snprintf(bp->mii_bus->id, MII_BUS_ID_SIZE, "%s-%x",
bp->pdev->name, bp->pdev->id);
bp->mii_bus->priv = bp;
- bp->mii_bus->parent = &bp->dev->dev;
+ bp->mii_bus->parent = &bp->pdev->dev;
pdata = dev_get_platdata(&bp->pdev->dev);
dev_set_drvdata(&bp->dev->dev, bp->mii_bus);
struct phy_device *phydev;
phydev = mdiobus_scan(bp->mii_bus, i);
- if (IS_ERR(phydev)) {
+ if (IS_ERR(phydev) &&
+ PTR_ERR(phydev) != -ENODEV) {
err = PTR_ERR(phydev);
break;
}
if (err)
goto err_out_free_netdev;
+ err = macb_mii_init(bp);
+ if (err)
+ goto err_out_free_netdev;
+
+ phydev = bp->phy_dev;
+
+ netif_carrier_off(dev);
+
err = register_netdev(dev);
if (err) {
dev_err(&pdev->dev, "Cannot register net device, aborting.\n");
- goto err_out_unregister_netdev;
+ goto err_out_unregister_mdio;
}
- err = macb_mii_init(bp);
- if (err)
- goto err_out_unregister_netdev;
-
- netif_carrier_off(dev);
+ phy_attached_info(phydev);
netdev_info(dev, "Cadence %s rev 0x%08x at 0x%08lx irq %d (%pM)\n",
macb_is_gem(bp) ? "GEM" : "MACB", macb_readl(bp, MID),
dev->base_addr, dev->irq, dev->dev_addr);
- phydev = bp->phy_dev;
- phy_attached_info(phydev);
-
return 0;
-err_out_unregister_netdev:
- unregister_netdev(dev);
+err_out_unregister_mdio:
+ phy_disconnect(bp->phy_dev);
+ mdiobus_unregister(bp->mii_bus);
+ mdiobus_free(bp->mii_bus);
+
+ /* Shutdown the PHY if there is a GPIO reset */
+ if (bp->reset_gpio)
+ gpiod_set_value(bp->reset_gpio, 0);
err_out_free_netdev:
free_netdev(dev);
/* Enable per-CPU interrupts on the CPU that is
* brought up.
*/
- smp_call_function_single(cpu, mvneta_percpu_enable,
- pp, true);
+ mvneta_percpu_enable(pp);
/* Enable per-CPU interrupt on the one CPU we care
* about.
/* Disable per-CPU interrupts on the CPU that is
* brought down.
*/
- smp_call_function_single(cpu, mvneta_percpu_disable,
- pp, true);
+ mvneta_percpu_disable(pp);
break;
case CPU_DEAD:
return 0;
pep->phy = mdiobus_scan(pep->smi_bus, pep->phy_addr);
+ if (IS_ERR(pep->phy))
+ return PTR_ERR(pep->phy);
if (!pep->phy)
return -ENODEV;
bool "Mellanox Technologies ConnectX-4 Ethernet support"
depends on NETDEVICES && ETHERNET && PCI && MLX5_CORE
select PTP_1588_CLOCK
+ select VXLAN if MLX5_CORE=y
default n
---help---
Ethernet support in Mellanox Technologies ConnectX-4 NIC.
