Commit c14fec39 authored by David S. Miller's avatar David S. Miller

Merge branch 'hv_netvsc-next'

Stephen Hemminger says:

====================
hv_netvsc changes

These are mostly about improving the handling of interaction between
the virtual network device (netvsc) and the SR-IOV VF network device.
====================
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
parents dd5a3005 f7ad75b7
...@@ -649,6 +649,8 @@ struct multi_recv_comp { ...@@ -649,6 +649,8 @@ struct multi_recv_comp {
struct netvsc_stats { struct netvsc_stats {
u64 packets; u64 packets;
u64 bytes; u64 bytes;
u64 broadcast;
u64 multicast;
struct u64_stats_sync syncp; struct u64_stats_sync syncp;
}; };
...@@ -695,9 +697,8 @@ struct net_device_context { ...@@ -695,9 +697,8 @@ struct net_device_context {
bool start_remove; bool start_remove;
/* State to manage the associated VF interface. */ /* State to manage the associated VF interface. */
struct net_device *vf_netdev; struct net_device __rcu *vf_netdev;
bool vf_inject;
atomic_t vf_use_cnt;
/* 1: allocated, serial number is valid. 0: not allocated */ /* 1: allocated, serial number is valid. 0: not allocated */
u32 vf_alloc; u32 vf_alloc;
/* Serial number of the VF to team with */ /* Serial number of the VF to team with */
......
...@@ -635,7 +635,7 @@ static void netvsc_send_tx_complete(struct netvsc_device *net_device, ...@@ -635,7 +635,7 @@ static void netvsc_send_tx_complete(struct netvsc_device *net_device,
q_idx = nvsc_packet->q_idx; q_idx = nvsc_packet->q_idx;
channel = incoming_channel; channel = incoming_channel;
dev_kfree_skb_any(skb); dev_consume_skb_any(skb);
} }
num_outstanding_sends = num_outstanding_sends =
...@@ -944,7 +944,7 @@ int netvsc_send(struct hv_device *device, ...@@ -944,7 +944,7 @@ int netvsc_send(struct hv_device *device,
} }
if (msdp->skb) if (msdp->skb)
dev_kfree_skb_any(msdp->skb); dev_consume_skb_any(msdp->skb);
if (xmit_more && !packet->cp_partial) { if (xmit_more && !packet->cp_partial) {
msdp->skb = skb; msdp->skb = skb;
......
...@@ -667,51 +667,23 @@ int netvsc_recv_callback(struct hv_device *device_obj, ...@@ -667,51 +667,23 @@ int netvsc_recv_callback(struct hv_device *device_obj,
{ {
struct net_device *net = hv_get_drvdata(device_obj); struct net_device *net = hv_get_drvdata(device_obj);
struct net_device_context *net_device_ctx = netdev_priv(net); struct net_device_context *net_device_ctx = netdev_priv(net);
struct net_device *vf_netdev;
struct sk_buff *skb; struct sk_buff *skb;
struct sk_buff *vf_skb;
struct netvsc_stats *rx_stats; struct netvsc_stats *rx_stats;
u32 bytes_recvd = packet->total_data_buflen;
int ret = 0;
if (!net || net->reg_state != NETREG_REGISTERED) if (net->reg_state != NETREG_REGISTERED)
return NVSP_STAT_FAIL; return NVSP_STAT_FAIL;
if (READ_ONCE(net_device_ctx->vf_inject)) { /*
atomic_inc(&net_device_ctx->vf_use_cnt); * If necessary, inject this packet into the VF interface.
if (!READ_ONCE(net_device_ctx->vf_inject)) { * On Hyper-V, multicast and brodcast packets are only delivered
/* * to the synthetic interface (after subjecting these to
* We raced; just move on. * policy filters on the host). Deliver these via the VF
*/ * interface in the guest.
atomic_dec(&net_device_ctx->vf_use_cnt); */
goto vf_injection_done; vf_netdev = rcu_dereference(net_device_ctx->vf_netdev);
} if (vf_netdev && (vf_netdev->flags & IFF_UP))
net = vf_netdev;
/*
* Inject this packet into the VF inerface.
