_ops = {
static void vrf_get_drvinfo(struct net_device *dev,
struct ethtool_drvinfo *info)
{
- strlcpy(info->driver, DRV_NAME, sizeof(info->driver));
- strlcpy(info->version, DRV_VERSION, sizeof(info->version));
+ strscpy(info->driver, DRV_NAME, sizeof(info->driver));
+ strscpy(info->version, DRV_VERSION, sizeof(info->version));
}
static const struct ethtool_ops vrf_ethtool_ops = {
@@ -1146,7 +1507,8 @@ static int vrf_fib_rule(const struct net_device *dev, __u8 family, bool add_it)
struct sk_buff *skb;
int err;
- if (family == AF_INET6 && !ipv6_mod_enabled())
+ if ((family == AF_INET6 || family == RTNL_FAMILY_IP6MR) &&
+ !ipv6_mod_enabled())
return 0;
skb = nlmsg_new(vrf_fib_rule_nl_size(), GFP_KERNEL);
@@ -1176,14 +1538,12 @@ static int vrf_fib_rule(const struct net_device *dev, __u8 family, bool add_it)
nlmsg_end(skb, nlh);
- /* fib_nl_{new,del}rule handling looks for net from skb->sk */
- skb->sk = dev_net(dev)->rtnl;
if (add_it) {
- err = fib_nl_newrule(skb, nlh, NULL);
+ err = fib_newrule(dev_net(dev), skb, nlh, NULL, true);
if (err == -EEXIST)
err = 0;
} else {
- err = fib_nl_delrule(skb, nlh, NULL);
+ err = fib_delrule(dev_net(dev), skb, nlh, NULL, true);
if (err == -ENOENT)
err = 0;
}
@@ -1255,10 +1615,10 @@ static void vrf_setup(struct net_device *dev)
eth_hw_addr_random(dev);
/* don't acquire vrf device's netif_tx_lock when transmitting */
- dev->features |= NETIF_F_LLTX;
+ dev->lltx = true;
/* don't allow vrf devices to change network namespaces. */
- dev->features |= NETIF_F_NETNS_LOCAL;
+ dev->netns_immutable = true;
/* does not make sense for a VLAN to be added to a vrf device */
dev->features |= NETIF_F_VLAN_CHALLENGED;
@@ -1273,6 +1633,18 @@ static void vrf_setup(struct net_device *dev)
/* default to no qdisc; user can add if desired */
dev->priv_flags |= IFF_NO_QUEUE;
+ dev->priv_flags |= IFF_NO_RX_HANDLER;
+ dev->priv_flags |= IFF_LIVE_ADDR_CHANGE;
+
+ /* VRF devices do not care about MTU, but if the MTU is set
+ * too low then the ipv4 and ipv6 protocols are disabled
+ * which breaks networking.
+ */
+ dev->min_mtu = IPV6_MIN_MTU;
+ dev->max_mtu = IP6_MAX_MTU;
+ dev->mtu = dev->max_mtu;
+
+ dev->pcpu_stat_type = NETDEV_PCPU_STAT_DSTATS;
}
static int vrf_validate(struct nlattr *tb[], struct nlattr *data[],
@@ -1299,14 +1671,18 @@ static void vrf_dellink(struct net_device *dev, struct list_head *head)
netdev_for_each_lower_dev(dev, port_dev, iter)
vrf_del_slave(dev, port_dev);
+ vrf_map_unregister_dev(dev);
+
unregister_netdevice_queue(dev, head);
}
-static int vrf_newlink(struct net *src_net, struct net_device *dev,
- struct nlattr *tb[], struct nlattr *data[],
+static int vrf_newlink(struct net_device *dev,
+ struct rtnl_newlink_params *params,
struct netlink_ext_ack *extack)
{
struct net_vrf *vrf = netdev_priv(dev);
+ struct nlattr **data = params->data;
+ struct netns_vrf *nn_vrf;
bool *add_fib_rules;
struct net *net;
int err;
@@ -1329,11 +1705,26 @@ static int vrf_newlink(struct net *src_net, struct net_device *dev,
if (err)
goto out;
+ /* mapping between table_id and vrf;
+ * note: such binding could not be done in the dev init function
+ * because dev->ifindex id is not available yet.
+ */
+ vrf->ifindex = dev->ifindex;
+
+ err = vrf_map_register_dev(dev, extack);
+ if (err) {
+ unregister_netdevice(dev);
+ goto out;
+ }
+
net = dev_net(dev);
- add_fib_rules = net_generic(net, vrf_net_id);
+ nn_vrf = net_generic(net, vrf_net_id);
+
+ add_fib_rules = &nn_vrf->add_fib_rules;
if (*add_fib_rules) {
err = vrf_add_fib_rules(dev);
if (err) {
+ vrf_map_unregister_dev(dev);
unregister_netdevice(dev);
goto out;
}
@@ -1420,20 +1811,164 @@ static struct notifier_block vrf_notifier_block __read_mostly = {
.notifier_call = vrf_device_event,
};
+static int vrf_map_init(struct vrf_map *vmap)
+{
+ spin_lock_init(&vmap->vmap_lock);
+ hash_init(vmap->ht);
+
+ vmap->strict_mode = false;
+
+ return 0;
+}
+
+#ifdef CONFIG_SYSCTL
+static bool vrf_strict_mode(struct vrf_map *vmap)
+{
+ bool strict_mode;
+
+ vrf_map_lock(vmap);
+ strict_mode = vmap->strict_mode;
+ vrf_map_unlock(vmap);
+
+ return strict_mode;
+}
+
+static int vrf_strict_mode_change(struct vrf_map *vmap, bool new_mode)
+{
+ bool *cur_mode;
+ int res = 0;
+
+ vrf_map_lock(vmap);
+
+ cur_mode = &vmap->strict_mode;
+ if (*cur_mode == new_mode)
+ goto unlock;
+
+ if (*cur_mode) {
+ /* disable strict mode */
+ *cur_mode = false;
+ } else {
+ if (vmap->shared_tables) {
+ /* we cannot allow strict_mode because there are some
+ * vrfs that share one or more tables.
