From: Daniel Borkmann <daniel@xxxxxxxxxxxxx>
[ Upstream commit 34d21de99cea9cb17967874313e5b0262527833c ]
Move {l,t,d}stats allocation to the core and let netdevs pick the stats
type they need. That way the driver doesn't have to bother with error
handling (allocation failure checking, making sure free happens in the
right spot, etc) - all happening in the core.
Co-developed-by: Jakub Kicinski <kuba@xxxxxxxxxx>
Signed-off-by: Jakub Kicinski <kuba@xxxxxxxxxx>
Signed-off-by: Daniel Borkmann <daniel@xxxxxxxxxxxxx>
Reviewed-by: Nikolay Aleksandrov <razor@xxxxxxxxxxxxx>
Cc: David Ahern <dsahern@xxxxxxxxxx>
Link: https://lore.kernel.org/r/20231114004220.6495-3-daniel@xxxxxxxxxxxxx
Signed-off-by: Martin KaFai Lau <martin.lau@xxxxxxxxxx>
Stable-dep-of: 024ee930cb3c ("bpf: Fix dev's rx stats for bpf_redirect_peer traffic")
[ Note: Simplified vrf bits to reduce patch given unrelated to the fix ]
Signed-off-by: Daniel Borkmann <daniel@xxxxxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
---
drivers/net/veth.c | 16 +------------
drivers/net/vrf.c | 24 ++++++--------------
include/linux/netdevice.h | 30 ++++++++++++++++++++++----
net/core/dev.c | 53 +++++++++++++++++++++++++++++++++++++++++++---
4 files changed, 85 insertions(+), 38 deletions(-)
--- a/drivers/net/veth.c
+++ b/drivers/net/veth.c
@@ -1381,25 +1381,12 @@ static void veth_free_queues(struct net_
static int veth_dev_init(struct net_device *dev)
{
- int err;
-
- dev->lstats = netdev_alloc_pcpu_stats(struct pcpu_lstats);
- if (!dev->lstats)
- return -ENOMEM;
-
- err = veth_alloc_queues(dev);
- if (err) {
- free_percpu(dev->lstats);
- return err;
- }
-
- return 0;
+ return veth_alloc_queues(dev);
}
static void veth_dev_free(struct net_device *dev)
{
veth_free_queues(dev);
- free_percpu(dev->lstats);
}
#ifdef CONFIG_NET_POLL_CONTROLLER
@@ -1625,6 +1612,7 @@ static void veth_setup(struct net_device
NETIF_F_HW_VLAN_STAG_RX);
dev->needs_free_netdev = true;
dev->priv_destructor = veth_dev_free;
+ dev->pcpu_stat_type = NETDEV_PCPU_STAT_LSTATS;
dev->max_mtu = ETH_MAX_MTU;
dev->hw_features = VETH_FEATURES;
--- a/drivers/net/vrf.c
+++ b/drivers/net/vrf.c
@@ -121,22 +121,12 @@ struct net_vrf {
int ifindex;
};
-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;
-};
-
static void vrf_rx_stats(struct net_device *dev, int len)
{
struct pcpu_dstats *dstats = this_cpu_ptr(dev->dstats);
u64_stats_update_begin(&dstats->syncp);
- dstats->rx_pkts++;
+ dstats->rx_packets++;
dstats->rx_bytes += len;
u64_stats_update_end(&dstats->syncp);
}
@@ -161,10 +151,10 @@ static void vrf_get_stats64(struct net_d
do {
start = u64_stats_fetch_begin_irq(&dstats->syncp);
tbytes = dstats->tx_bytes;
- tpkts = dstats->tx_pkts;
- tdrops = dstats->tx_drps;
+ tpkts = dstats->tx_packets;
+ tdrops = dstats->tx_drops;