* On Hyper-V, multicast and brodcast packets
* are only delivered on the synthetic interface
* (after subjecting these to policy filters on
* the host). Deliver these via the VF interface
* in the guest.
*/
vf_skb = netvsc_alloc_recv_skb(net_device_ctx->vf_netdev,
packet, csum_info, *data,
vlan_tci);
if (vf_skb != NULL) {
++net_device_ctx->vf_netdev->stats.rx_packets;
net_device_ctx->vf_netdev->stats.rx_bytes +=
bytes_recvd;
netif_receive_skb(vf_skb);
} else {
++net->stats.rx_dropped;
ret = NVSP_STAT_FAIL;
}
atomic_dec(&net_device_ctx->vf_use_cnt);
return ret;
}
vf_injection_done:
rx_stats = this_cpu_ptr(net_device_ctx->rx_stats);
/* Allocate a skb - TODO direct I/O to pages? */ /* Allocate a skb - TODO direct I/O to pages? */
skb = netvsc_alloc_recv_skb(net, packet, csum_info, *data, vlan_tci); skb = netvsc_alloc_recv_skb(net, packet, csum_info, *data, vlan_tci);
...@@ -719,12 +691,25 @@ int netvsc_recv_callback(struct hv_device *device_obj, ...@@ -719,12 +691,25 @@ int netvsc_recv_callback(struct hv_device *device_obj,
++net->stats.rx_dropped; ++net->stats.rx_dropped;
return NVSP_STAT_FAIL; return NVSP_STAT_FAIL;
} }
skb_record_rx_queue(skb, channel->
offermsg.offer.sub_channel_index);
if (net != vf_netdev)
skb_record_rx_queue(skb,
channel->offermsg.offer.sub_channel_index);
/*
* Even if injecting the packet, record the statistics
* on the synthetic device because modifying the VF device
* statistics will not work correctly.
*/
rx_stats = this_cpu_ptr(net_device_ctx->rx_stats);
u64_stats_update_begin(&rx_stats->syncp); u64_stats_update_begin(&rx_stats->syncp);
rx_stats->packets++; rx_stats->packets++;
rx_stats->bytes += packet->total_data_buflen; rx_stats->bytes += packet->total_data_buflen;
if (skb->pkt_type == PACKET_BROADCAST)
++rx_stats->broadcast;
else if (skb->pkt_type == PACKET_MULTICAST)
++rx_stats->multicast;
u64_stats_update_end(&rx_stats->syncp); u64_stats_update_end(&rx_stats->syncp);
/* /*
...@@ -967,7 +952,7 @@ static struct rtnl_link_stats64 *netvsc_get_stats64(struct net_device *net, ...@@ -967,7 +952,7 @@ static struct rtnl_link_stats64 *netvsc_get_stats64(struct net_device *net,
cpu); cpu);
struct netvsc_stats *rx_stats = per_cpu_ptr(ndev_ctx->rx_stats, struct netvsc_stats *rx_stats = per_cpu_ptr(ndev_ctx->rx_stats,
cpu); cpu);
u64 tx_packets, tx_bytes, rx_packets, rx_bytes; u64 tx_packets, tx_bytes, rx_packets, rx_bytes, rx_multicast;
unsigned int start; unsigned int start;
do { do {
...@@ -980,12 +965,14 @@ static struct rtnl_link_stats64 *netvsc_get_stats64(struct net_device *net, ...@@ -980,12 +965,14 @@ static struct rtnl_link_stats64 *netvsc_get_stats64(struct net_device *net,
start = u64_stats_fetch_begin_irq(&rx_stats->syncp); start = u64_stats_fetch_begin_irq(&rx_stats->syncp);
rx_packets = rx_stats->packets; rx_packets = rx_stats->packets;
rx_bytes = rx_stats->bytes; rx_bytes = rx_stats->bytes;
rx_multicast = rx_stats->multicast + rx_stats->broadcast;
} while (u64_stats_fetch_retry_irq(&rx_stats->syncp, start)); } while (u64_stats_fetch_retry_irq(&rx_stats->syncp, start));
t->tx_bytes += tx_bytes; t->tx_bytes += tx_bytes;
t->tx_packets += tx_packets; t->tx_packets += tx_packets;
t->rx_bytes += rx_bytes; t->rx_bytes += rx_bytes;
t->rx_packets += rx_packets; t->rx_packets += rx_packets;
