mpls: Packet stats
Having MPLS packet stats is useful for observing network operation and
for diagnosing network problems. In the absence of anything better,
RFC2863 and RFC3813 are used for guidance for which stats to expose
and the semantics of them. In particular rx_noroutes maps to in
unknown protos in RFC2863. The stats are exposed to userspace via
AF_MPLS attributes embedded in the IFLA_STATS_AF_SPEC attribute of
RTM_GETSTATS messages.
All the introduced fields are 64-bit, even error ones, to ensure no
overflow with long uptimes. Per-CPU counters are used to avoid
cache-line contention on the commonly used fields. The other fields
have also been made per-CPU for code to avoid performance problems in
error conditions on the assumption that on some platforms the cost of
atomic operations could be more expensive than sending the packet
(which is what would be done in the success case). If that's not the
case, we could instead not use per-CPU counters for these fields.
Only unicast and non-fragment are exposed at the moment, but other
counters can be exposed in the future either by adding to the end of
struct mpls_link_stats or by additional netlink attributes in the
AF_MPLS IFLA_STATS_AF_SPEC nested attribute.
Signed-off-by: Robert Shearman <rshearma@brocade.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
diff --git a/net/mpls/af_mpls.c b/net/mpls/af_mpls.c
index 15fe976..4dc8196 100644
--- a/net/mpls/af_mpls.c
+++ b/net/mpls/af_mpls.c
@@ -8,6 +8,7 @@
#include <linux/ipv6.h>
#include <linux/mpls.h>
#include <linux/vmalloc.h>
+#include <linux/percpu.h>
#include <net/ip.h>
#include <net/dst.h>
#include <net/sock.h>
@@ -17,8 +18,8 @@
#include <net/netns/generic.h>
#if IS_ENABLED(CONFIG_IPV6)
#include <net/ipv6.h>
-#include <net/addrconf.h>
#endif
+#include <net/addrconf.h>
#include <net/nexthop.h>
#include "internal.h"
@@ -48,11 +49,6 @@
return rt;
}
-static inline struct mpls_dev *mpls_dev_get(const struct net_device *dev)
-{
- return rcu_dereference_rtnl(dev->mpls_ptr);
-}
-
bool mpls_output_possible(const struct net_device *dev)
{
return dev && (dev->flags & IFF_UP) && netif_carrier_ok(dev);
@@ -98,6 +94,31 @@
}
EXPORT_SYMBOL_GPL(mpls_pkt_too_big);
+void mpls_stats_inc_outucastpkts(struct net_device *dev,
+ const struct sk_buff *skb)
+{
+ struct mpls_dev *mdev;
+
+ if (skb->protocol == htons(ETH_P_MPLS_UC)) {
+ mdev = mpls_dev_get(dev);
+ if (mdev)
+ MPLS_INC_STATS_LEN(mdev, skb->len,
+ tx_packets,
+ tx_bytes);
+ } else if (skb->protocol == htons(ETH_P_IP)) {
+ IP_UPD_PO_STATS(dev_net(dev), IPSTATS_MIB_OUT, skb->len);
+#if IS_ENABLED(CONFIG_IPV6)
+ } else if (skb->protocol == htons(ETH_P_IPV6)) {
+ struct inet6_dev *in6dev = __in6_dev_get(dev);
+
+ if (in6dev)
+ IP6_UPD_PO_STATS(dev_net(dev), in6dev,
+ IPSTATS_MIB_OUT, skb->len);
+#endif
+ }
+}
+EXPORT_SYMBOL_GPL(mpls_stats_inc_outucastpkts);
+
static u32 mpls_multipath_hash(struct mpls_route *rt,
struct sk_buff *skb, bool bos)
{
@@ -253,6 +274,7 @@
struct mpls_nh *nh;