struct mlx5e_vxlan_db vxlan;
struct mlx5e_params params;
+ struct workqueue_struct *wq;
struct work_struct update_carrier_work;
struct work_struct set_rx_mode_work;
struct delayed_work update_stats_work;
mutex_lock(&priv->state_lock);
if (test_bit(MLX5E_STATE_OPENED, &priv->state)) {
mlx5e_update_stats(priv);
- schedule_delayed_work(dwork,
- msecs_to_jiffies(
- MLX5E_UPDATE_STATS_INTERVAL));
+ queue_delayed_work(priv->wq, dwork,
+ msecs_to_jiffies(MLX5E_UPDATE_STATS_INTERVAL));
}
mutex_unlock(&priv->state_lock);
}
switch (event) {
case MLX5_DEV_EVENT_PORT_UP:
case MLX5_DEV_EVENT_PORT_DOWN:
- schedule_work(&priv->update_carrier_work);
+ queue_work(priv->wq, &priv->update_carrier_work);
break;
default:
mlx5e_update_carrier(priv);
mlx5e_timestamp_init(priv);
- schedule_delayed_work(&priv->update_stats_work, 0);
+ queue_delayed_work(priv->wq, &priv->update_stats_work, 0);
return 0;
{
struct mlx5e_priv *priv = netdev_priv(dev);
- schedule_work(&priv->set_rx_mode_work);
+ queue_work(priv->wq, &priv->set_rx_mode_work);
}
static int mlx5e_set_mac(struct net_device *netdev, void *addr)
ether_addr_copy(netdev->dev_addr, saddr->sa_data);
netif_addr_unlock_bh(netdev);
- schedule_work(&priv->set_rx_mode_work);
+ queue_work(priv->wq, &priv->set_rx_mode_work);
return 0;
}
if (!mlx5e_vxlan_allowed(priv->mdev))
return;
- mlx5e_vxlan_add_port(priv, be16_to_cpu(port));
+ mlx5e_vxlan_queue_work(priv, sa_family, be16_to_cpu(port), 1);
}
static void mlx5e_del_vxlan_port(struct net_device *netdev,
if (!mlx5e_vxlan_allowed(priv->mdev))
return;
- mlx5e_vxlan_del_port(priv, be16_to_cpu(port));
+ mlx5e_vxlan_queue_work(priv, sa_family, be16_to_cpu(port), 0);
}
static netdev_features_t mlx5e_vxlan_features_check(struct mlx5e_priv *priv,
priv = netdev_priv(netdev);
+ priv->wq = create_singlethread_workqueue("mlx5e");
+ if (!priv->wq)
+ goto err_free_netdev;
+
err = mlx5_alloc_map_uar(mdev, &priv->cq_uar, false);
if (err) {
mlx5_core_err(mdev, "alloc_map uar failed, %d\n", err);
- goto err_free_netdev;
+ goto err_destroy_wq;
}
err = mlx5_core_alloc_pd(mdev, &priv->pdn);
vxlan_get_rx_port(netdev);
mlx5e_enable_async_events(priv);
- schedule_work(&priv->set_rx_mode_work);
+ queue_work(priv->wq, &priv->set_rx_mode_work);
return priv;
err_unmap_free_uar:
mlx5_unmap_free_uar(mdev, &priv->cq_uar);
+err_destroy_wq:
+ destroy_workqueue(priv->wq);
+
err_free_netdev:
free_netdev(netdev);
set_bit(MLX5E_STATE_DESTROYING, &priv->state);
- schedule_work(&priv->set_rx_mode_work);
+ queue_work(priv->wq, &priv->set_rx_mode_work);
mlx5e_disable_async_events(priv);
- flush_scheduled_work();
+ flush_workqueue(priv->wq);
if (test_bit(MLX5_INTERFACE_STATE_SHUTDOWN, &mdev->intf_state)) {
netif_device_detach(netdev);
mutex_lock(&priv->state_lock);
mlx5_core_dealloc_transport_domain(priv->mdev, priv->tdn);
mlx5_core_dealloc_pd(priv->mdev, priv->pdn);
mlx5_unmap_free_uar(priv->mdev, &priv->cq_uar);
+ cancel_delayed_work_sync(&priv->update_stats_work);
+ destroy_workqueue(priv->wq);
if (!test_bit(MLX5_INTERFACE_STATE_SHUTDOWN, &mdev->intf_state))
free_netdev(netdev);
void mlx5_unmap_free_uar(struct mlx5_core_dev *mdev, struct mlx5_uar *uar)
{
- iounmap(uar->map);
- iounmap(uar->bf_map);
+ if (uar->map)
+ iounmap(uar->map);
+ else
+ iounmap(uar->bf_map);