t->multicast += rx_multicast;
} }
t->tx_dropped = net->stats.tx_dropped; t->tx_dropped = net->stats.tx_dropped;
...@@ -1215,22 +1202,44 @@ static void netvsc_free_netdev(struct net_device *netdev) ...@@ -1215,22 +1202,44 @@ static void netvsc_free_netdev(struct net_device *netdev)
free_netdev(netdev); free_netdev(netdev);
} }
static struct net_device *get_netvsc_net_device(char *mac) static struct net_device *get_netvsc_bymac(const u8 *mac)
{ {
struct net_device *dev, *found = NULL; struct net_device *dev;
ASSERT_RTNL(); ASSERT_RTNL();
for_each_netdev(&init_net, dev) { for_each_netdev(&init_net, dev) {
if (memcmp(dev->dev_addr, mac, ETH_ALEN) == 0) { if (dev->netdev_ops != &device_ops)
if (dev->netdev_ops != &device_ops) continue; /* not a netvsc device */
continue;
found = dev; if (ether_addr_equal(mac, dev->perm_addr))
break; return dev;
} }
return NULL;
}
static struct net_device *get_netvsc_byref(struct net_device *vf_netdev)
{
struct net_device *dev;
ASSERT_RTNL();
for_each_netdev(&init_net, dev) {
struct net_device_context *net_device_ctx;
if (dev->netdev_ops != &device_ops)
continue; /* not a netvsc device */
net_device_ctx = netdev_priv(dev);
if (net_device_ctx->nvdev == NULL)
continue; /* device is removed */
if (rtnl_dereference(net_device_ctx->vf_netdev) == vf_netdev)
return dev; /* a match */
} }
return found; return NULL;
} }
static int netvsc_register_vf(struct net_device *vf_netdev) static int netvsc_register_vf(struct net_device *vf_netdev)
...@@ -1238,9 +1247,8 @@ static int netvsc_register_vf(struct net_device *vf_netdev) ...@@ -1238,9 +1247,8 @@ static int netvsc_register_vf(struct net_device *vf_netdev)
struct net_device *ndev; struct net_device *ndev;
struct net_device_context *net_device_ctx; struct net_device_context *net_device_ctx;
struct netvsc_device *netvsc_dev; struct netvsc_device *netvsc_dev;
const struct ethtool_ops *eth_ops = vf_netdev->ethtool_ops;
if (eth_ops == NULL || eth_ops == &ethtool_ops) if (vf_netdev->addr_len != ETH_ALEN)
return NOTIFY_DONE; return NOTIFY_DONE;
/* /*
...@@ -1248,13 +1256,13 @@ static int netvsc_register_vf(struct net_device *vf_netdev) ...@@ -1248,13 +1256,13 @@ static int netvsc_register_vf(struct net_device *vf_netdev)
* associate with the VF interface. If we don't find a matching * associate with the VF interface. If we don't find a matching
* synthetic interface, move on. * synthetic interface, move on.
*/ */
ndev = get_netvsc_net_device(vf_netdev->dev_addr); ndev = get_netvsc_bymac(vf_netdev->perm_addr);
if (!ndev) if (!ndev)
return NOTIFY_DONE; return NOTIFY_DONE;
net_device_ctx = netdev_priv(ndev); net_device_ctx = netdev_priv(ndev);
netvsc_dev = net_device_ctx->nvdev; netvsc_dev = net_device_ctx->nvdev;
if (!netvsc_dev || net_device_ctx->vf_netdev) if (!netvsc_dev || rtnl_dereference(net_device_ctx->vf_netdev))
return NOTIFY_DONE; return NOTIFY_DONE;
netdev_info(ndev, "VF registering: %s\n", vf_netdev->name); netdev_info(ndev, "VF registering: %s\n", vf_netdev->name);
...@@ -1262,46 +1270,26 @@ static int netvsc_register_vf(struct net_device *vf_netdev) ...@@ -1262,46 +1270,26 @@ static int netvsc_register_vf(struct net_device *vf_netdev)
* Take a reference on the module. * Take a reference on the module.
