[ovs-dev,ovn,v4] ovn-northd: Add IGMP Relay support
diff mbox series

Message ID 1565958045-10037-1-git-send-email-dceara@redhat.com
State Accepted
Headers show
Series
  • [ovs-dev,ovn,v4] ovn-northd: Add IGMP Relay support
Related show

Commit Message

Dumitru Ceara Aug. 16, 2019, 12:20 p.m. UTC
Add a new configuration option 'mcast_relay' to the Logical_Router:options
in the OVN Northbound database.

If a router is configured with 'mcast_relay' enabled then ovn-northd
will install Logical_Flows to allow IP multicast traffic to be routed
between Logical_Switches. The logical router will aggregate all IGMP
groups from attached logical switches and modify the routing pipeline in
the following way:
- Table S_ROUTER_IN_IP_INPUT: add flow allowing IP multicast traffic
  if mcast_relay is enabled on the datapath.
- Table S_ROUTER_IN_IP_ROUTING: add flow matching the group address,
  update TTL and set outport="<Multicast_Group> associated with the
  IGMP group". Continue to next table.
- Table S_ROUTER_IN_ARP_RESOLVE: bypass ARP resolve for IP multicast
  traffic and continue to next table.
- Table S_ROUTER_OUT_DELIVERY: add flow matching IP multicast traffic
  and set ETH.SRC to the MAC address of the logical port on which
  traffic is forwarded.

Signed-off-by: Dumitru Ceara <dceara@redhat.com>
Acked-by: Mark Michelson <mmichels@redhat.com>

---
v4:
- Address Numan's comment and use the predicate for ip4.mcast and
  add a new ip4.src_mcast predicate.
v3:
- Address Mark's comment and move setting of the outport in the IP
  Routing stage.
- Update commit message.
- Fix some typos.
v2:
- Optimize flooding to multicast router ports.
- Fix check for source IP multicast in router pipeline.
- Use an enum for OVN_MCAST_*_KEY definitions to avoid hard to debug
  errors due to typos when adding new OVN_MCAST_*_KEY values.
- Fix ovn-northd man page for IGMP.
---
 NEWS                    |   1 +
 lib/logical-fields.c    |   2 +
 lib/mcast-group-index.h |  13 +-
 northd/ovn-northd.8.xml |  79 +++++++-
 northd/ovn-northd.c     | 505 ++++++++++++++++++++++++++++++++++++------------
 ovn-nb.xml              |   6 +
 ovn-sb.xml              |   2 +
 tests/ovn.at            | 199 +++++++++++++++++--
 8 files changed, 656 insertions(+), 151 deletions(-)

Comments

Numan Siddique Aug. 16, 2019, 12:53 p.m. UTC | #1
On Fri, Aug 16, 2019 at 5:51 PM Dumitru Ceara <dceara@redhat.com> wrote:

> Add a new configuration option 'mcast_relay' to the Logical_Router:options
> in the OVN Northbound database.
>
> If a router is configured with 'mcast_relay' enabled then ovn-northd
> will install Logical_Flows to allow IP multicast traffic to be routed
> between Logical_Switches. The logical router will aggregate all IGMP
> groups from attached logical switches and modify the routing pipeline in
> the following way:
> - Table S_ROUTER_IN_IP_INPUT: add flow allowing IP multicast traffic
>   if mcast_relay is enabled on the datapath.
> - Table S_ROUTER_IN_IP_ROUTING: add flow matching the group address,
>   update TTL and set outport="<Multicast_Group> associated with the
>   IGMP group". Continue to next table.
> - Table S_ROUTER_IN_ARP_RESOLVE: bypass ARP resolve for IP multicast
>   traffic and continue to next table.
> - Table S_ROUTER_OUT_DELIVERY: add flow matching IP multicast traffic
>   and set ETH.SRC to the MAC address of the logical port on which
>   traffic is forwarded.
>
> Signed-off-by: Dumitru Ceara <dceara@redhat.com>
> Acked-by: Mark Michelson <mmichels@redhat.com>
>

Acked-by: Numan Siddique <nusiddiq@redhat.com>

Just one comment inline.