mlx5_cmd_free_uar(mdev, uar->index);
}
EXPORT_SYMBOL(mlx5_unmap_free_uar);
return vxlan;
}
-int mlx5e_vxlan_add_port(struct mlx5e_priv *priv, u16 port)
+static void mlx5e_vxlan_add_port(struct work_struct *work)
{
+ struct mlx5e_vxlan_work *vxlan_work =
+ container_of(work, struct mlx5e_vxlan_work, work);
+ struct mlx5e_priv *priv = vxlan_work->priv;
struct mlx5e_vxlan_db *vxlan_db = &priv->vxlan;
+ u16 port = vxlan_work->port;
struct mlx5e_vxlan *vxlan;
int err;
- err = mlx5e_vxlan_core_add_port_cmd(priv->mdev, port);
- if (err)
- return err;
+ if (mlx5e_vxlan_core_add_port_cmd(priv->mdev, port))
+ goto free_work;
vxlan = kzalloc(sizeof(*vxlan), GFP_KERNEL);
- if (!vxlan) {
- err = -ENOMEM;
+ if (!vxlan)
goto err_delete_port;
- }
vxlan->udp_port = port;
if (err)
goto err_free;
- return 0;
+ goto free_work;
err_free:
kfree(vxlan);
err_delete_port:
mlx5e_vxlan_core_del_port_cmd(priv->mdev, port);
- return err;
+free_work:
+ kfree(vxlan_work);
}
static void __mlx5e_vxlan_core_del_port(struct mlx5e_priv *priv, u16 port)
kfree(vxlan);
}
-void mlx5e_vxlan_del_port(struct mlx5e_priv *priv, u16 port)
+static void mlx5e_vxlan_del_port(struct work_struct *work)
{
- if (!mlx5e_vxlan_lookup_port(priv, port))
- return;
+ struct mlx5e_vxlan_work *vxlan_work =
+ container_of(work, struct mlx5e_vxlan_work, work);
+ struct mlx5e_priv *priv = vxlan_work->priv;
+ u16 port = vxlan_work->port;
__mlx5e_vxlan_core_del_port(priv, port);
+
+ kfree(vxlan_work);
+}
+
+void mlx5e_vxlan_queue_work(struct mlx5e_priv *priv, sa_family_t sa_family,
+ u16 port, int add)
+{
+ struct mlx5e_vxlan_work *vxlan_work;
+
+ vxlan_work = kmalloc(sizeof(*vxlan_work), GFP_ATOMIC);
+ if (!vxlan_work)
+ return;
+
+ if (add)
+ INIT_WORK(&vxlan_work->work, mlx5e_vxlan_add_port);
+ else
+ INIT_WORK(&vxlan_work->work, mlx5e_vxlan_del_port);
+
+ vxlan_work->priv = priv;
+ vxlan_work->port = port;
+ vxlan_work->sa_family = sa_family;
+ queue_work(priv->wq, &vxlan_work->work);
}
void mlx5e_vxlan_cleanup(struct mlx5e_priv *priv)
u16 udp_port;
};
+struct mlx5e_vxlan_work {
+ struct work_struct work;
+ struct mlx5e_priv *priv;
+ sa_family_t sa_family;
+ u16 port;
+};
+
static inline bool mlx5e_vxlan_allowed(struct mlx5_core_dev *mdev)
{
return (MLX5_CAP_ETH(mdev, tunnel_stateless_vxlan) &&
}
void mlx5e_vxlan_init(struct mlx5e_priv *priv);
-int mlx5e_vxlan_add_port(struct mlx5e_priv *priv, u16 port);
-void mlx5e_vxlan_del_port(struct mlx5e_priv *priv, u16 port);
+void mlx5e_vxlan_queue_work(struct mlx5e_priv *priv, sa_family_t sa_family,
+ u16 port, int add);
struct mlx5e_vxlan *mlx5e_vxlan_lookup_port(struct mlx5e_priv *priv, u16 port);
void mlx5e_vxlan_cleanup(struct mlx5e_priv *priv);
return (struct ethhdr *)skb_mac_header(skb);
}
+static inline struct ethhdr *inner_eth_hdr(const struct sk_buff *skb)
+{
+ return (struct ethhdr *)skb_inner_mac_header(skb);
+}
+
int eth_header_parse(const struct sk_buff *skb, unsigned char *haddr);
extern ssize_t sysfs_format_mac(char *buf, const unsigned char *addr, int len);
(skb->inner_protocol_type != ENCAP_TYPE_ETHER ||
skb->inner_protocol != htons(ETH_P_TEB) ||
(skb_inner_mac_header(skb) - skb_transport_header(skb) !=
- sizeof(struct udphdr) + sizeof(struct vxlanhdr))))
+ sizeof(struct udphdr) + sizeof(struct vxlanhdr)) ||
+ (skb->ip_summed != CHECKSUM_NONE &&
+ !can_checksum_protocol(features, inner_eth_hdr(skb)->h_proto))))
return features & ~(NETIF_F_CSUM_MASK | NETIF_F_GSO_MASK);
return features;
#include "bat_v_elp.h"
#include "bat_v_ogm.h"
+#include "hard-interface.h"
#include "hash.h"
#include "originator.h"
#include "packet.h"
+static void batadv_v_iface_activate(struct batadv_hard_iface *hard_iface)
+{
+ /* B.A.T.M.A.N. V does not use any queuing mechanism, therefore it can
+ * set the interface as ACTIVE right away, without any risk of race
+ * condition
+ */
+ if (hard_iface->if_status == BATADV_IF_TO_BE_ACTIVATED)
+ hard_iface->if_status = BATADV_IF_ACTIVE;
+}
+
static int batadv_v_iface_enable(struct batadv_hard_iface *hard_iface)
{
int ret;
static struct batadv_algo_ops batadv_batman_v __read_mostly = {
.name = "BATMAN_V",
+ .bat_iface_activate = batadv_v_iface_activate,
.bat_iface_enable = batadv_v_iface_enable,
.bat_iface_disable = batadv_v_iface_disable,
.bat_iface_update_mac = batadv_v_iface_update_mac,
* be sent to
* @bat_priv: the bat priv with all the soft interface information
* @ip_dst: ipv4 to look up in the DHT
+ * @vid: VLAN identifier
*
* An originator O is selected if and only if its DHT_ID value is one of three
* closest values (from the LEFT, with wrap around if needed) then the hash
* Return: the candidate array of size BATADV_DAT_CANDIDATE_NUM.
*/
static struct batadv_dat_candidate *
-batadv_dat_select_candidates(struct batadv_priv *bat_priv, __be32 ip_dst)
+batadv_dat_select_candidates(struct batadv_priv *bat_priv, __be32 ip_dst,
+ unsigned short vid)
{
int select;
batadv_dat_addr_t last_max = BATADV_DAT_ADDR_MAX, ip_key;
return NULL;
dat.ip = ip_dst;
- dat.vid = 0;
+ dat.vid = vid;
ip_key = (batadv_dat_addr_t)batadv_hash_dat(&dat,
BATADV_DAT_ADDR_MAX);
* @bat_priv: the bat priv with all the soft interface information
* @skb: payload to send
* @ip: the DHT key
+ * @vid: VLAN identifier
* @packet_subtype: unicast4addr packet subtype to use
*
* This function copies the skb with pskb_copy() and is sent as unicast packet
*/
static bool batadv_dat_send_data(struct batadv_priv *bat_priv,
struct sk_buff *skb, __be32 ip,
- int packet_subtype)
+ unsigned short vid, int packet_subtype)
{
int i;
bool ret = false;
struct sk_buff *tmp_skb;
struct batadv_dat_candidate *cand;
- cand = batadv_dat_select_candidates(bat_priv, ip);
+ cand = batadv_dat_select_candidates(bat_priv, ip, vid);
if (!cand)
goto out;
ret = true;
} else {
/* Send the request to the DHT */
- ret = batadv_dat_send_data(bat_priv, skb, ip_dst,
+ ret = batadv_dat_send_data(bat_priv, skb, ip_dst, vid,
BATADV_P_DAT_DHT_GET);
}
out:
/* Send the ARP reply to the candidates for both the IP addresses that
* the node obtained from the ARP reply
*/
- batadv_dat_send_data(bat_priv, skb, ip_src, BATADV_P_DAT_DHT_PUT);
- batadv_dat_send_data(bat_priv, skb, ip_dst, BATADV_P_DAT_DHT_PUT);
+ batadv_dat_send_data(bat_priv, skb, ip_src, vid, BATADV_P_DAT_DHT_PUT);
+ batadv_dat_send_data(bat_priv, skb, ip_dst, vid, BATADV_P_DAT_DHT_PUT);
}
/**
batadv_update_min_mtu(hard_iface->soft_iface);
+ if (bat_priv->bat_algo_ops->bat_iface_activate)
+ bat_priv->bat_algo_ops->bat_iface_activate(hard_iface);
+
out:
if (primary_if)
batadv_hardif_put(primary_if);
{
struct hlist_node *node_tmp;
struct batadv_neigh_node *neigh_node;
- struct batadv_hardif_neigh_node *hardif_neigh;
struct batadv_neigh_ifinfo *neigh_ifinfo;
struct batadv_algo_ops *bao;
batadv_neigh_ifinfo_put(neigh_ifinfo);
}
- hardif_neigh = batadv_hardif_neigh_get(neigh_node->if_incoming,
- neigh_node->addr);
- if (hardif_neigh) {
- /* batadv_hardif_neigh_get() increases refcount too */
- batadv_hardif_neigh_put(hardif_neigh);
- batadv_hardif_neigh_put(hardif_neigh);
- }
+ batadv_hardif_neigh_put(neigh_node->hardif_neigh);
if (bao->bat_neigh_free)
bao->bat_neigh_free(neigh_node);
neigh_node->orig_node = orig_node;
neigh_node->last_seen = jiffies;
+ /* increment unique neighbor refcount */
+ kref_get(&hardif_neigh->refcount);
+ neigh_node->hardif_neigh = hardif_neigh;
+
/* extra reference for return */
kref_init(&neigh_node->refcount);
kref_get(&neigh_node->refcount);
hlist_add_head_rcu(&neigh_node->list, &orig_node->neigh_list);
spin_unlock_bh(&orig_node->neigh_list_lock);
- /* increment unique neighbor refcount */
- kref_get(&hardif_neigh->refcount);
-
batadv_dbg(BATADV_DBG_BATMAN, orig_node->bat_priv,
"Creating new neighbor %pM for orig_node %pM on interface %s\n",
neigh_addr, orig_node->orig, hard_iface->net_dev->name);
tt_local_entry = container_of(ref, struct batadv_tt_local_entry,
common.refcount);
+ batadv_softif_vlan_put(tt_local_entry->vlan);
+
kfree_rcu(tt_local_entry, common.rcu);
}
kref_get(&tt_local->common.refcount);
tt_local->last_seen = jiffies;
tt_local->common.added_at = tt_local->last_seen;
+ tt_local->vlan = vlan;
/* the batman interface mac and multicast addresses should never be
* purged
struct batadv_tt_common_entry *tt_common_entry;
struct batadv_tt_local_entry *tt_local;
struct batadv_hard_iface *primary_if;
- struct batadv_softif_vlan *vlan;
struct hlist_head *head;
unsigned short vid;
u32 i;
last_seen_msecs = last_seen_msecs % 1000;
no_purge = tt_common_entry->flags & np_flag;
-
- vlan = batadv_softif_vlan_get(bat_priv, vid);
- if (!vlan) {
- seq_printf(seq, "Cannot retrieve VLAN %d\n",
- BATADV_PRINT_VID(vid));
- continue;
- }
-
seq_printf(seq,
" * %pM %4i [%c%c%c%c%c%c] %3u.%03u (%#.8x)\n",
tt_common_entry->addr,
BATADV_TT_CLIENT_ISOLA) ? 'I' : '.'),
no_purge ? 0 : last_seen_secs,
no_purge ? 0 : last_seen_msecs,
- vlan->tt.crc);
-
- batadv_softif_vlan_put(vlan);
+ tt_local->vlan->tt.crc);
}
rcu_read_unlock();
}
{
struct batadv_tt_local_entry *tt_local_entry;
u16 flags, curr_flags = BATADV_NO_FLAGS;
- struct batadv_softif_vlan *vlan;
void *tt_entry_exists;
tt_local_entry = batadv_tt_local_hash_find(bat_priv, addr, vid);
/* extra call to free the local tt entry */
batadv_tt_local_entry_put(tt_local_entry);
- /* decrease the reference held for this vlan */
- vlan = batadv_softif_vlan_get(bat_priv, vid);
- if (!vlan)
- goto out;
-
- batadv_softif_vlan_put(vlan);
- batadv_softif_vlan_put(vlan);
-
out:
if (tt_local_entry)
batadv_tt_local_entry_put(tt_local_entry);
spinlock_t *list_lock; /* protects write access to the hash lists */
struct batadv_tt_common_entry *tt_common_entry;
struct batadv_tt_local_entry *tt_local;
- struct batadv_softif_vlan *vlan;
struct hlist_node *node_tmp;
struct hlist_head *head;
u32 i;
struct batadv_tt_local_entry,
common);
- /* decrease the reference held for this vlan */
- vlan = batadv_softif_vlan_get(bat_priv,
- tt_common_entry->vid);
- if (vlan) {
- batadv_softif_vlan_put(vlan);
- batadv_softif_vlan_put(vlan);
- }
-
batadv_tt_local_entry_put(tt_local);
}
spin_unlock_bh(list_lock);
struct batadv_hashtable *hash = bat_priv->tt.local_hash;
struct batadv_tt_common_entry *tt_common;
struct batadv_tt_local_entry *tt_local;
- struct batadv_softif_vlan *vlan;
struct hlist_node *node_tmp;
struct hlist_head *head;
spinlock_t *list_lock; /* protects write access to the hash lists */
struct batadv_tt_local_entry,
common);
- /* decrease the reference held for this vlan */
- vlan = batadv_softif_vlan_get(bat_priv, tt_common->vid);
- if (vlan) {
- batadv_softif_vlan_put(vlan);
- batadv_softif_vlan_put(vlan);
- }
-
batadv_tt_local_entry_put(tt_local);
}
spin_unlock_bh(list_lock);
* @ifinfo_lock: lock protecting private ifinfo members and list
* @if_incoming: pointer to incoming hard-interface
* @last_seen: when last packet via this neighbor was received
+ * @hardif_neigh: hardif_neigh of this neighbor
* @refcount: number of contexts the object is used
* @rcu: struct used for freeing in an RCU-safe manner
*/
spinlock_t ifinfo_lock; /* protects ifinfo_list and its members */
struct batadv_hard_iface *if_incoming;
unsigned long last_seen;
+ struct batadv_hardif_neigh_node *hardif_neigh;
struct kref refcount;
struct rcu_head rcu;
};
* struct batadv_tt_local_entry - translation table local entry data
* @common: general translation table data
* @last_seen: timestamp used for purging stale tt local entries
+ * @vlan: soft-interface vlan of the entry
*/
struct batadv_tt_local_entry {
struct batadv_tt_common_entry common;
unsigned long last_seen;
+ struct batadv_softif_vlan *vlan;
};
/**
* struct batadv_algo_ops - mesh algorithm callbacks
* @list: list node for the batadv_algo_list
* @name: name of the algorithm
+ * @bat_iface_activate: start routing mechanisms when hard-interface is brought
+ * up
* @bat_iface_enable: init routing info when hard-interface is enabled
* @bat_iface_disable: de-init routing info when hard-interface is disabled
* @bat_iface_update_mac: (re-)init mac addresses of the protocol information
struct batadv_algo_ops {
struct hlist_node list;
char *name;
+ void (*bat_iface_activate)(struct batadv_hard_iface *hard_iface);
int (*bat_iface_enable)(struct batadv_hard_iface *hard_iface);
void (*bat_iface_disable)(struct batadv_hard_iface *hard_iface);
void (*bat_iface_update_mac)(struct batadv_hard_iface *hard_iface);
if (skb->ip_summed != CHECKSUM_NONE &&
!can_checksum_protocol(features, type)) {
- features &= ~NETIF_F_CSUM_MASK;
+ features &= ~(NETIF_F_CSUM_MASK | NETIF_F_GSO_MASK);
} else if (illegal_highdma(skb->dev, skb)) {
features &= ~NETIF_F_SG;
}
static int ila_encap_nlsize(struct lwtunnel_state *lwtstate)
{
- /* No encapsulation overhead */
- return 0;
+ return nla_total_size(sizeof(u64)); /* ILA_ATTR_LOCATOR */
}
static int ila_encap_cmp(struct lwtunnel_state *a, struct lwtunnel_state *b)
/*
* This is the only path that sets tc->t_sock. Send and receive trust that
- * it is set. The RDS_CONN_CONNECTED bit protects those paths from being
+ * it is set. The RDS_CONN_UP bit protects those paths from being
* called while it isn't set.
*/
void rds_tcp_set_callbacks(struct socket *sock, struct rds_connection *conn)
if (!tc)
return -ENOMEM;
+ mutex_init(&tc->t_conn_lock);
tc->t_sock = NULL;
tc->t_tinc = NULL;
tc->t_tinc_hdr_rem = sizeof(struct rds_header);
struct list_head t_tcp_node;
struct rds_connection *conn;
+ /* t_conn_lock synchronizes the connection establishment between
+ * rds_tcp_accept_one and rds_tcp_conn_connect
+ */
+ struct mutex t_conn_lock;
struct socket *t_sock;
void *t_orig_write_space;
void *t_orig_data_ready;
struct socket *sock = NULL;
struct sockaddr_in src, dest;
int ret;
+ struct rds_tcp_connection *tc = conn->c_transport_data;
+
+ mutex_lock(&tc->t_conn_lock);
+ if (rds_conn_up(conn)) {
+ mutex_unlock(&tc->t_conn_lock);
+ return 0;
+ }
ret = sock_create_kern(rds_conn_net(conn), PF_INET,
SOCK_STREAM, IPPROTO_TCP, &sock);
if (ret < 0)
}
out:
+ mutex_unlock(&tc->t_conn_lock);
if (sock)
sock_release(sock);
return ret;
struct rds_connection *conn;
int ret;
struct inet_sock *inet;
- struct rds_tcp_connection *rs_tcp;
+ struct rds_tcp_connection *rs_tcp = NULL;
+ int conn_state;
+ struct sock *nsk;
ret = sock_create_kern(sock_net(sock->sk), sock->sk->sk_family,
sock->sk->sk_type, sock->sk->sk_protocol,
* rds_tcp_state_change() will do that cleanup
*/
rs_tcp = (struct rds_tcp_connection *)conn->c_transport_data;
- if (rs_tcp->t_sock &&
- ntohl(inet->inet_saddr) < ntohl(inet->inet_daddr)) {
- struct sock *nsk = new_sock->sk;
-
- nsk->sk_user_data = NULL;
- nsk->sk_prot->disconnect(nsk, 0);
- tcp_done(nsk);
- new_sock = NULL;
- ret = 0;
- goto out;
- } else if (rs_tcp->t_sock) {
- rds_tcp_restore_callbacks(rs_tcp->t_sock, rs_tcp);
- conn->c_outgoing = 0;
- }
-
rds_conn_transition(conn, RDS_CONN_DOWN, RDS_CONN_CONNECTING);
+ mutex_lock(&rs_tcp->t_conn_lock);
+ conn_state = rds_conn_state(conn);
+ if (conn_state != RDS_CONN_CONNECTING && conn_state != RDS_CONN_UP)
+ goto rst_nsk;
+ if (rs_tcp->t_sock) {
+ /* Need to resolve a duelling SYN between peers.
+ * We have an outstanding SYN to this peer, which may
+ * potentially have transitioned to the RDS_CONN_UP state,
+ * so we must quiesce any send threads before resetting
+ * c_transport_data.
+ */
+ wait_event(conn->c_waitq,
+ !test_bit(RDS_IN_XMIT, &conn->c_flags));
+ if (ntohl(inet->inet_saddr) < ntohl(inet->inet_daddr)) {
+ goto rst_nsk;
+ } else if (rs_tcp->t_sock) {
+ rds_tcp_restore_callbacks(rs_tcp->t_sock, rs_tcp);
+ conn->c_outgoing = 0;
+ }
+ }
rds_tcp_set_callbacks(new_sock, conn);
- rds_connect_complete(conn);
+ rds_connect_complete(conn); /* marks RDS_CONN_UP */
+ new_sock = NULL;
+ ret = 0;
+ goto out;
+rst_nsk:
+ /* reset the newly returned accept sock and bail */
+ nsk = new_sock->sk;
+ rds_tcp_stats_inc(s_tcp_listen_closed_stale);
+ nsk->sk_user_data = NULL;
+ nsk->sk_prot->disconnect(nsk, 0);
+ tcp_done(nsk);
new_sock = NULL;
ret = 0;
-
out:
+ if (rs_tcp)
+ mutex_unlock(&rs_tcp->t_conn_lock);
if (new_sock)
sock_release(new_sock);
return ret;
sch->q.qlen++;
}
+/* netem can't properly corrupt a megapacket (like we get from GSO), so instead
+ * when we statistically choose to corrupt one, we instead segment it, returning
+ * the first packet to be corrupted, and re-enqueue the remaining frames
+ */
+static struct sk_buff *netem_segment(struct sk_buff *skb, struct Qdisc *sch)
+{
+ struct sk_buff *segs;
+ netdev_features_t features = netif_skb_features(skb);
+
+ segs = skb_gso_segment(skb, features & ~NETIF_F_GSO_MASK);
+
+ if (IS_ERR_OR_NULL(segs)) {
+ qdisc_reshape_fail(skb, sch);
+ return NULL;
+ }
+ consume_skb(skb);
+ return segs;
+}
+
/*
* Insert one skb into qdisc.
* Note: parent depends on return value to account for queue length.
/* We don't fill cb now as skb_unshare() may invalidate it */
struct netem_skb_cb *cb;
struct sk_buff *skb2;
+ struct sk_buff *segs = NULL;
+ unsigned int len = 0, last_len, prev_len = qdisc_pkt_len(skb);
+ int nb = 0;
int count = 1;
+ int rc = NET_XMIT_SUCCESS;
/* Random duplication */
if (q->duplicate && q->duplicate >= get_crandom(&q->dup_cor))
* do it now in software before we mangle it.
*/
if (q->corrupt && q->corrupt >= get_crandom(&q->corrupt_cor)) {
+ if (skb_is_gso(skb)) {
+ segs = netem_segment(skb, sch);
+ if (!segs)
+ return NET_XMIT_DROP;
+ } else {
+ segs = skb;
+ }
+
+ skb = segs;
+ segs = segs->next;
+
if (!(skb = skb_unshare(skb, GFP_ATOMIC)) ||
(skb->ip_summed == CHECKSUM_PARTIAL &&
- skb_checksum_help(skb)))
- return qdisc_drop(skb, sch);
+ skb_checksum_help(skb))) {
+ rc = qdisc_drop(skb, sch);
+ goto finish_segs;
+ }
skb->data[prandom_u32() % skb_headlen(skb)] ^=
1<<(prandom_u32() % 8);
sch->qstats.requeues++;
}
+finish_segs:
+ if (segs) {
+ while (segs) {
+ skb2 = segs->next;
+ segs->next = NULL;
+ qdisc_skb_cb(segs)->pkt_len = segs->len;
+ last_len = segs->len;
+ rc = qdisc_enqueue(segs, sch);
+ if (rc != NET_XMIT_SUCCESS) {
+ if (net_xmit_drop_count(rc))
+ qdisc_qstats_drop(sch);
+ } else {
+ nb++;
+ len += last_len;
+ }
+ segs = skb2;
+ }
+ sch->q.qlen += nb;
+ if (nb > 1)
+ qdisc_tree_reduce_backlog(sch, 1 - nb, prev_len - len);
+ }
return NET_XMIT_SUCCESS;
}