Cong Wang
2012-Dec-04 09:56 UTC
[Bridge] [PATCH net-next v2] bridge: implement multicast fast leave
V2: make the toggle per-port Fast leave allows bridge to immediately stops the multicast traffic on the port receives IGMP Leave when IGMP snooping is enabled, no timeouts are observed. Cc: Herbert Xu <herbert at gondor.apana.org.au> Cc: Stephen Hemminger <shemminger at vyatta.com> Cc: "David S. Miller" <davem at davemloft.net> Signed-off-by: Cong Wang <amwang at redhat.com> --- net/bridge/br_multicast.c | 21 +++++++++++++++++++++ net/bridge/br_private.h | 1 + net/bridge/br_sysfs_if.c | 20 ++++++++++++++++++++ 3 files changed, 42 insertions(+), 0 deletions(-) diff --git a/net/bridge/br_multicast.c b/net/bridge/br_multicast.c index d53e4f4..02da618 100644 --- a/net/bridge/br_multicast.c +++ b/net/bridge/br_multicast.c @@ -1226,6 +1226,27 @@ static void br_multicast_leave_group(struct net_bridge *br, if (!mp) goto out; + if (port && port->multicast_fast_leave) { + struct net_bridge_port_group __rcu **pp; + + for (pp = &mp->ports; + (p = mlock_dereference(*pp, br)) != NULL; + pp = &p->next) { + if (p->port != port) + continue; + + rcu_assign_pointer(*pp, p->next); + hlist_del_init(&p->mglist); + del_timer(&p->timer); + call_rcu_bh(&p->rcu, br_multicast_free_pg); + + if (!mp->ports && !mp->mglist && + netif_running(br->dev)) + mod_timer(&mp->timer, jiffies); + } + goto out; + } + now = jiffies; time = now + br->multicast_last_member_count * br->multicast_last_member_interval; diff --git a/net/bridge/br_private.h b/net/bridge/br_private.h index 6484069..8f0e789 100644 --- a/net/bridge/br_private.h +++ b/net/bridge/br_private.h @@ -142,6 +142,7 @@ struct net_bridge_port #ifdef CONFIG_BRIDGE_IGMP_SNOOPING u32 multicast_startup_queries_sent; unsigned char multicast_router; + unsigned char multicast_fast_leave; struct timer_list multicast_router_timer; struct timer_list multicast_query_timer; struct hlist_head mglist; diff --git a/net/bridge/br_sysfs_if.c b/net/bridge/br_sysfs_if.c index 7ff95ba..dc484ac 100644 --- a/net/bridge/br_sysfs_if.c +++ b/net/bridge/br_sysfs_if.c @@ -172,6 +172,25 @@ static int store_multicast_router(struct net_bridge_port *p, } static BRPORT_ATTR(multicast_router, S_IRUGO | S_IWUSR, show_multicast_router, store_multicast_router); + +static ssize_t show_multicast_fast_leave(struct net_bridge_port *p, + char *buf) +{ + return sprintf(buf, "%d\n", p->multicast_fast_leave); +} + +static int store_multicast_fast_leave(struct net_bridge_port *p, + unsigned long v) +{ + if (p->br->multicast_disabled) + return -EINVAL; + + p->multicast_fast_leave = !!v; + return 0; +} + +static BRPORT_ATTR(multicast_fast_leave, S_IRUGO | S_IWUSR, + show_multicast_fast_leave, store_multicast_fast_leave); #endif static const struct brport_attribute *brport_attrs[] = { @@ -195,6 +214,7 @@ static const struct brport_attribute *brport_attrs[] = { &brport_attr_root_block, #ifdef CONFIG_BRIDGE_IGMP_SNOOPING &brport_attr_multicast_router, + &brport_attr_multicast_fast_leave, #endif NULL };
Stephen Hemminger
2012-Dec-04 16:44 UTC
[Bridge] [PATCH net-next v2] bridge: implement multicast fast leave
On Tue, 4 Dec 2012 17:56:40 +0800 Cong Wang <amwang at redhat.com> wrote:> V2: make the toggle per-port > > Fast leave allows bridge to immediately stops the multicast > traffic on the port receives IGMP Leave when IGMP snooping is enabled, > no timeouts are observed. > > Cc: Herbert Xu <herbert at gondor.apana.org.au> > Cc: Stephen Hemminger <shemminger at vyatta.com> > Cc: "David S. Miller" <davem at davemloft.net> > Signed-off-by: Cong Wang <amwang at redhat.com>Ok. as is. A good followup change would be to migrate these multicast flags into the existing flags bits (see hairpin etc), and make them manageable via netlink.
David Miller
2012-Dec-05 21:25 UTC
[Bridge] [PATCH net-next v3] bridge: implement multicast fast leave
From: Cong Wang <amwang at redhat.com> Date: Wed, 5 Dec 2012 15:19:51 +0800> V3: make it a flag > V2: make the toggle per-port > > Fast leave allows bridge to immediately stops the multicast > traffic on the port receives IGMP Leave when IGMP snooping is enabled, > no timeouts are observed. > > Cc: Herbert Xu <herbert at gondor.apana.org.au> > Cc: Stephen Hemminger <shemminger at vyatta.com> > Cc: "David S. Miller" <davem at davemloft.net> > Signed-off-by: Cong Wang <amwang at redhat.com>Applied.