rbytes = dstats->rx_bytes;
- rpkts = dstats->rx_pkts;
+ rpkts = dstats->rx_packets;
} while (u64_stats_fetch_retry_irq(&dstats->syncp, start));
stats->tx_bytes += tbytes;
stats->tx_packets += tpkts;
@@ -421,7 +411,7 @@ static int vrf_local_xmit(struct sk_buff
if (likely(__netif_rx(skb) == NET_RX_SUCCESS))
vrf_rx_stats(dev, len);
else
- this_cpu_inc(dev->dstats->rx_drps);
+ this_cpu_inc(dev->dstats->rx_drops);
return NETDEV_TX_OK;
}
@@ -616,11 +606,11 @@ static netdev_tx_t vrf_xmit(struct sk_bu
struct pcpu_dstats *dstats = this_cpu_ptr(dev->dstats);
u64_stats_update_begin(&dstats->syncp);
- dstats->tx_pkts++;
+ dstats->tx_packets++;
dstats->tx_bytes += len;
u64_stats_update_end(&dstats->syncp);
} else {
- this_cpu_inc(dev->dstats->tx_drps);
+ this_cpu_inc(dev->dstats->tx_drops);
}
return ret;
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -1747,6 +1747,13 @@ enum netdev_ml_priv_type {
ML_PRIV_CAN,
};
+enum netdev_stat_type {
+ NETDEV_PCPU_STAT_NONE,
+ NETDEV_PCPU_STAT_LSTATS, /* struct pcpu_lstats */
+ NETDEV_PCPU_STAT_TSTATS, /* struct pcpu_sw_netstats */
+ NETDEV_PCPU_STAT_DSTATS, /* struct pcpu_dstats */
+};
+
/**
* struct net_device - The DEVICE structure.
*
@@ -1941,10 +1948,14 @@ enum netdev_ml_priv_type {
*
* @ml_priv: Mid-layer private
* @ml_priv_type: Mid-layer private type
- * @lstats: Loopback statistics
- * @tstats: Tunnel statistics
- * @dstats: Dummy statistics
- * @vstats: Virtual ethernet statistics
+ *
+ * @pcpu_stat_type: Type of device statistics which the core should
+ * allocate/free: none, lstats, tstats, dstats. none
+ * means the driver is handling statistics allocation/
+ * freeing internally.
+ * @lstats: Loopback statistics: packets, bytes
+ * @tstats: Tunnel statistics: RX/TX packets, RX/TX bytes
+ * @dstats: Dummy statistics: RX/TX/drop packets, RX/TX bytes
*
* @garp_port: GARP
* @mrp_port: MRP
@@ -2287,6 +2298,7 @@ struct net_device {
void *ml_priv;
enum netdev_ml_priv_type ml_priv_type;
+ enum netdev_stat_type pcpu_stat_type:8;
union {
struct pcpu_lstats __percpu *lstats;
struct pcpu_sw_netstats __percpu *tstats;
@@ -2670,6 +2682,16 @@ struct pcpu_sw_netstats {
struct u64_stats_sync syncp;
} __aligned(4 * sizeof(u64));
+struct pcpu_dstats {
+ u64 rx_packets;
+ u64 rx_bytes;
+ u64 rx_drops;
+ u64 tx_packets;
+ u64 tx_bytes;
+ u64 tx_drops;
+ struct u64_stats_sync syncp;
+} __aligned(8 * sizeof(u64));
+
struct pcpu_lstats {
u64_stats_t packets;
u64_stats_t bytes;
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -9991,6 +9991,46 @@ void netif_tx_stop_all_queues(struct net
}
EXPORT_SYMBOL(netif_tx_stop_all_queues);
+static int netdev_do_alloc_pcpu_stats(struct net_device *dev)
+{
+ void __percpu *v;
+
+ switch (dev->pcpu_stat_type) {
+ case NETDEV_PCPU_STAT_NONE:
+ return 0;
+ case NETDEV_PCPU_STAT_LSTATS:
+ v = dev->lstats = netdev_alloc_pcpu_stats(struct pcpu_lstats);
+ break;
+ case NETDEV_PCPU_STAT_TSTATS:
+ v = dev->tstats = netdev_alloc_pcpu_stats(struct pcpu_sw_netstats);
+ break;
+ case NETDEV_PCPU_STAT_DSTATS:
+ v = dev->dstats = netdev_alloc_pcpu_stats(struct pcpu_dstats);
+ break;
+ default:
+ return -EINVAL;
+ }
+
+ return v ? 0 : -ENOMEM;
+}
+
+static void netdev_do_free_pcpu_stats(struct net_device *dev)
+{
+ switch (dev->pcpu_stat_type) {
+ case NETDEV_PCPU_STAT_NONE:
+ return;
+ case NETDEV_PCPU_STAT_LSTATS:
+ free_percpu(dev->lstats);
+ break;
+ case NETDEV_PCPU_STAT_TSTATS:
+ free_percpu(dev->tstats);
+ break;
+ case NETDEV_PCPU_STAT_DSTATS:
+ free_percpu(dev->dstats);
+ break;
+ }
+}
+
/**
* register_netdevice() - register a network device
* @dev: device to register
@@ -10051,11 +10091,15 @@ int register_netdevice(struct net_device
goto err_uninit;
}
+ ret = netdev_do_alloc_pcpu_stats(dev);
+ if (ret)
+ goto err_uninit;
+
ret = -EBUSY;
if (!dev->ifindex)
dev->ifindex = dev_new_index(net);
else if (__dev_get_by_index(net, dev->ifindex))
- goto err_uninit;
+ goto err_free_pcpu;
/* Transfer changeable features to wanted_features and enable
* software offloads (GSO and GRO).
@@ -10102,14 +10146,14 @@ int register_netdevice(struct net_device
ret = call_netdevice_notifiers(NETDEV_POST_INIT, dev);
ret = notifier_to_errno(ret);
if (ret)
- goto err_uninit;
+ goto err_free_pcpu;
ret = netdev_register_kobject(dev);
write_lock(&dev_base_lock);
dev->reg_state = ret ? NETREG_UNREGISTERED : NETREG_REGISTERED;
write_unlock(&dev_base_lock);
if (ret)
- goto err_uninit;
+ goto err_free_pcpu;
__netdev_update_features(dev);
@@ -10156,6 +10200,8 @@ int register_netdevice(struct net_device
out:
return ret;
+err_free_pcpu:
+ netdev_do_free_pcpu_stats(dev);
err_uninit:
if (dev->netdev_ops->ndo_uninit)
dev->netdev_ops->ndo_uninit(dev);
@@ -10409,6 +10455,7 @@ void netdev_run_todo(void)
WARN_ON(rcu_access_pointer(dev->ip_ptr));
WARN_ON(rcu_access_pointer(dev->ip6_ptr));
+ netdev_do_free_pcpu_stats(dev);
if (dev->priv_destructor)
dev->priv_destructor(dev);
if (dev->needs_free_netdev)
Patches currently in stable-queue which might be from daniel@xxxxxxxxxxxxx are
queue-6.1/tcp_bpf-fix-the-sk_mem_uncharge-logic-in-tcp_bpf_sen.patch
queue-6.1/bpf-fix-dev-s-rx-stats-for-bpf_redirect_peer-traffic.patch
queue-6.1/bpf-remove-unnecessary-kfree-im_node-in-lpm_trie_upd.patch
queue-6.1/bpf-fix-helper-writes-to-read-only-maps.patch
queue-6.1/net-move-l-t-d-stats-allocation-to-core-and-convert-veth-vrf.patch
queue-6.1/veth-use-tstats-per-cpu-traffic-counters.patch
queue-6.1/bpf-handle-bpf_exist-and-bpf_noexist-for-lpm-trie.patch
queue-6.1/tools-override-makefile-arch-variable-if-defined-but.patch