Thanks
Numan


>
> ---
> v4:
> - Address Numan's comment and use the predicate for ip4.mcast and
>   add a new ip4.src_mcast predicate.
> v3:
> - Address Mark's comment and move setting of the outport in the IP
>   Routing stage.
> - Update commit message.
> - Fix some typos.
> v2:
> - Optimize flooding to multicast router ports.
> - Fix check for source IP multicast in router pipeline.
> - Use an enum for OVN_MCAST_*_KEY definitions to avoid hard to debug
>   errors due to typos when adding new OVN_MCAST_*_KEY values.
> - Fix ovn-northd man page for IGMP.
> ---
>  NEWS                    |   1 +
>  lib/logical-fields.c    |   2 +
>  lib/mcast-group-index.h |  13 +-
>  northd/ovn-northd.8.xml |  79 +++++++-
>  northd/ovn-northd.c     | 505
> ++++++++++++++++++++++++++++++++++++------------
>  ovn-nb.xml              |   6 +
>  ovn-sb.xml              |   2 +
>  tests/ovn.at            | 199 +++++++++++++++++--
>  8 files changed, 656 insertions(+), 151 deletions(-)
>
> diff --git a/NEWS b/NEWS
> index f476984..73045d6 100644
> --- a/NEWS
> +++ b/NEWS
> @@ -39,6 +39,7 @@ Post-v2.11.0
>         logical groups which results in tunnels only been formed between
>         members of the same transport zone(s).
>       * Support for new logical switch port type - 'virtual'.
> +     * Support for IGMP Snooping/Querier and Relay.
>     - New QoS type "linux-netem" on Linux.
>     - Added support for TLS Server Name Indication (SNI).
>
> diff --git a/lib/logical-fields.c b/lib/logical-fields.c
> index 4ad5bf4..8fb591c 100644
> --- a/lib/logical-fields.c
> +++ b/lib/logical-fields.c
> @@ -156,6 +156,8 @@ ovn_init_symtab(struct shash *symtab)
>
>      expr_symtab_add_field(symtab, "ip4.src", MFF_IPV4_SRC, "ip4", false);
>      expr_symtab_add_field(symtab, "ip4.dst", MFF_IPV4_DST, "ip4", false);
> +    expr_symtab_add_predicate(symtab, "ip4.src_mcast",
> +                              "ip4.src[28..31] == 0xe");
>      expr_symtab_add_predicate(symtab, "ip4.mcast", "ip4.dst[28..31] ==
> 0xe");
>
>      expr_symtab_add_predicate(symtab, "icmp4", "ip4 && ip.proto == 1");
> diff --git a/lib/mcast-group-index.h b/lib/mcast-group-index.h
> index 15a1592..cb49ad7 100644
> --- a/lib/mcast-group-index.h
> +++ b/lib/mcast-group-index.h
> @@ -20,8 +20,17 @@ struct ovsdb_idl;
>
>  struct sbrec_datapath_binding;
>
> -#define OVN_MCAST_FLOOD_TUNNEL_KEY   65535
> -#define OVN_MCAST_UNKNOWN_TUNNEL_KEY (OVN_MCAST_FLOOD_TUNNEL_KEY - 1)
> +#define OVN_MIN_MULTICAST 32768
> +#define OVN_MAX_MULTICAST 65535
> +
> +enum ovn_mcast_tunnel_keys {
> +
> +    OVN_MCAST_FLOOD_TUNNEL_KEY = OVN_MIN_MULTICAST,
> +    OVN_MCAST_UNKNOWN_TUNNEL_KEY,
> +    OVN_MCAST_MROUTER_FLOOD_TUNNEL_KEY,
> +    OVN_MIN_IP_MULTICAST,
> +    OVN_MAX_IP_MULTICAST = OVN_MAX_MULTICAST,
> +};
>
>  struct ovsdb_idl_index *mcast_group_index_create(struct ovsdb_idl *);
>  const struct sbrec_multicast_group *
> diff --git a/northd/ovn-northd.8.xml b/northd/ovn-northd.8.xml
> index 6d2fbe3..d45bb15 100644
> --- a/northd/ovn-northd.8.xml
> +++ b/northd/ovn-northd.8.xml
> @@ -947,10 +947,40 @@ output;
>
>      <ul>
>        <li>
> -        A priority-100 flow that outputs all packets with an Ethernet
> broadcast
> +        A priority-100 flow that punts all IGMP packets to
> +        <code>ovn-controller</code> if IGMP snooping is enabled on the
> +        logical switch.
> +      </li>
> +
> +      <li>
> +        Priority-90 flows that forward registered IP multicast traffic to
> +        their corresponding multicast group, which <code>ovn-northd</code>
> +        creates based on learnt <ref table="IGMP_Group"
> db="OVN_Southbound"/>
> +        entries.  The flows also forward packets to the
> +        <code>MC_MROUTER_FLOOD</code> multicast group, which
> +        <code>ovn-nortdh</code> populates with all the logical ports that
> +        are connected to logical routers with
> +        <ref column="options" table="Logical_Router"/>:mcast_relay='true'.
> +      </li>
> +
> +      <li>
> +        A priority-85 flow that forwards all IP multicast traffic
> destined to
> +        224.0.0.X to the <code>MC_FLOOD</code> multicast group, which
> +        <code>ovn-northd</code> populates with all enabled logical ports.
> +      </li>
> +
> +      <li>
> +        A priority-80 flow that forwards all unregistered IP multicast
> traffic
> +        to the <code>MC_MROUTER_FLOOD</code> multicast group, if any.
> +        Otherwise the flow drops all unregistered IP multicast packets.
> This
> +        flow is added only if <ref column="other_config"
> +        table="Logical_Switch"/>:mcast_flood_unregistered='false'.
> +      </li>
> +
> +      <li>
> +        A priority-70 flow that outputs all packets with an Ethernet
> broadcast
>          or multicast <code>eth.dst</code> to the <code>MC_FLOOD</code>
> -        multicast group, which <code>ovn-northd</code> populates with all
> -        enabled logical ports.
> +        multicast group.
>        </li>
>
>        <li>