struct mpls_entry_decoded dec;
struct net_device *out_dev;
+ struct mpls_dev *out_mdev;
struct mpls_dev *mdev;
unsigned int hh_len;
unsigned int new_header_size;
@@ -262,17 +284,25 @@
/* Careful this entire function runs inside of an rcu critical section */
mdev = mpls_dev_get(dev);
- if (!mdev || !mdev->input_enabled)
+ if (!mdev)
goto drop;
+ MPLS_INC_STATS_LEN(mdev, skb->len, rx_packets,
+ rx_bytes);
+
+ if (!mdev->input_enabled) {
+ MPLS_INC_STATS(mdev, rx_dropped);
+ goto drop;
+ }
+
if (skb->pkt_type != PACKET_HOST)
- goto drop;
+ goto err;
if ((skb = skb_share_check(skb, GFP_ATOMIC)) == NULL)
- goto drop;
+ goto err;
if (!pskb_may_pull(skb, sizeof(*hdr)))
- goto drop;
+ goto err;
/* Read and decode the label */
hdr = mpls_hdr(skb);
@@ -285,33 +315,35 @@
skb_orphan(skb);
rt = mpls_route_input_rcu(net, dec.label);
- if (!rt)
+ if (!rt) {
+ MPLS_INC_STATS(mdev, rx_noroute);
goto drop;
+ }
nh = mpls_select_multipath(rt, skb, dec.bos);
if (!nh)
- goto drop;
-
- /* Find the output device */
- out_dev = rcu_dereference(nh->nh_dev);
- if (!mpls_output_possible(out_dev))
- goto drop;
+ goto err;
if (skb_warn_if_lro(skb))
- goto drop;
+ goto err;
skb_forward_csum(skb);
/* Verify ttl is valid */
if (dec.ttl <= 1)
- goto drop;
+ goto err;
dec.ttl -= 1;
+ /* Find the output device */
+ out_dev = rcu_dereference(nh->nh_dev);
+ if (!mpls_output_possible(out_dev))
+ goto tx_err;
+
/* Verify the destination can hold the packet */
new_header_size = mpls_nh_header_size(nh);
mtu = mpls_dev_mtu(out_dev);
if (mpls_pkt_too_big(skb, mtu - new_header_size))
- goto drop;
+ goto tx_err;
hh_len = LL_RESERVED_SPACE(out_dev);
if (!out_dev->header_ops)
@@ -319,7 +351,7 @@
/* Ensure there is enough space for the headers in the skb */
if (skb_cow(skb, hh_len + new_header_size))
- goto drop;
+ goto tx_err;
skb->dev = out_dev;
skb->protocol = htons(ETH_P_MPLS_UC);
@@ -327,7 +359,7 @@
if (unlikely(!new_header_size && dec.bos)) {
/* Penultimate hop popping */
if (!mpls_egress(rt, skb, dec))
- goto drop;
+ goto err;
} else {
bool bos;
int i;
@@ -343,6 +375,8 @@
}
}
+ mpls_stats_inc_outucastpkts(out_dev, skb);
+
/* If via wasn't specified then send out using device address */
if (nh->nh_via_table == MPLS_NEIGH_TABLE_UNSPEC)
err = neigh_xmit(NEIGH_LINK_TABLE, out_dev,
@@ -355,6 +389,13 @@
__func__, err);
return 0;
+tx_err:
+ out_mdev = out_dev ? mpls_dev_get(out_dev) : NULL;
+ if (out_mdev)
+ MPLS_INC_STATS(out_mdev, tx_errors);
+ goto drop;
+err:
+ MPLS_INC_STATS(mdev, rx_errors);
drop:
kfree_skb(skb);
return NET_RX_DROP;
@@ -853,6 +894,70 @@
return err;
}
+static void mpls_get_stats(struct mpls_dev *mdev,
+ struct mpls_link_stats *stats)
+{
+ struct mpls_pcpu_stats *p;
+ int i;
+
+ memset(stats, 0, sizeof(*stats));
+
+ for_each_possible_cpu(i) {
+ struct mpls_link_stats local;
+ unsigned int start;
+
+ p = per_cpu_ptr(mdev->stats, i);
+ do {
+ start = u64_stats_fetch_begin(&p->syncp);
+ local = p->stats;
+ } while (u64_stats_fetch_retry(&p->syncp, start));