*/ */
try_module_get(THIS_MODULE); try_module_get(THIS_MODULE);
net_device_ctx->vf_netdev = vf_netdev;
return NOTIFY_OK;
}
static void netvsc_inject_enable(struct net_device_context *net_device_ctx) dev_hold(vf_netdev);
{ rcu_assign_pointer(net_device_ctx->vf_netdev, vf_netdev);
net_device_ctx->vf_inject = true; return NOTIFY_OK;
}
static void netvsc_inject_disable(struct net_device_context *net_device_ctx)
{
net_device_ctx->vf_inject = false;
/* Wait for currently active users to drain out. */
while (atomic_read(&net_device_ctx->vf_use_cnt) != 0)
udelay(50);
} }
static int netvsc_vf_up(struct net_device *vf_netdev) static int netvsc_vf_up(struct net_device *vf_netdev)
{ {
struct net_device *ndev; struct net_device *ndev;
struct netvsc_device *netvsc_dev; struct netvsc_device *netvsc_dev;
const struct ethtool_ops *eth_ops = vf_netdev->ethtool_ops;
struct net_device_context *net_device_ctx; struct net_device_context *net_device_ctx;
if (eth_ops == &ethtool_ops) ndev = get_netvsc_byref(vf_netdev);
return NOTIFY_DONE;
ndev = get_netvsc_net_device(vf_netdev->dev_addr);
if (!ndev) if (!ndev)
return NOTIFY_DONE; return NOTIFY_DONE;
net_device_ctx = netdev_priv(ndev); net_device_ctx = netdev_priv(ndev);
netvsc_dev = net_device_ctx->nvdev; netvsc_dev = net_device_ctx->nvdev;
if (!netvsc_dev || !net_device_ctx->vf_netdev)
return NOTIFY_DONE;
netdev_info(ndev, "VF up: %s\n", vf_netdev->name); netdev_info(ndev, "VF up: %s\n", vf_netdev->name);
netvsc_inject_enable(net_device_ctx);
/* /*
* Open the device before switching data path. * Open the device before switching data path.
...@@ -1327,23 +1315,15 @@ static int netvsc_vf_down(struct net_device *vf_netdev) ...@@ -1327,23 +1315,15 @@ static int netvsc_vf_down(struct net_device *vf_netdev)
struct net_device *ndev; struct net_device *ndev;
struct netvsc_device *netvsc_dev; struct netvsc_device *netvsc_dev;
struct net_device_context *net_device_ctx; struct net_device_context *net_device_ctx;
const struct ethtool_ops *eth_ops = vf_netdev->ethtool_ops;
if (eth_ops == &ethtool_ops)
return NOTIFY_DONE;
ndev = get_netvsc_net_device(vf_netdev->dev_addr); ndev = get_netvsc_byref(vf_netdev);
if (!ndev) if (!ndev)
return NOTIFY_DONE; return NOTIFY_DONE;
net_device_ctx = netdev_priv(ndev); net_device_ctx = netdev_priv(ndev);
netvsc_dev = net_device_ctx->nvdev; netvsc_dev = net_device_ctx->nvdev;
if (!netvsc_dev || !net_device_ctx->vf_netdev)
return NOTIFY_DONE;
netdev_info(ndev, "VF down: %s\n", vf_netdev->name); netdev_info(ndev, "VF down: %s\n", vf_netdev->name);
netvsc_inject_disable(net_device_ctx);
netvsc_switch_datapath(ndev, false); netvsc_switch_datapath(ndev, false);
netdev_info(ndev, "Data path switched from VF: %s\n", vf_netdev->name); netdev_info(ndev, "Data path switched from VF: %s\n", vf_netdev->name);
rndis_filter_close(netvsc_dev); rndis_filter_close(netvsc_dev);