+ */
+ res = -EBUSY;
+ goto unlock;
+ }
+
+ /* no tables are shared among vrfs, so we can go back
+ * to 1:1 association between a vrf with its table.
+ */
+ *cur_mode = true;
+ }
+
+unlock:
+ vrf_map_unlock(vmap);
+
+ return res;
+}
+
+static int vrf_shared_table_handler(const struct ctl_table *table, int write,
+ void *buffer, size_t *lenp, loff_t *ppos)
+{
+ struct net *net = (struct net *)table->extra1;
+ struct vrf_map *vmap = netns_vrf_map(net);
+ int proc_strict_mode = 0;
+ struct ctl_table tmp = {
+ .procname = table->procname,
+ .data = &proc_strict_mode,
+ .maxlen = sizeof(int),
+ .mode = table->mode,
+ .extra1 = SYSCTL_ZERO,
+ .extra2 = SYSCTL_ONE,
+ };
+ int ret;
+
+ if (!write)
+ proc_strict_mode = vrf_strict_mode(vmap);
+
+ ret = proc_dointvec_minmax(&tmp, write, buffer, lenp, ppos);
+
+ if (write && ret == 0)
+ ret = vrf_strict_mode_change(vmap, (bool)proc_strict_mode);
+
+ return ret;
+}
+
+static const struct ctl_table vrf_table[] = {
+ {
+ .procname = "strict_mode",
+ .data = NULL,
+ .maxlen = sizeof(int),
+ .mode = 0644,
+ .proc_handler = vrf_shared_table_handler,
+ /* set by the vrf_netns_init */
+ .extra1 = NULL,
+ },
+};
+
+static int vrf_netns_init_sysctl(struct net *net, struct netns_vrf *nn_vrf)
+{
+ struct ctl_table *table;
+
+ table = kmemdup(vrf_table, sizeof(vrf_table), GFP_KERNEL);
+ if (!table)
+ return -ENOMEM;
+
+ /* init the extra1 parameter with the reference to current netns */
+ table[0].extra1 = net;
+
+ nn_vrf->ctl_hdr = register_net_sysctl_sz(net, "net/vrf", table,
+ ARRAY_SIZE(vrf_table));
+ if (!nn_vrf->ctl_hdr) {
+ kfree(table);
+ return -ENOMEM;
+ }
+
+ return 0;
+}
+
+static void vrf_netns_exit_sysctl(struct net *net)
+{
+ struct netns_vrf *nn_vrf = net_generic(net, vrf_net_id);
+ const struct ctl_table *table;
+
+ table = nn_vrf->ctl_hdr->ctl_table_arg;
+ unregister_net_sysctl_table(nn_vrf->ctl_hdr);
+ kfree(table);
+}
+#else
+static int vrf_netns_init_sysctl(struct net *net, struct netns_vrf *nn_vrf)
+{
+ return 0;
+}
+
+static void vrf_netns_exit_sysctl(struct net *net)
+{
+}
+#endif
+
/* Initialize per network namespace state */
static int __net_init vrf_netns_init(struct net *net)
{
- bool *add_fib_rules = net_generic(net, vrf_net_id);
+ struct netns_vrf *nn_vrf = net_generic(net, vrf_net_id);
- *add_fib_rules = true;
+ nn_vrf->add_fib_rules = true;
+ vrf_map_init(&nn_vrf->vmap);
- return 0;
+ return vrf_netns_init_sysctl(net, nn_vrf);
+}
+
+static void __net_exit vrf_netns_exit(struct net *net)
+{
+ vrf_netns_exit_sysctl(net);
}
static struct pernet_operations vrf_net_ops __net_initdata = {
.init = vrf_netns_init,
+ .exit = vrf_netns_exit,
.id = &vrf_net_id,
- .size = sizeof(bool),
+ .size = sizeof(struct netns_vrf),
};
static int __init vrf_init_module(void)
@@ -1446,14 +1981,24 @@ static int __init vrf_init_module(void)
if (rc < 0)
goto error;
+ rc = l3mdev_table_lookup_register(L3MDEV_TYPE_VRF,
+ vrf_ifindex_lookup_by_table_id);
+ if (rc < 0)
+ goto unreg_pernet;
+
rc = rtnl_link_register(&vrf_link_ops);
- if (rc < 0) {
- unregister_pernet_subsys(&vrf_net_ops);
- goto error;
- }
+ if (rc < 0)
+ goto table_lookup_unreg;
return 0;
+table_lookup_unreg:
+ l3mdev_table_lookup_unregister(L3MDEV_TYPE_VRF,
+ vrf_ifindex_lookup_by_table_id);
+
+unreg_pernet:
+ unregister_pernet_subsys(&vrf_net_ops);
+
error:
unregister_netdevice_notifier(&vrf_notifier_block);
return rc;
generated by cgit (git 2.34.1) at 2025-12-24 20:46:53 +0000