> @@ -1228,6 +1258,14 @@ output;
>
>        <li>
>          <p>
> +          A priority-95 flow allows IP multicast traffic if
> +          <ref column="options"
> table="Logical_Router"/>:mcast_relay='true',
> +          otherwise drops it.
> +        </p>
> +      </li>
> +
> +      <li>
> +        <p>
>            ICMP echo reply.  These flows reply to ICMP echo requests
> received
>            for the router's IP address.  Let <var>A</var> be an IP address
>            owned by a router port.  Then, for each <var>A</var> that is
> @@ -1941,6 +1979,16 @@ output;
>      <ul>
>        <li>
>          <p>
> +          Priority-500 flows that match IP multicast traffic destined to
> +          groups registered on any of the attached switches and sets
> +          <code>outport</code> to the associated multicast group that will
> +          eventually flood the traffic to all interested attached logical
> +          switches. The flows also decrement TTL.
> +        </p>
> +      </li>
> +
> +      <li>
> +        <p>
>            For distributed logical routers where one of the logical router
>            ports specifies a <code>redirect-chassis</code>, a priority-400
>            logical flow for each ip source/destination couple that matches
> the
> @@ -2074,6 +2122,15 @@ next;
>      <ul>
>        <li>
>          <p>
> +          A priority-500 flow that matches IP multicast traffic that was
> +          allowed in the routing pipeline. For this kind of traffic the
> +          <code>outport</code> was already set so the flow just advances
> to
> +          the next table.
> +        </p>
> +      </li>
> +
> +      <li>
> +        <p>
>            For distributed logical routers where one of the logical router
>            ports specifies a <code>redirect-chassis</code>, a priority-400
>            logical flow with match <code>REGBIT_DISTRIBUTED_NAT == 1</code>
> @@ -2641,9 +2698,19 @@ clone {
>      <h3>Egress Table 3: Delivery</h3>
>
>      <p>
> -      Packets that reach this table are ready for delivery.  It contains
> -      priority-100 logical flows that match packets on each enabled
> logical
> -      router port, with action <code>output;</code>.
> +      Packets that reach this table are ready for delivery.  It contains:
> +      <ul>
> +        <li>
> +          Priority-110 logical flows that match IP multicast packets on
> each
> +          enabled logical router port and modify the Ethernet source
> address
> +          of the packets to the Ethernet address of the port and then
> execute
> +          action <code>output;</code>.
> +        </li>
> +        <li>
> +          Priority-100 logical flows that match packets on each enabled
> +          logical router port, with action <code>output;</code>.
> +        </li>
> +      </ul>
>      </p>
>
>  </manpage>
> diff --git a/northd/ovn-northd.c b/northd/ovn-northd.c
> index e861344..592d3b9 100644
> --- a/northd/ovn-northd.c
> +++ b/northd/ovn-northd.c
> @@ -433,32 +433,52 @@ struct ipam_info {
>      bool mac_only;
>  };
>
> -#define OVN_MIN_MULTICAST 32768
> -#define OVN_MAX_MULTICAST OVN_MCAST_FLOOD_TUNNEL_KEY
> -BUILD_ASSERT_DECL(OVN_MIN_MULTICAST < OVN_MAX_MULTICAST);
> -
> -#define OVN_MIN_IP_MULTICAST OVN_MIN_MULTICAST
> -#define OVN_MAX_IP_MULTICAST (OVN_MCAST_UNKNOWN_TUNNEL_KEY - 1)
> -BUILD_ASSERT_DECL(OVN_MAX_IP_MULTICAST >= OVN_MIN_MULTICAST);
> -
>  /*
>   * Multicast snooping and querier per datapath configuration.
>   */
> +struct mcast_switch_info {
> +
> +    bool enabled;               /* True if snooping enabled. */
> +    bool querier;               /* True if querier enabled. */
> +    bool flood_unregistered;    /* True if unregistered multicast should
> be
> +                                 * flooded.
> +                                 */
> +    bool flood_relay;           /* True if the switch is connected to a
> +                                 * multicast router and unregistered
> multicast
> +                                 * should be flooded to the mrouter. Only
> +                                 * applicable if flood_unregistered ==
> false.
> +                                 */
> +
> +    int64_t table_size;         /* Max number of IP multicast groups. */
> +    int64_t idle_timeout;       /* Timeout after which an idle group is
> +                                 * flushed.
> +                                 */
> +    int64_t query_interval;     /* Interval between multicast queries. */
> +    char *eth_src;              /* ETH src address of the multicast
> queries. */
> +    char *ipv4_src;             /* IP src address of the multicast
> queries. */
> +    int64_t query_max_response; /* Expected time after which reports
> should
> +                                 * be received for queries that were sent
> out.
> +                                 */
> +
> +    uint32_t active_flows;      /* Current number of active IP multicast
> +                                 * flows.
> +                                 */
> +};
