mirrors/linux-stable
1
0
Fork 0
mirror of https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git synced 2024-10-03 23:58:05 +00:00
Code Issues Releases Wiki Activity

vxlan: Use FDB_HASH_SIZE hash_locks to reduce contention

Browse source 
The monolithic hash_lock could cause huge contention when
inserting/deletiing vxlan_fdbs into the fdb_head.

Use FDB_HASH_SIZE hash_locks to protect insertions/deletions
of vxlan_fdbs into the fdb_head hash table.

Suggested-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Litao jiao <jiaolitao@raisecom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit is contained in:
Litao jiao 2019-06-06 17:57:58 +08:00 committed by David S. Miller
parent 4daa95af7f
commit fe1e0713bb
2 changed files with 60 additions and 34 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

92
drivers/net/vxlan.c
View file

@ -471,14 +471,19 @@ static u32 eth_vni_hash(const unsigned char *addr, __be32 vni)
return jhash_2words(key, vni, vxlan_salt) & (FDB_HASH_SIZE - 1);
}
static u32 fdb_head_index(struct vxlan_dev *vxlan, const u8 *mac, __be32 vni)
{
if (vxlan->cfg.flags & VXLAN_F_COLLECT_METADATA)
return eth_vni_hash(mac, vni);
else
return eth_hash(mac);
}
/* Hash chain to use given mac address */
static inline struct hlist_head *vxlan_fdb_head(struct vxlan_dev *vxlan,
const u8 *mac, __be32 vni)
{
if (vxlan->cfg.flags & VXLAN_F_COLLECT_METADATA)
return &vxlan->fdb_head[eth_vni_hash(mac, vni)];
else
return &vxlan->fdb_head[eth_hash(mac)];
return &vxlan->fdb_head[fdb_head_index(vxlan, mac, vni)];
}
/* Look up Ethernet address in forwarding table */
@ -593,8 +598,8 @@ int vxlan_fdb_replay(const struct net_device *dev, __be32 vni,
return -EINVAL;
vxlan = netdev_priv(dev);
spin_lock_bh(&vxlan->hash_lock);
for (h = 0; h < FDB_HASH_SIZE; ++h) {
spin_lock_bh(&vxlan->hash_lock[h]);
hlist_for_each_entry(f, &vxlan->fdb_head[h], hlist) {
if (f->vni == vni) {
list_for_each_entry(rdst, &f->remotes, list) {
@ -602,14 +607,16 @@ int vxlan_fdb_replay(const struct net_device *dev, __be32 vni,
f, rdst,
extack);
if (rc)
goto out;
goto unlock;
}
}
}
spin_unlock_bh(&vxlan->hash_lock[h]);
}
return 0;
out:
spin_unlock_bh(&vxlan->hash_lock);
unlock:
spin_unlock_bh(&vxlan->hash_lock[h]);
return rc;
}
EXPORT_SYMBOL_GPL(vxlan_fdb_replay);
@ -625,14 +632,15 @@ void vxlan_fdb_clear_offload(const struct net_device *dev, __be32 vni)
return;
vxlan = netdev_priv(dev);
spin_lock_bh(&vxlan->hash_lock);
for (h = 0; h < FDB_HASH_SIZE; ++h) {
spin_lock_bh(&vxlan->hash_lock[h]);
hlist_for_each_entry(f, &vxlan->fdb_head[h], hlist)
if (f->vni == vni)
list_for_each_entry(rdst, &f->remotes, list)
rdst->offloaded = false;
spin_unlock_bh(&vxlan->hash_lock[h]);
}
spin_unlock_bh(&vxlan->hash_lock);
}
EXPORT_SYMBOL_GPL(vxlan_fdb_clear_offload);
@ -1108,6 +1116,7 @@ static int vxlan_fdb_add(struct ndmsg *ndm, struct nlattr *tb[],
__be16 port;
__be32 src_vni, vni;
u32 ifindex;
u32 hash_index;
int err;
if (!(ndm->ndm_state & (NUD_PERMANENT|NUD_REACHABLE))) {
@ -1126,12 +1135,13 @@ static int vxlan_fdb_add(struct ndmsg *ndm, struct nlattr *tb[],
if (vxlan->default_dst.remote_ip.sa.sa_family != ip.sa.sa_family)
return -EAFNOSUPPORT;
spin_lock_bh(&vxlan->hash_lock);
hash_index = fdb_head_index(vxlan, addr, src_vni);
spin_lock_bh(&vxlan->hash_lock[hash_index]);
err = vxlan_fdb_update(vxlan, addr, &ip, ndm->ndm_state, flags,
port, src_vni, vni, ifindex,
