Commit b099ce26 authored by Eric W. Biederman's avatar Eric W. Biederman Committed by David S. Miller

net: Batch inet_twsk_purge

This function walks the whole hashtable so there is no point in
passing it a network namespace.  Instead I purge all timewait
sockets from dead network namespaces that I find.  If the namespace
is one of the once I am trying to purge I am guaranteed no new timewait
sockets can be formed so this will get them all.  If the namespace
is one I am not acting for it might form a few more but I will
call inet_twsk_purge again and  shortly to get rid of them.  In
any even if the network namespace is dead timewait sockets are
useless.

Move the calls of inet_twsk_purge into batch_exit routines so
that if I am killing a bunch of namespaces at once I will just
call inet_twsk_purge once and save a lot of redundant unnecessary
work.

My simple 4k network namespace exit test the cleanup time dropped from
roughly 8.2s to 1.6s.  While the time spent running inet_twsk_purge fell
to about 2ms.  1ms for ipv4 and 1ms for ipv6.
Signed-off-by: default avatarEric W. Biederman <ebiederm@xmission.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
parent 575f4cd5
......@@ -212,14 +212,14 @@ extern void inet_twsk_schedule(struct inet_timewait_sock *tw,
extern void inet_twsk_deschedule(struct inet_timewait_sock *tw,
struct inet_timewait_death_row *twdr);
extern void inet_twsk_purge(struct net *net, struct inet_hashinfo *hashinfo,
extern void inet_twsk_purge(struct inet_hashinfo *hashinfo,
struct inet_timewait_death_row *twdr, int family);
static inline
struct net *twsk_net(const struct inet_timewait_sock *twsk)
{
#ifdef CONFIG_NET_NS
return twsk->tw_net;
return rcu_dereference(twsk->tw_net);
#else
return &init_net;
#endif
......@@ -229,7 +229,7 @@ static inline
void twsk_net_set(struct inet_timewait_sock *twsk, struct net *net)
{
#ifdef CONFIG_NET_NS
twsk->tw_net = net;
rcu_assign_pointer(twsk->tw_net, net);
#endif
}
#endif /* _INET_TIMEWAIT_SOCK_ */
......@@ -421,7 +421,7 @@ void inet_twdr_twcal_tick(unsigned long data)
EXPORT_SYMBOL_GPL(inet_twdr_twcal_tick);
void inet_twsk_purge(struct net *net, struct inet_hashinfo *hashinfo,
void inet_twsk_purge(struct inet_hashinfo *hashinfo,
struct inet_timewait_death_row *twdr, int family)
{
struct inet_timewait_sock *tw;
......@@ -436,15 +436,15 @@ void inet_twsk_purge(struct net *net, struct inet_hashinfo *hashinfo,
restart:
sk_nulls_for_each_rcu(sk, node, &head->twchain) {
tw = inet_twsk(sk);
if (!net_eq(twsk_net(tw), net) ||
tw->tw_family != family)
if ((tw->tw_family != family) ||
atomic_read(&twsk_net(tw)->count))
continue;
if (unlikely(!atomic_inc_not_zero(&tw->tw_refcnt)))
continue;
if (unlikely(!net_eq(twsk_net(tw), net) ||
tw->tw_family != family)) {
if (unlikely((tw->tw_family != family) ||
atomic_read(&twsk_net(tw)->count))) {
inet_twsk_put(tw);
goto restart;
}
......
......@@ -2529,12 +2529,17 @@ static int __net_init tcp_sk_init(struct net *net)
static void __net_exit tcp_sk_exit(struct net *net)
{
inet_ctl_sock_destroy(net->ipv4.tcp_sock);
inet_twsk_purge(net, &tcp_hashinfo, &tcp_death_row, AF_INET);
}
static void __net_exit tcp_sk_exit_batch(struct list_head *net_exit_list)
{
inet_twsk_purge(&tcp_hashinfo, &tcp_death_row, AF_INET);
}
static struct pernet_operations __net_initdata tcp_sk_ops = {
.init = tcp_sk_init,
.exit = tcp_sk_exit,
.init = tcp_sk_init,
.exit = tcp_sk_exit,
.exit_batch = tcp_sk_exit_batch,
};
void __init tcp_v4_init(void)
......
......@@ -2184,12 +2184,17 @@ static int tcpv6_net_init(struct net *net)
static void tcpv6_net_exit(struct net *net)
{
inet_ctl_sock_destroy(net->ipv6.tcp_sk);
inet_twsk_purge(net, &tcp_hashinfo, &tcp_death_row, AF_INET6);
}
static void tcpv6_net_exit_batch(struct list_head *net_exit_list)
{
inet_twsk_purge(&tcp_hashinfo, &tcp_death_row, AF_INET6);
}
static struct pernet_operations tcpv6_net_ops = {
.init = tcpv6_net_init,
.exit = tcpv6_net_exit,
.init = tcpv6_net_init,
.exit = tcpv6_net_exit,
.exit_batch = tcpv6_net_exit_batch,
};
int __init tcpv6_init(void)
......
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