> +
> +struct mcast_router_info {
> +    bool relay; /* True if the router should relay IP multicast. */
> +};
> +
>  struct mcast_info {
> -    bool enabled;
> -    bool querier;
> -    bool flood_unregistered;
> -
> -    int64_t table_size;
> -    int64_t idle_timeout;
> -    int64_t query_interval;
> -    char *eth_src;
> -    char *ipv4_src;
> -    int64_t  query_max_response;
> -
> -    struct hmap group_tnlids;
> -    uint32_t group_tnlid_hint;
> -    uint32_t active_flows;
> +
> +    struct hmap group_tnlids;  /* Group tunnel IDs in use on this DP. */
> +    uint32_t group_tnlid_hint; /* Hint for allocating next group tunnel
> ID. */
> +    struct ovs_list groups;    /* List of groups learnt on this DP. */
> +
> +    union {
> +        struct mcast_switch_info sw;  /* Switch specific multicast info.
> */
> +        struct mcast_router_info rtr; /* Router specific multicast info.
> */
> +    };
>  };
>
>  static uint32_t
> @@ -559,6 +579,7 @@ ovn_datapath_create(struct hmap *datapaths, const
> struct uuid *key,
>  }
>
>  static void ovn_ls_port_group_destroy(struct hmap *nb_pgs);
> +static void destroy_mcast_info_for_datapath(struct ovn_datapath *od);
>
>  static void
>  ovn_datapath_destroy(struct hmap *datapaths, struct ovn_datapath *od)
> @@ -572,12 +593,7 @@ ovn_datapath_destroy(struct hmap *datapaths, struct
> ovn_datapath *od)
>          bitmap_free(od->ipam_info.allocated_ipv4s);
>          free(od->router_ports);
>          ovn_ls_port_group_destroy(&od->nb_pgs);
> -
> -        if (od->nbs) {
> -            free(od->mcast_info.eth_src);
> -            free(od->mcast_info.ipv4_src);
> -            destroy_tnlids(&od->mcast_info.group_tnlids);
> -        }
> +        destroy_mcast_info_for_datapath(od);
>
>          free(od);
>      }
> @@ -714,23 +730,28 @@ init_ipam_info_for_datapath(struct ovn_datapath *od)
>  }
>
>  static void
> -init_mcast_info_for_datapath(struct ovn_datapath *od)
> +init_mcast_info_for_router_datapath(struct ovn_datapath *od)
>  {
> -    if (!od->nbs) {
> -        return;
> -    }
> +    struct mcast_router_info *mcast_rtr_info = &od->mcast_info.rtr;
>
> -    struct mcast_info *mcast_info = &od->mcast_info;
> +    mcast_rtr_info->relay = smap_get_bool(&od->nbr->options,
> "mcast_relay",
> +                                          false);
> +}
>
> -    mcast_info->enabled =
> +static void
> +init_mcast_info_for_switch_datapath(struct ovn_datapath *od)
> +{
> +    struct mcast_switch_info *mcast_sw_info = &od->mcast_info.sw;
> +
> +    mcast_sw_info->enabled =
>          smap_get_bool(&od->nbs->other_config, "mcast_snoop", false);
> -    mcast_info->querier =
> +    mcast_sw_info->querier =
>          smap_get_bool(&od->nbs->other_config, "mcast_querier", true);
> -    mcast_info->flood_unregistered =
> +    mcast_sw_info->flood_unregistered =
>          smap_get_bool(&od->nbs->other_config, "mcast_flood_unregistered",
>                        false);
>
> -    mcast_info->table_size =
> +    mcast_sw_info->table_size =
>          smap_get_ullong(&od->nbs->other_config, "mcast_table_size",
>                          OVN_MCAST_DEFAULT_MAX_ENTRIES);
>
> @@ -742,54 +763,94 @@ init_mcast_info_for_datapath(struct ovn_datapath *od)
>      } else if (idle_timeout > OVN_MCAST_MAX_IDLE_TIMEOUT_S) {
>          idle_timeout = OVN_MCAST_MAX_IDLE_TIMEOUT_S;
>      }
> -    mcast_info->idle_timeout = idle_timeout;
> +    mcast_sw_info->idle_timeout = idle_timeout;
>
>      uint32_t query_interval =
>          smap_get_ullong(&od->nbs->other_config, "mcast_query_interval",
> -                        mcast_info->idle_timeout / 2);
> +                        mcast_sw_info->idle_timeout / 2);
>      if (query_interval < OVN_MCAST_MIN_QUERY_INTERVAL_S) {
>          query_interval = OVN_MCAST_MIN_QUERY_INTERVAL_S;
>      } else if (query_interval > OVN_MCAST_MAX_QUERY_INTERVAL_S) {
>          query_interval = OVN_MCAST_MAX_QUERY_INTERVAL_S;
>      }
> -    mcast_info->query_interval = query_interval;
> +    mcast_sw_info->query_interval = query_interval;
>
> -    mcast_info->eth_src =
> +    mcast_sw_info->eth_src =
>          nullable_xstrdup(smap_get(&od->nbs->other_config,
> "mcast_eth_src"));
> -    mcast_info->ipv4_src =
> +    mcast_sw_info->ipv4_src =
>          nullable_xstrdup(smap_get(&od->nbs->other_config,
> "mcast_ip4_src"));
>
> -    mcast_info->query_max_response =
> +    mcast_sw_info->query_max_response =
>          smap_get_ullong(&od->nbs->other_config,
> "mcast_query_max_response",
>                          OVN_MCAST_DEFAULT_QUERY_MAX_RESPONSE_S);
>
> -    hmap_init(&mcast_info->group_tnlids);
> -    mcast_info->group_tnlid_hint = OVN_MIN_IP_MULTICAST;
> -    mcast_info->active_flows = 0;
> +    mcast_sw_info->active_flows = 0;
> +}
> +
> +static void
> +init_mcast_info_for_datapath(struct ovn_datapath *od)
> +{
> +    if (!od->nbr && !od->nbs) {
> +        return;
> +    }
> +
> +    hmap_init(&od->mcast_info.group_tnlids);
> +    od->mcast_info.group_tnlid_hint = OVN_MIN_IP_MULTICAST;
> +    ovs_list_init(&od->mcast_info.groups);
> +
> +    if (od->nbs) {
> +        init_mcast_info_for_switch_datapath(od);
> +    } else {
> +        init_mcast_info_for_router_datapath(od);
> +    }
> +}
> +
> +static void
> +destroy_mcast_info_for_switch_datapath(struct ovn_datapath *od)
> +{
> +    struct mcast_switch_info *mcast_sw_info = &od->mcast_info.sw;
> +
> +    free(mcast_sw_info->eth_src);
> +    free(mcast_sw_info->ipv4_src);
> +}
> +
> +static void
> +destroy_mcast_info_for_datapath(struct ovn_datapath *od)
> +{
> +    if (!od->nbr && !od->nbs) {
> +        return;
> +    }
> +
> +    if (od->nbs) {
> +        destroy_mcast_info_for_switch_datapath(od);
> +    }
> +
> +    destroy_tnlids(&od->mcast_info.group_tnlids);
>  }
>
>  static void
> -store_mcast_info_for_datapath(const struct sbrec_ip_multicast *sb,
> -                              struct ovn_datapath *od)
> +store_mcast_info_for_switch_datapath(const struct sbrec_ip_multicast *sb,
> +                                     struct ovn_datapath *od)
>  {
> -    struct mcast_info *mcast_info = &od->mcast_info;
> +    struct mcast_switch_info *mcast_sw_info = &od->mcast_info.sw;
>
>      sbrec_ip_multicast_set_datapath(sb, od->sb);
> -    sbrec_ip_multicast_set_enabled(sb, &mcast_info->enabled, 1);
> -    sbrec_ip_multicast_set_querier(sb, &mcast_info->querier, 1);
> -    sbrec_ip_multicast_set_table_size(sb, &mcast_info->table_size, 1);
> -    sbrec_ip_multicast_set_idle_timeout(sb, &mcast_info->idle_timeout, 1);
> +    sbrec_ip_multicast_set_enabled(sb, &mcast_sw_info->enabled, 1);
> +    sbrec_ip_multicast_set_querier(sb, &mcast_sw_info->querier, 1);
> +    sbrec_ip_multicast_set_table_size(sb, &mcast_sw_info->table_size, 1);
> +    sbrec_ip_multicast_set_idle_timeout(sb, &mcast_sw_info->idle_timeout,
> 1);
>      sbrec_ip_multicast_set_query_interval(sb,
> -                                          &mcast_info->query_interval, 1);
> +                                          &mcast_sw_info->query_interval,
> 1);
>      sbrec_ip_multicast_set_query_max_resp(sb,
> -
> &mcast_info->query_max_response, 1);
> +
> &mcast_sw_info->query_max_response,
> +                                          1);
>
> -    if (mcast_info->eth_src) {
> -        sbrec_ip_multicast_set_eth_src(sb, mcast_info->eth_src);
> +    if (mcast_sw_info->eth_src) {
> +        sbrec_ip_multicast_set_eth_src(sb, mcast_sw_info->eth_src);
>      }
>
> -    if (mcast_info->ipv4_src) {
> -        sbrec_ip_multicast_set_ip4_src(sb, mcast_info->ipv4_src);
> +    if (mcast_sw_info->ipv4_src) {
> +        sbrec_ip_multicast_set_ip4_src(sb, mcast_sw_info->ipv4_src);
>      }
>  }
>
> @@ -906,6 +967,7 @@ join_datapaths(struct northd_context *ctx, struct hmap
> *datapaths,
>                                       NULL, nbr, NULL);
>              ovs_list_push_back(nb_only, &od->list);
>          }
> +        init_mcast_info_for_datapath(od);
>          ovs_list_push_back(lr_list, &od->lr_list);
>      }
>  }
> @@ -1999,6 +2061,13 @@ join_logical_ports(struct northd_context *ctx,
>                      break;
>                  }
>              }
> +
> +            /* If the router is multicast enabled then set relay on the
> switch
> +             * datapath.
> +             */
> +            if (peer->od && peer->od->mcast_info.rtr.relay) {
> +                op->od->mcast_info.sw.flood_relay = true;
> +            }
>          } else if (op->nbrp && op->nbrp->peer && !op->derived) {
>              struct ovn_port *peer = ovn_port_find(ports, op->nbrp->peer);
>              if (peer) {
> @@ -2846,6 +2915,10 @@ struct multicast_group {
>  static const struct multicast_group mc_flood =
>      { MC_FLOOD, OVN_MCAST_FLOOD_TUNNEL_KEY };
>
> +#define MC_MROUTER_FLOOD "_MC_mrouter_flood"
> +static const struct multicast_group mc_mrouter_flood =
> +    { MC_MROUTER_FLOOD, OVN_MCAST_MROUTER_FLOOD_TUNNEL_KEY };
> +
>  #define MC_UNKNOWN "_MC_unknown"
>  static const struct multicast_group mc_unknown =
>      { MC_UNKNOWN, OVN_MCAST_UNKNOWN_TUNNEL_KEY };
> @@ -2955,7 +3028,8 @@ ovn_multicast_update_sbrec(const struct
> ovn_multicast *mc,
>   */
>  struct ovn_igmp_group_entry {
>      struct ovs_list list_node; /* Linkage in the list of entries. */
> -    const struct sbrec_igmp_group *sb;
> +    size_t n_ports;
> +    struct ovn_port **ports;
>  };
>
>  /*
> @@ -2964,12 +3038,13 @@ struct ovn_igmp_group_entry {
>   */
>  struct ovn_igmp_group {
>      struct hmap_node hmap_node; /* Index on 'datapath' and 'address'. */
> +    struct ovs_list list_node;  /* Linkage in the per-dp igmp group list.
> */
>
>      struct ovn_datapath *datapath;
>      struct in6_addr address; /* Multicast IPv6-mapped-IPv4 or IPv4
> address. */
>      struct multicast_group mcgroup;
>
> -    struct ovs_list sb_entries; /* List of SB entries for this group. */
> +    struct ovs_list entries; /* List of SB entries for this group. */
>  };
>
>  static uint32_t
> @@ -2997,77 +3072,120 @@ ovn_igmp_group_find(struct hmap *igmp_groups,