ndm->ndm_flags | NTF_VXLAN_ADDED_BY_USER,
true, extack);
spin_unlock_bh(&vxlan->hash_lock);
spin_unlock_bh(&vxlan->hash_lock[hash_index]);
return err;
}
@ -1179,16 +1189,18 @@ static int vxlan_fdb_delete(struct ndmsg *ndm, struct nlattr *tb[],
__be32 src_vni, vni;
__be16 port;
u32 ifindex;
u32 hash_index;
int err;
err = vxlan_fdb_parse(tb, vxlan, &ip, &port, &src_vni, &vni, &ifindex);
if (err)
return err;
spin_lock_bh(&vxlan->hash_lock);
hash_index = fdb_head_index(vxlan, addr, src_vni);
spin_lock_bh(&vxlan->hash_lock[hash_index]);
err = __vxlan_fdb_delete(vxlan, addr, ip, port, src_vni, vni, ifindex,
true);
spin_unlock_bh(&vxlan->hash_lock);
spin_unlock_bh(&vxlan->hash_lock[hash_index]);
return err;
}
@ -1300,8 +1312,10 @@ static bool vxlan_snoop(struct net_device *dev,
f->updated = jiffies;
vxlan_fdb_notify(vxlan, f, rdst, RTM_NEWNEIGH, true, NULL);
} else {
u32 hash_index = fdb_head_index(vxlan, src_mac, vni);
/* learned new entry */
spin_lock(&vxlan->hash_lock);
spin_lock(&vxlan->hash_lock[hash_index]);
/* close off race between vxlan_flush and incoming packets */
if (netif_running(dev))
@ -1312,7 +1326,7 @@ static bool vxlan_snoop(struct net_device *dev,
vni,
vxlan->default_dst.remote_vni,
ifindex, NTF_SELF, true, NULL);
spin_unlock(&vxlan->hash_lock);
spin_unlock(&vxlan->hash_lock[hash_index]);
}
return false;
@ -2702,7 +2716,7 @@ static void vxlan_cleanup(struct timer_list *t)
for (h = 0; h < FDB_HASH_SIZE; ++h) {
struct hlist_node *p, *n;
spin_lock(&vxlan->hash_lock);
spin_lock(&vxlan->hash_lock[h]);
hlist_for_each_safe(p, n, &vxlan->fdb_head[h]) {
struct vxlan_fdb *f
= container_of(p, struct vxlan_fdb, hlist);
@ -2724,7 +2738,7 @@ static void vxlan_cleanup(struct timer_list *t)
} else if (time_before(timeout, next_timer))
next_timer = timeout;
}
spin_unlock(&vxlan->hash_lock);
spin_unlock(&vxlan->hash_lock[h]);
}
mod_timer(&vxlan->age_timer, next_timer);
@ -2767,12 +2781,13 @@ static int vxlan_init(struct net_device *dev)
static void vxlan_fdb_delete_default(struct vxlan_dev *vxlan, __be32 vni)
{
struct vxlan_fdb *f;
u32 hash_index = fdb_head_index(vxlan, all_zeros_mac, vni);
spin_lock_bh(&vxlan->hash_lock);
spin_lock_bh(&vxlan->hash_lock[hash_index]);
f = __vxlan_find_mac(vxlan, all_zeros_mac, vni);
if (f)
vxlan_fdb_destroy(vxlan, f, true, true);
spin_unlock_bh(&vxlan->hash_lock);
spin_unlock_bh(&vxlan->hash_lock[hash_index]);
}
static void vxlan_uninit(struct net_device *dev)
@ -2817,9 +2832,10 @@ static void vxlan_flush(struct vxlan_dev *vxlan, bool do_all)
{
unsigned int h;
spin_lock_bh(&vxlan->hash_lock);
for (h = 0; h < FDB_HASH_SIZE; ++h) {
struct hlist_node *p, *n;
spin_lock_bh(&vxlan->hash_lock[h]);
hlist_for_each_safe(p, n, &vxlan->fdb_head[h]) {
struct vxlan_fdb *f
= container_of(p, struct vxlan_fdb, hlist);
@ -2829,8 +2845,8 @@ static void vxlan_flush(struct vxlan_dev *vxlan, bool do_all)
if (!is_zero_ether_addr(f->eth_addr))
vxlan_fdb_destroy(vxlan, f, true, true);
}
spin_unlock_bh(&vxlan->hash_lock[h]);
}
spin_unlock_bh(&vxlan->hash_lock);
}
/* Cleanup timer and forwarding table on shutdown */
@ -3014,7 +3030,6 @@ static void vxlan_setup(struct net_device *dev)
dev->max_mtu = ETH_MAX_MTU;
INIT_LIST_HEAD(&vxlan->next);
spin_lock_init(&vxlan->hash_lock);
timer_setup(&vxlan->age_timer, vxlan_cleanup, TIMER_DEFERRABLE);
@ -3022,8 +3037,10 @@ static void vxlan_setup(struct net_device *dev)
gro_cells_init(&vxlan->gro_cells, dev);
for (h = 0; h < FDB_HASH_SIZE; ++h)
for (h = 0; h < FDB_HASH_SIZE; ++h) {
spin_lock_init(&vxlan->hash_lock[h]);
INIT_HLIST_HEAD(&vxlan->fdb_head[h]);
}
}
static void vxlan_ether_setup(struct net_device *dev)
@ -3917,7 +3934,9 @@ static int vxlan_changelink(struct net_device *dev, struct nlattr *tb[],
/* handle default dst entry */
if (!vxlan_addr_equal(&conf.remote_ip, &dst->remote_ip)) {
spin_lock_bh(&vxlan->hash_lock);
u32 hash_index = fdb_head_index(vxlan, all_zeros_mac, conf.vni);
spin_lock_bh(&vxlan->hash_lock[hash_index]);
if (!vxlan_addr_any(&conf.remote_ip)) {
err = vxlan_fdb_update(vxlan, all_zeros_mac,
&conf.remote_ip,
@ -3928,7 +3947,7 @@ static int vxlan_changelink(struct net_device *dev, struct nlattr *tb[],
conf.remote_ifindex,
NTF_SELF, true, extack);
if (err) {
spin_unlock_bh(&vxlan->hash_lock);
spin_unlock_bh(&vxlan->hash_lock[hash_index]);
return err;
}
}
@ -3940,7 +3959,7 @@ static int vxlan_changelink(struct net_device *dev, struct nlattr *tb[],
dst->remote_vni,
dst->remote_ifindex,
true);
spin_unlock_bh(&vxlan->hash_lock);
spin_unlock_bh(&vxlan->hash_lock[hash_index]);
}
if (conf.age_interval != vxlan->cfg.age_interval)
@ -4195,8 +4214,11 @@ vxlan_fdb_offloaded_set(struct net_device *dev,
struct vxlan_dev *vxlan = netdev_priv(dev);
struct vxlan_rdst *rdst;
struct vxlan_fdb *f;
u32 hash_index;
spin_lock_bh(&vxlan->hash_lock);
hash_index = fdb_head_index(vxlan, fdb_info->eth_addr, fdb_info->vni);
spin_lock_bh(&vxlan->hash_lock[hash_index]);
f = vxlan_find_mac(vxlan, fdb_info->eth_addr, fdb_info->vni);
if (!f)
@ -4212,7 +4234,7 @@ vxlan_fdb_offloaded_set(struct net_device *dev,
rdst->offloaded = fdb_info->offloaded;
out:
spin_unlock_bh(&vxlan->hash_lock);
spin_unlock_bh(&vxlan->hash_lock[hash_index]);
}
static int
@ -4221,11 +4243,13 @@ vxlan_fdb_external_learn_add(struct net_device *dev,
{
struct vxlan_dev *vxlan = netdev_priv(dev);
struct netlink_ext_ack *extack;
u32 hash_index;
int err;
hash_index = fdb_head_index(vxlan, fdb_info->eth_addr, fdb_info->vni);
extack = switchdev_notifier_info_to_extack(&fdb_info->info);
spin_lock_bh(&vxlan->hash_lock);
spin_lock_bh(&vxlan->hash_lock[hash_index]);
err = vxlan_fdb_update(vxlan, fdb_info->eth_addr, &fdb_info->remote_ip,
NUD_REACHABLE,
NLM_F_CREATE | NLM_F_REPLACE,
@ -4235,7 +4259,7 @@ vxlan_fdb_external_learn_add(struct net_device *dev,
fdb_info->remote_ifindex,
NTF_USE | NTF_SELF | NTF_EXT_LEARNED,
false, extack);
spin_unlock_bh(&vxlan->hash_lock);
spin_unlock_bh(&vxlan->hash_lock[hash_index]);
return err;
}
@ -4246,9 +4270,11 @@ vxlan_fdb_external_learn_del(struct net_device *dev,
{
struct vxlan_dev *vxlan = netdev_priv(dev);
struct vxlan_fdb *f;
u32 hash_index;
int err = 0;
spin_lock_bh(&vxlan->hash_lock);
hash_index = fdb_head_index(vxlan, fdb_info->eth_addr, fdb_info->vni);
spin_lock_bh(&vxlan->hash_lock[hash_index]);
f = vxlan_find_mac(vxlan, fdb_info->eth_addr, fdb_info->vni);
if (!f)
@ -4262,7 +4288,7 @@ vxlan_fdb_external_learn_del(struct net_device *dev,
fdb_info->remote_ifindex,
false);
spin_unlock_bh(&vxlan->hash_lock);
spin_unlock_bh(&vxlan->hash_lock[hash_index]);
return err;
}

2
include/net/vxlan.h
View file

@ -242,7 +242,7 @@ struct vxlan_dev {
struct vxlan_rdst default_dst; /* default destination */
struct timer_list age_timer;
spinlock_t hash_lock;
spinlock_t hash_lock[FDB_HASH_SIZE];
unsigned int addrcnt;
struct gro_cells gro_cells;