...@@ -1359,23 +1339,19 @@ static int netvsc_unregister_vf(struct net_device *vf_netdev) ...@@ -1359,23 +1339,19 @@ static int netvsc_unregister_vf(struct net_device *vf_netdev)
{ {
struct net_device *ndev; struct net_device *ndev;
struct netvsc_device *netvsc_dev; struct netvsc_device *netvsc_dev;
const struct ethtool_ops *eth_ops = vf_netdev->ethtool_ops;
struct net_device_context *net_device_ctx; struct net_device_context *net_device_ctx;
if (eth_ops == &ethtool_ops) ndev = get_netvsc_byref(vf_netdev);
return NOTIFY_DONE;
ndev = get_netvsc_net_device(vf_netdev->dev_addr);
if (!ndev) if (!ndev)
return NOTIFY_DONE; return NOTIFY_DONE;
net_device_ctx = netdev_priv(ndev); net_device_ctx = netdev_priv(ndev);
netvsc_dev = net_device_ctx->nvdev; netvsc_dev = net_device_ctx->nvdev;
if (!netvsc_dev || !net_device_ctx->vf_netdev)
return NOTIFY_DONE;
netdev_info(ndev, "VF unregistering: %s\n", vf_netdev->name); netdev_info(ndev, "VF unregistering: %s\n", vf_netdev->name);
netvsc_inject_disable(net_device_ctx);
net_device_ctx->vf_netdev = NULL; RCU_INIT_POINTER(net_device_ctx->vf_netdev, NULL);
dev_put(vf_netdev);
module_put(THIS_MODULE); module_put(THIS_MODULE);
return NOTIFY_OK; return NOTIFY_OK;
} }
...@@ -1427,10 +1403,6 @@ static int netvsc_probe(struct hv_device *dev, ...@@ -1427,10 +1403,6 @@ static int netvsc_probe(struct hv_device *dev,
spin_lock_init(&net_device_ctx->lock); spin_lock_init(&net_device_ctx->lock);
INIT_LIST_HEAD(&net_device_ctx->reconfig_events); INIT_LIST_HEAD(&net_device_ctx->reconfig_events);
atomic_set(&net_device_ctx->vf_use_cnt, 0);
net_device_ctx->vf_netdev = NULL;
net_device_ctx->vf_inject = false;
net->netdev_ops = &device_ops; net->netdev_ops = &device_ops;
net->hw_features = NETVSC_HW_FEATURES; net->hw_features = NETVSC_HW_FEATURES;
...@@ -1539,13 +1511,21 @@ static int netvsc_netdev_event(struct notifier_block *this, ...@@ -1539,13 +1511,21 @@ static int netvsc_netdev_event(struct notifier_block *this,
{ {
struct net_device *event_dev = netdev_notifier_info_to_dev(ptr); struct net_device *event_dev = netdev_notifier_info_to_dev(ptr);
/* Skip our own events */
if (event_dev->netdev_ops == &device_ops)
return NOTIFY_DONE;
/* Avoid non-Ethernet type devices */
if (event_dev->type != ARPHRD_ETHER)
return NOTIFY_DONE;
/* Avoid Vlan dev with same MAC registering as VF */ /* Avoid Vlan dev with same MAC registering as VF */
if (event_dev->priv_flags & IFF_802_1Q_VLAN) if (event_dev->priv_flags & IFF_802_1Q_VLAN)
return NOTIFY_DONE; return NOTIFY_DONE;
/* Avoid Bonding master dev with same MAC registering as VF */ /* Avoid Bonding master dev with same MAC registering as VF */
if (event_dev->priv_flags & IFF_BONDING && if ((event_dev->priv_flags & IFF_BONDING) &&
event_dev->flags & IFF_MASTER) (event_dev->flags & IFF_MASTER))
return NOTIFY_DONE; return NOTIFY_DONE;
switch (event) { switch (event) {
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment