[Intel-wired-lan] [next-queue v6 PATCH 6/7] i40e: Add support for exposing switch port statistics via port netdevs
Sridhar Samudrala
sridhar.samudrala at intel.com
Thu Mar 30 00:22:54 UTC 2017
By default stats counted by HW are returned via the original
ndo_get_stats64() api. Stats counted in SW are returned via
ndo_get_offload_stats() api.
Small script to demonstrate port stats in switchdev mode.
PF: p4p1, VFs: p4p1_0,p4p1_1 VF Port Reps:p4p1-vf0, p4p1-vf1
PF Port rep: p4p1-pf
# rmmod i40e; modprobe i40e
# devlink dev eswitch set pci/0000:05:00.0 mode switchdev
# echo 2 > /sys/class/net/enp5s0f0/device/sriov_numvfs
# ip link set p4p1 vf 0 mac 00:11:22:33:44:55
# ip link set p4p1 vf 1 mac 00:11:22:33:44:56
# rmmod i40evf; modprobe i40evf
/* Create 2 namespaces and move the VFs to the corresponding ns */
# ip netns add ns0
# ip link set p4p1_0 netns ns0
# ip netns exec ns0 ip addr add 192.168.1.10/24 dev p4p1_0
# ip netns exec ns0 ip link set p4p1_0 up
# ip netns add ns1
# ip link set p4p1_1 netns ns1
# ip netns exec ns1 ip addr add 192.168.1.11/24 dev p4p1_1
# ip netns exec ns1 ip link set p4p1_1 up
/* bring up pf and port netdevs */
# ip addr add 192.168.1.1/24 dev p4p1
# ip link set p4p1 up
# ip link set p4p1-vf0 up
# ip link set p4p1-vf1 up
# ip link set p4p1-pf up
# ip netns exec ns0 ping -c3 192.168.1.11 /* VF0 -> VF1 */
# ip netns exec ns1 ping -c3 192.168.1.10 /* VF1 -> VF0 */
# ping -c3 192.168.1.10 /* PF -> VF0 */
# ping -c3 192.168.1.11 /* PF -> VF1 */
/* VF0 -> IP in same subnet - broadcasts will be seen on p4p1-vf0 */
# ip netns exec ns0 ping -c1 -W1 192.168.1.200
/* VF1 -> IP in same subnet- broadcasts will be seen on p4p1-vf1 */
# ip netns exec ns0 ping -c1 -W1 192.168.1.200
/* port rep VF0 -> IP in same subnet - broadcasts will be seen on p4p1_0 */
# ping -I p4p1-vf0 -c1 -W1 192.168.1.200
/* port rep VF1 -> IP in same subnet - broadcasts will be seen on p4p1_1 */
# ping -I p4p1-vf1 -c1 -W1 192.168.1.200
HW STATS
# ip netns exec ns0 ip -s l show p4p1_0
41: p4p1_0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT qlen 1000
link/ether 00:11:22:33:44:55 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped overrun mcast
1274 21 0 0 0 0
TX: bytes packets errors dropped carrier collsns
980 14 0 0 0 0
# ip -s l show p4p1-vf0
37: p4p1-vf0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP mode DEFAULT qlen 1000
link/ether f6:07:98:0e:cd:97 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped overrun mcast
980 14 0 0 0 0
TX: bytes packets errors dropped carrier collsns
1274 21 0 0 0 0
# ip netns exec ns1 ip -s l show p4p1_1
42: p4p1_1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT qlen 1000
link/ether 00:11:22:33:44:56 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped overrun mcast
1246 19 0 0 0 0
TX: bytes packets errors dropped carrier collsns
1078 15 0 0 0 0
# ip -s l show p4p1-vf1
38: p4p1-vf1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP mode DEFAULT qlen 1000
link/ether 2a:cf:ff:6a:f3:66 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped overrun mcast
1078 15 0 0 0 0
TX: bytes packets errors dropped carrier collsns
1246 19 0 0 0 0
# ip -s l show p4p1
34: p4p1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT qlen 1000
link/ether 3c:fd:fe:a3:18:f8 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped overrun mcast
1134 17 0 0 0 0
TX: bytes packets errors dropped carrier collsns
966 19 0 0 0 0
# ip -s l show p4p1-pf
36: p4p1-pf: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP mode DEFAULT qlen 1000
link/ether da:0f:67:fe:2e:66 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped overrun mcast
966 19 0 0 0 0
TX: bytes packets errors dropped carrier collsns
882 17 0 0 0 0
SW STATS
# ifstat -a -x c
#kernel
Interface RX Pkts/Rate TX Pkts/Rate RX Data/Rate TX Data/Rate
RX Errs/Drop TX Errs/Drop RX Over/Rate TX Coll/Rate
p4p1-pf 0 0 3 0 0 0 126 0
0 0 0 0 0 0 0 0
p4p1-vf0 4 0 6 0 184 0 252 0
0 0 0 0 0 0 0 0
p4p1-vf1 3 0 3 0 138 0 126 0
0 0 0 0 0 0 0 0
Signed-off-by: Sridhar Samudrala <sridhar.samudrala at intel.com>
---
drivers/net/ethernet/intel/i40e/i40e.h | 10 +++
drivers/net/ethernet/intel/i40e/i40e_main.c | 125 ++++++++++++++++++++++++++++
drivers/net/ethernet/intel/i40e/i40e_txrx.c | 24 +++++-
3 files changed, 158 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/intel/i40e/i40e.h b/drivers/net/ethernet/intel/i40e/i40e.h
index ac11005..72e11b2 100644
--- a/drivers/net/ethernet/intel/i40e/i40e.h
+++ b/drivers/net/ethernet/intel/i40e/i40e.h
@@ -328,8 +328,18 @@ enum i40e_port_netdev_type {
I40E_PORT_NETDEV_VF
};
+struct port_netdev_pcpu_stats {
+ u64 tx_packets;
+ u64 tx_bytes;
+ u64 tx_drops;
+ u64 rx_packets;
+ u64 rx_bytes;
+ struct u64_stats_sync syncp;
+};
+
/* Port representor netdev private structure */
struct i40e_port_netdev_priv {
+ struct port_netdev_pcpu_stats __percpu *stats;
enum i40e_port_netdev_type type; /* type - PF or VF */
struct metadata_dst *dst; /* port id */
void *f; /* ptr to PF or VF struct */
diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index e9c5c6b..4f0eebc 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -11001,10 +11001,123 @@ static int i40e_port_netdev_stop(struct net_device *dev)
return err;
}
+/**
+ * i40e_port_netdev_get_stats64
+ * @dev: network interface device structure
+ * @stats: netlink stats structure
+ *
+ * Fills the hw statistics from the VSI corresponding to the associated port
+ * netdev
+ **/
+static void
+i40e_port_netdev_get_stats64(struct net_device *netdev,
+ struct rtnl_link_stats64 *stats)
+{
+ struct i40e_port_netdev_priv *priv = netdev_priv(netdev);
+ struct i40e_vf *vf;
+ struct i40e_pf *pf;
+ struct i40e_vsi *vsi;
+ struct i40e_eth_stats *estats;
+
+ switch (priv->type) {
+ case I40E_PORT_NETDEV_VF:
+ vf = (struct i40e_vf *)priv->f;
+ pf = vf->pf;
+ vsi = pf->vsi[vf->lan_vsi_idx];
+ break;
+ case I40E_PORT_NETDEV_PF:
+ pf = (struct i40e_pf *)priv->f;
+ vsi = pf->vsi[pf->lan_vsi];
+ break;
+ default:
+ return;
+ }
+
+ i40e_update_stats(vsi);
+ estats = &vsi->eth_stats;
+
+ /* TX and RX stats are flipped as we are returning the stats as seen
+ * at the switch port corresponding to the VF.
+ */
+ stats->rx_packets = estats->tx_unicast + estats->tx_multicast +
+ estats->tx_broadcast;
+ stats->tx_packets = estats->rx_unicast + estats->rx_multicast +
+ estats->rx_broadcast;
+ stats->rx_bytes = estats->tx_bytes;
+ stats->tx_bytes = estats->rx_bytes;
+ stats->rx_dropped = estats->tx_discards;
+ stats->tx_dropped = estats->rx_discards;
+}
+
+/**
+ * i40e_port_netdev_get_cpu_hit_stats64
+ * @dev: network interface device structure
+ * @stats: netlink stats structure
+ *
+ * stats are filled from the priv structure. correspond to the packets
+ * that are seen by the cpu and sent/received via port netdev.
+ **/
+static int
+i40e_port_netdev_get_cpu_hit_stats64(const struct net_device *dev,
+ struct rtnl_link_stats64 *stats)
+{
+ struct i40e_port_netdev_priv *priv = netdev_priv(dev);
+ int i;
+
+ for_each_possible_cpu(i) {
+ struct port_netdev_pcpu_stats *port_netdev_stats;
+ u64 tbytes, tpkts, tdrops, rbytes, rpkts;
+ unsigned int start;
+
+ port_netdev_stats = per_cpu_ptr(priv->stats, i);
+ do {
+ start = u64_stats_fetch_begin_irq(&port_netdev_stats->syncp);
+ tbytes = port_netdev_stats->tx_bytes;
+ tpkts = port_netdev_stats->tx_packets;
+ tdrops = port_netdev_stats->tx_drops;
+ rbytes = port_netdev_stats->rx_bytes;
+ rpkts = port_netdev_stats->rx_packets;
+ } while (u64_stats_fetch_retry_irq(&port_netdev_stats->syncp, start));
+ stats->tx_bytes += tbytes;
+ stats->tx_packets += tpkts;
+ stats->tx_dropped += tdrops;
+ stats->rx_bytes += rbytes;
+ stats->rx_packets += rpkts;
+ }
+
+ return 0;
+}
+
+static bool
+i40e_port_netdev_has_offload_stats(const struct net_device *dev, int attr_id)
+{
+ switch (attr_id) {
+ case IFLA_OFFLOAD_XSTATS_CPU_HIT:
+ return true;
+ }
+
+ return false;
+}
+
+static int
+i40e_port_netdev_get_offload_stats(int attr_id, const struct net_device *dev,
+ void *sp)
+{
+ switch (attr_id) {
+ case IFLA_OFFLOAD_XSTATS_CPU_HIT:
+ return i40e_port_netdev_get_cpu_hit_stats64(dev, sp);
+ }
+
+ return -EINVAL;
+}
+
static const struct net_device_ops i40e_port_netdev_ops = {
.ndo_open = i40e_port_netdev_open,
.ndo_stop = i40e_port_netdev_stop,
.ndo_start_xmit = i40e_port_netdev_start_xmit,
+ .ndo_get_stats64 = i40e_port_netdev_get_stats64,
+ .ndo_has_offload_stats = i40e_port_netdev_has_offload_stats,
+ .ndo_get_offload_stats = i40e_port_netdev_get_offload_stats,
};
/**
@@ -11077,6 +11190,16 @@ int i40e_alloc_port_netdev(void *f, enum i40e_port_netdev_type type)
return -EINVAL;
}
+ priv->stats = netdev_alloc_pcpu_stats(struct port_netdev_pcpu_stats);
+ if (!priv->stats) {
+ dev_err(&pf->pdev->dev,
+ "alloc_pcpu_stats failed for port netdev: %s\n",
+ port_netdev->name);
+ dst_release((struct dst_entry *)priv->dst);
+ free_netdev(port_netdev);
+ return -ENOMEM;
+ }
+
port_netdev->netdev_ops = &i40e_port_netdev_ops;
eth_hw_addr_random(port_netdev);
@@ -11144,6 +11267,7 @@ void i40e_free_port_netdev(void *f, enum i40e_port_netdev_type type)
pf->port_netdev->name);
priv = netdev_priv(pf->port_netdev);
dst_release((struct dst_entry *)priv->dst);
+ free_percpu(priv->stats);
unregister_netdev(pf->port_netdev);
free_netdev(pf->port_netdev);
pf->port_netdev = NULL;
@@ -11163,6 +11287,7 @@ void i40e_free_port_netdev(void *f, enum i40e_port_netdev_type type)
vf->port_netdev->name);
priv = netdev_priv(vf->port_netdev);
dst_release((struct dst_entry *)priv->dst);
+ free_percpu(priv->stats);
unregister_netdev(vf->port_netdev);
free_netdev(vf->port_netdev);
vf->port_netdev = NULL;
diff --git a/drivers/net/ethernet/intel/i40e/i40e_txrx.c b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
index 86d2510..449a35c 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
@@ -1309,7 +1309,9 @@ static bool i40e_alloc_mapped_page(struct i40e_ring *rx_ring,
static void i40e_handle_lpbk_skb(struct i40e_ring *rx_ring, struct sk_buff *skb)
{
struct i40e_q_vector *q_vector = rx_ring->q_vector;
+ struct port_netdev_pcpu_stats *port_netdev_stats;
struct i40e_pf *pf = rx_ring->vsi->back;
+ struct i40e_port_netdev_priv *priv;
struct sk_buff *nskb;
struct i40e_vf *vf;
struct ethhdr *eth;
@@ -1334,6 +1336,12 @@ static void i40e_handle_lpbk_skb(struct i40e_ring *rx_ring, struct sk_buff *skb)
break;
nskb->offload_fwd_mark = 1;
nskb->dev = vf->port_netdev;
+ priv = netdev_priv(vf->port_netdev);
+ port_netdev_stats = this_cpu_ptr(priv->stats);
+ u64_stats_update_begin(&port_netdev_stats->syncp);
+ port_netdev_stats->rx_packets++;
+ port_netdev_stats->rx_bytes += nskb->len;
+ u64_stats_update_end(&port_netdev_stats->syncp);
napi_gro_receive(&q_vector->napi, nskb);
break;
}
@@ -3283,6 +3291,7 @@ netdev_tx_t i40e_port_netdev_start_xmit(struct sk_buff *skb,
struct i40e_vsi *vsi;
struct i40e_pf *pf;
struct i40e_vf *vf;
+ int ret;
switch (priv->type) {
case I40E_PORT_NETDEV_VF:
@@ -3302,5 +3311,18 @@ netdev_tx_t i40e_port_netdev_start_xmit(struct sk_buff *skb,
skb_dst_set(skb, &priv->dst->dst);
skb->dev = vsi->netdev;
- return dev_queue_xmit(skb);
+ ret = dev_queue_xmit(skb);
+ if (likely(ret == NET_XMIT_SUCCESS || ret == NET_XMIT_CN)) {
+ struct port_netdev_pcpu_stats *port_netdev_stats;
+
+ port_netdev_stats = this_cpu_ptr(priv->stats);
+ u64_stats_update_begin(&port_netdev_stats->syncp);
+ port_netdev_stats->tx_packets++;
+ port_netdev_stats->tx_bytes += skb->len;
+ u64_stats_update_end(&port_netdev_stats->syncp);
+ } else {
+ this_cpu_inc(priv->stats->tx_drops);
+ }
+
+ return ret;
}
--
1.8.3.1
More information about the Intel-wired-lan
mailing list