+
+ stats->rx_packets += local.rx_packets;
+ stats->rx_bytes += local.rx_bytes;
+ stats->tx_packets += local.tx_packets;
+ stats->tx_bytes += local.tx_bytes;
+ stats->rx_errors += local.rx_errors;
+ stats->tx_errors += local.tx_errors;
+ stats->rx_dropped += local.rx_dropped;
+ stats->tx_dropped += local.tx_dropped;
+ stats->rx_noroute += local.rx_noroute;
+ }
+}
+
+static int mpls_fill_stats_af(struct sk_buff *skb,
+ const struct net_device *dev)
+{
+ struct mpls_link_stats *stats;
+ struct mpls_dev *mdev;
+ struct nlattr *nla;
+
+ mdev = mpls_dev_get(dev);
+ if (!mdev)
+ return -ENODATA;
+
+ nla = nla_reserve_64bit(skb, MPLS_STATS_LINK,
+ sizeof(struct mpls_link_stats),
+ MPLS_STATS_UNSPEC);
+ if (!nla)
+ return -EMSGSIZE;
+
+ stats = nla_data(nla);
+ mpls_get_stats(mdev, stats);
+
+ return 0;
+}
+
+static size_t mpls_get_stats_af_size(const struct net_device *dev)
+{
+ struct mpls_dev *mdev;
+
+ mdev = mpls_dev_get(dev);
+ if (!mdev)
+ return 0;
+
+ return nla_total_size_64bit(sizeof(struct mpls_link_stats));
+}
+
#define MPLS_PERDEV_SYSCTL_OFFSET(field) \
(&((struct mpls_dev *)0)->field)
@@ -911,6 +1016,7 @@
{
struct mpls_dev *mdev;
int err = -ENOMEM;
+ int i;
ASSERT_RTNL();
@@ -918,6 +1024,17 @@
if (!mdev)
return ERR_PTR(err);
+ mdev->stats = alloc_percpu(struct mpls_pcpu_stats);
+ if (!mdev->stats)
+ goto free;
+
+ for_each_possible_cpu(i) {
+ struct mpls_pcpu_stats *mpls_stats;
+
+ mpls_stats = per_cpu_ptr(mdev->stats, i);
+ u64_stats_init(&mpls_stats->syncp);
+ }
+
err = mpls_dev_sysctl_register(dev, mdev);
if (err)
goto free;
@@ -927,10 +1044,19 @@
return mdev;
free:
+ free_percpu(mdev->stats);
kfree(mdev);
return ERR_PTR(err);
}
+static void mpls_dev_destroy_rcu(struct rcu_head *head)
+{
+ struct mpls_dev *mdev = container_of(head, struct mpls_dev, rcu);
+
+ free_percpu(mdev->stats);
+ kfree(mdev);
+}
+
static void mpls_ifdown(struct net_device *dev, int event)
{
struct mpls_route __rcu **platform_label;
@@ -1045,7 +1171,7 @@
if (mdev) {
mpls_dev_sysctl_unregister(mdev);
RCU_INIT_POINTER(dev->mpls_ptr, NULL);
- kfree_rcu(mdev, rcu);
+ call_rcu(&mdev->rcu, mpls_dev_destroy_rcu);
}
break;
case NETDEV_CHANGENAME:
@@ -1706,6 +1832,12 @@
.exit = mpls_net_exit,
};
+static struct rtnl_af_ops mpls_af_ops __read_mostly = {
+ .family = AF_MPLS,
+ .fill_stats_af = mpls_fill_stats_af,
+ .get_stats_af_size = mpls_get_stats_af_size,
+};
+
static int __init mpls_init(void)
{
int err;
@@ -1722,6 +1854,8 @@
dev_add_pack(&mpls_packet_type);
+ rtnl_af_register(&mpls_af_ops);
+
rtnl_register(PF_MPLS, RTM_NEWROUTE, mpls_rtm_newroute, NULL, NULL);
rtnl_register(PF_MPLS, RTM_DELROUTE, mpls_rtm_delroute, NULL, NULL);
rtnl_register(PF_MPLS, RTM_GETROUTE, NULL, mpls_dump_routes, NULL);
@@ -1738,6 +1872,7 @@
static void __exit mpls_exit(void)
{
rtnl_unregister_all(PF_MPLS);
+ rtnl_af_unregister(&mpls_af_ops);
dev_remove_pack(&mpls_packet_type);
unregister_netdevice_notifier(&mpls_dev_notifier);
unregister_pernet_subsys(&mpls_net_ops);