>      return NULL;
>  }
>
> -static void
> +static struct ovn_igmp_group *
>  ovn_igmp_group_add(struct northd_context *ctx, struct hmap *igmp_groups,
>                     struct ovn_datapath *datapath,
> -                   const struct sbrec_igmp_group *sb_igmp_group)
> +                   const struct in6_addr *address,
> +                   const char *address_s)
>  {
> -    struct in6_addr group_address;
> -    ovs_be32 ipv4;
> -
> -    if (ip_parse(sb_igmp_group->address, &ipv4)) {
> -        group_address = in6_addr_mapped_ipv4(ipv4);
> -    } else if (!ipv6_parse(sb_igmp_group->address, &group_address)) {
> -        static struct vlog_rate_limit rl = VLOG_RATE_LIMIT_INIT(1, 1);
> -        VLOG_WARN_RL(&rl, "invalid IGMP group address: %s",
> -                     sb_igmp_group->address);
> -        return;
> -    }
> -
>      struct ovn_igmp_group *igmp_group =
> -        ovn_igmp_group_find(igmp_groups, datapath, &group_address);
> +        ovn_igmp_group_find(igmp_groups, datapath, address);
>
>      if (!igmp_group) {
>          igmp_group = xmalloc(sizeof *igmp_group);
>
>          const struct sbrec_multicast_group *mcgroup =
> -            mcast_group_lookup(ctx->sbrec_mcast_group_by_name_dp,
> -                               sb_igmp_group->address, datapath->sb);
> +            mcast_group_lookup(ctx->sbrec_mcast_group_by_name_dp,
> address_s,
> +                               datapath->sb);
>
>          igmp_group->datapath = datapath;
> -        igmp_group->address = group_address;
> +        igmp_group->address = *address;
>          if (mcgroup) {
>              igmp_group->mcgroup.key = mcgroup->tunnel_key;
>              add_tnlid(&datapath->mcast_info.group_tnlids,
> mcgroup->tunnel_key);
>          } else {
>              igmp_group->mcgroup.key = 0;
>          }
> -        igmp_group->mcgroup.name = sb_igmp_group->address;
> -        ovs_list_init(&igmp_group->sb_entries);
> +        igmp_group->mcgroup.name = address_s;
> +        ovs_list_init(&igmp_group->entries);
>
>          hmap_insert(igmp_groups, &igmp_group->hmap_node,
> -                    ovn_igmp_group_hash(datapath, &group_address));
> +                    ovn_igmp_group_hash(datapath, address));
> +        ovs_list_push_back(&datapath->mcast_info.groups,
> +                           &igmp_group->list_node);
> +    }
> +
> +    return igmp_group;
> +}
> +
> +static bool
> +ovn_igmp_group_get_address(const struct sbrec_igmp_group *sb_igmp_group,
> +                           struct in6_addr *address)
> +{
> +    ovs_be32 ipv4;
> +
> +    if (ip_parse(sb_igmp_group->address, &ipv4)) {
> +        *address = in6_addr_mapped_ipv4(ipv4);
> +        return true;
> +    }
> +    if (!ipv6_parse(sb_igmp_group->address, address)) {
> +        return false;
>      }
> +    return true;
> +}
>
> +static struct ovn_port **
> +ovn_igmp_group_get_ports(const struct sbrec_igmp_group *sb_igmp_group,
> +                         size_t *n_ports, struct hmap *ovn_ports)
> +{
> +    struct ovn_port **ports = xmalloc(sb_igmp_group->n_ports * sizeof
> *ports);
> +
> +     *n_ports = 0;
> +     for (size_t i = 0; i < sb_igmp_group->n_ports; i++) {
> +        ports[(*n_ports)] =
> +            ovn_port_find(ovn_ports,
> sb_igmp_group->ports[i]->logical_port);
> +        if (ports[(*n_ports)]) {
> +            (*n_ports)++;
> +        }
> +    }
> +
> +    return ports;
> +}
> +
> +static void
> +ovn_igmp_group_add_entry(struct ovn_igmp_group *igmp_group,
> +                         struct ovn_port **ports, size_t n_ports)
> +{
>      struct ovn_igmp_group_entry *entry = xmalloc(sizeof *entry);
>
> -    entry->sb = sb_igmp_group;
> -    ovs_list_push_back(&igmp_group->sb_entries , &entry->list_node);
> +    entry->ports = ports;
> +    entry->n_ports = n_ports;
> +    ovs_list_push_back(&igmp_group->entries, &entry->list_node);
> +}
> +
> +static void
> +ovn_igmp_group_destroy_entry(struct ovn_igmp_group_entry *entry)
> +{
> +    free(entry->ports);
> +}
> +
> +static bool
> +ovn_igmp_group_allocate_id(struct ovn_igmp_group *igmp_group)
> +{
> +    if (igmp_group->mcgroup.key == 0) {
> +        struct mcast_info *mcast_info = &igmp_group->datapath->mcast_info;
> +        igmp_group->mcgroup.key =
> ovn_mcast_group_allocate_key(mcast_info);
> +    }
> +
> +    if (igmp_group->mcgroup.key == 0) {
> +        return false;
> +    }
> +
> +    return true;
>  }
>
>  static void
>  ovn_igmp_group_aggregate_ports(struct ovn_igmp_group *igmp_group,
> -                               struct hmap *ovn_ports,
>                                 struct hmap *mcast_groups)
>  {
>      struct ovn_igmp_group_entry *entry;
>
> -    LIST_FOR_EACH_POP (entry, list_node, &igmp_group->sb_entries) {
> -        size_t n_oports = 0;
> -        struct ovn_port **oports =
> -            xmalloc(entry->sb->n_ports * sizeof *oports);
> -
> -        for (size_t i = 0; i < entry->sb->n_ports; i++) {
> -            oports[n_oports] =
> -                ovn_port_find(ovn_ports,
> entry->sb->ports[i]->logical_port);
> -            if (oports[n_oports]) {
> -                n_oports++;
> -            }
> -        }
> -
> +    LIST_FOR_EACH_POP (entry, list_node, &igmp_group->entries) {
>          ovn_multicast_add_ports(mcast_groups, igmp_group->datapath,
> -                                &igmp_group->mcgroup, oports, n_oports);
> -        free(oports);
> +                                &igmp_group->mcgroup, entry->ports,
> +                                entry->n_ports);
> +
> +        ovn_igmp_group_destroy_entry(entry);
>          free(entry);
>      }
>  }
> @@ -3079,10 +3197,12 @@ ovn_igmp_group_destroy(struct hmap *igmp_groups,
>      if (igmp_group) {
>          struct ovn_igmp_group_entry *entry;
>
> -        LIST_FOR_EACH_POP (entry, list_node, &igmp_group->sb_entries) {
> +        LIST_FOR_EACH_POP (entry, list_node, &igmp_group->entries) {
> +            ovn_igmp_group_destroy_entry(entry);
>              free(entry);
>          }
>          hmap_remove(igmp_groups, &igmp_group->hmap_node);
> +        ovs_list_remove(&igmp_group->list_node);
>          free(igmp_group);
>      }
>  }
> @@ -5286,7 +5406,9 @@ build_lswitch_flows(struct hmap *datapaths, struct
> hmap *ports,
>              continue;
>          }
>
> -        if (od->mcast_info.enabled) {
> +        struct mcast_switch_info *mcast_sw_info = &od->mcast_info.sw;
> +
> +        if (mcast_sw_info->enabled) {
>              /* Punt IGMP traffic to controller. */
>              ovn_lflow_add(lflows, od, S_SWITCH_IN_L2_LKUP, 100,
>                            "ip4 && ip.proto == 2", "igmp;");
> @@ -5299,9 +5421,16 @@ build_lswitch_flows(struct hmap *datapaths, struct
> hmap *ports,
>                            "outport = \""MC_FLOOD"\"; output;");
>
>              /* Drop unregistered IP multicast if not allowed. */
> -            if (!od->mcast_info.flood_unregistered) {
> -                ovn_lflow_add(lflows, od, S_SWITCH_IN_L2_LKUP, 80,
> -                              "ip4 && ip4.mcast", "drop;");
> +            if (!mcast_sw_info->flood_unregistered) {
> +                /* Forward unregistered IP multicast to mrouter (if any).
> */
> +                if (mcast_sw_info->flood_relay) {
> +                    ovn_lflow_add(lflows, od, S_SWITCH_IN_L2_LKUP, 80,
> +                                  "ip4 && ip4.mcast",
> +                                  "outport = \""MC_MROUTER_FLOOD"\";
> output;");
> +                } else {
> +                    ovn_lflow_add(lflows, od, S_SWITCH_IN_L2_LKUP, 80,
> +                                  "ip4 && ip4.mcast", "drop;");
> +                }
>              }
>          }
>
> @@ -5318,18 +5447,26 @@ build_lswitch_flows(struct hmap *datapaths, struct
> hmap *ports,
>              continue;
>          }
>
> -        struct mcast_info *mcast_info = &igmp_group->datapath->mcast_info;
> +        struct mcast_switch_info *mcast_sw_info =
> +            &igmp_group->datapath->mcast_info.sw;
>
> -        if (mcast_info->active_flows >= mcast_info->table_size) {
> +        if (mcast_sw_info->active_flows >= mcast_sw_info->table_size) {
>              continue;
>          }
> -        mcast_info->active_flows++;
> +        mcast_sw_info->active_flows++;
>
>          ds_clear(&match);
>          ds_clear(&actions);
>
>          ds_put_format(&match, "eth.mcast && ip4 && ip4.dst == %s ",
>                        igmp_group->mcgroup.name);
> +        /* Also flood traffic to all multicast routers with relay
> enabled. */
> +        if (mcast_sw_info->flood_relay) {
> +            ds_put_cstr(&actions,
> +                        "clone { "
> +                            "outport = \""MC_MROUTER_FLOOD "\"; output; "
> +                        "};");
> +        }
>          ds_put_format(&actions, "outport = \"%s\"; output; ",
>                        igmp_group->mcgroup.name);
>
> @@ -6209,7 +6346,7 @@ build_lrouter_flows(struct hmap *datapaths, struct
> hmap *ports,
>           * source or destination, and zero network source or destination
>           * (priority 100). */
>          ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_INPUT, 100,
> -                      "ip4.mcast || "
> +                      "ip4.src_mcast ||"
>                        "ip4.src == 255.255.255.255 || "
>                        "ip4.src == 127.0.0.0/8 || "
>                        "ip4.dst == 127.0.0.0/8 || "
> @@ -6217,6 +6354,16 @@ build_lrouter_flows(struct hmap *datapaths, struct
> hmap *ports,
>                        "ip4.dst == 0.0.0.0/8",
>                        "drop;");
>
> +        /* Allow multicast if relay enabled (priority 95). */
> +        ds_clear(&actions);
> +        if (od->mcast_info.rtr.relay) {
> +            ds_put_cstr(&actions, "next;");
> +        } else {
> +            ds_put_cstr(&actions, "drop;");
> +        }
> +        ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_INPUT, 95,
> +                      "ip4.mcast", ds_cstr(&actions));
> +
>

Instead of if/else how about
ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_INPUT, 95, "ip4.mcast",
od->mcast_info.rtr.relay ? "next": "drop");


         /* ARP reply handling.  Use ARP replies to populate the logical
>           * router's ARP table. */
>          ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_INPUT, 90, "arp.op == 2",
> @@ -7487,6 +7634,27 @@ build_lrouter_flows(struct hmap *datapaths, struct
> hmap *ports,
>          }
>      }
>
> +    /* IP Multicast lookup. Here we set the output port, adjust TTL and
> +     * advance to next table (priority 500).
> +     */
> +    HMAP_FOR_EACH (od, key_node, datapaths) {
> +        if (!od->nbr || !od->mcast_info.rtr.relay) {
> +            continue;
> +        }
> +        struct ovn_igmp_group *igmp_group;
> +
> +        LIST_FOR_EACH (igmp_group, list_node, &od->mcast_info.groups) {
> +            ds_clear(&match);
> +            ds_clear(&actions);
> +            ds_put_format(&match, "ip4 && ip4.dst == %s ",
> +                          igmp_group->mcgroup.name);
> +            ds_put_format(&actions, "outport = \"%s\"; ip.ttl--; next;",
> +                          igmp_group->mcgroup.name);
> +            ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_ROUTING, 500,
> +                          ds_cstr(&match), ds_cstr(&actions));
> +        }
> +    }
> +
>      /* Logical router ingress table 8: Policy.
>       *
>       * A packet that arrives at this table is an IP packet that should be
> @@ -7517,10 +7685,24 @@ build_lrouter_flows(struct hmap *datapaths, struct
> hmap *ports,
>
>      /* Local router ingress table 9: ARP Resolution.
>       *
> -     * Any packet that reaches this table is an IP packet whose next-hop
> IP
> -     * address is in reg0. (ip4.dst is the final destination.) This table
> -     * resolves the IP address in reg0 into an output port in outport and
> an
> -     * Ethernet address in eth.dst. */
> +     * Multicast packets already have the outport set so just advance to
> next
> +     * table (priority 500). */
> +    HMAP_FOR_EACH (od, key_node, datapaths) {
> +        if (!od->nbr) {
> +            continue;
> +        }
> +
> +        ovn_lflow_add(lflows, od, S_ROUTER_IN_ARP_RESOLVE, 500,
> +                      "ip4.mcast", "next;");
> +    }
> +
> +    /* Local router ingress table 9: ARP Resolution.
> +     *
> +     * Any unicast packet that reaches this table is an IP packet whose
> +     * next-hop IP address is in reg0. (ip4.dst is the final destination.)
> +     * This table resolves the IP address in reg0 into an output port in
> +     * outport and an Ethernet address in eth.dst.
> +     */
>      HMAP_FOR_EACH (op, key_node, ports) {
>          if (op->nbsp && !lsp_is_enabled(op->nbsp)) {
>              continue;
> @@ -8002,9 +8184,13 @@ build_lrouter_flows(struct hmap *datapaths, struct
> hmap *ports,
>          ovn_lflow_add(lflows, od, S_ROUTER_IN_ARP_REQUEST, 0, "1",
> "output;");
>      }
>
> -    /* Logical router egress table 1: Delivery (priority 100).
> +    /* Logical router egress table 1: Delivery (priority 100-110).
>       *
> -     * Priority 100 rules deliver packets to enabled logical ports. */
> +     * Priority 100 rules deliver packets to enabled logical ports.
> +     * Priority 110 rules match multicast packets and update the source
> +     * mac before delivering to enabled logical ports. IP multicast
> traffic
> +     * bypasses S_ROUTER_IN_IP_ROUTING route lookups.
> +     */
>      HMAP_FOR_EACH (op, key_node, ports) {
>          if (!op->nbrp) {
>              continue;
> @@ -8024,6 +8210,20 @@ build_lrouter_flows(struct hmap *datapaths, struct
> hmap *ports,
>              continue;
>          }
>
> +        /* If multicast relay is enabled then also adjust source mac for
> IP
> +         * multicast traffic.
> +         */
> +        if (op->od->mcast_info.rtr.relay) {
> +            ds_clear(&match);
> +            ds_clear(&actions);
> +            ds_put_format(&match, "ip4.mcast && outport == %s",
> +                          op->json_key);
> +            ds_put_format(&actions, "eth.src = %s; output;",
> +                          op->lrp_networks.ea_s);
> +            ovn_lflow_add(lflows, op->od, S_ROUTER_OUT_DELIVERY, 110,
> +                        ds_cstr(&match), ds_cstr(&actions));
> +        }
> +
>          ds_clear(&match);
>          ds_put_format(&match, "outport == %s", op->json_key);
>          ovn_lflow_add(lflows, op->od, S_ROUTER_OUT_DELIVERY, 100,
> @@ -8574,7 +8774,7 @@ build_ip_mcast(struct northd_context *ctx, struct
> hmap *datapaths)
>          if (!ip_mcast) {
>              ip_mcast = sbrec_ip_multicast_insert(ctx->ovnsb_txn);
>          }
> -        store_mcast_info_for_datapath(ip_mcast, od);
> +        store_mcast_info_for_switch_datapath(ip_mcast, od);
>      }
>
>      /* Delete southbound records without northbound matches. */
> @@ -8606,6 +8806,14 @@ build_mcast_groups(struct northd_context *ctx,
>
>          if (lsp_is_enabled(op->nbsp)) {
>              ovn_multicast_add(mcast_groups, &mc_flood, op);
> +
> +            /* If this port is connected to a multicast router then add it
> +             * to the MC_MROUTER_FLOOD group.
> +             */
> +            if (op->od->mcast_info.sw.flood_relay && op->peer &&
> +                    op->peer->od && op->peer->od->mcast_info.rtr.relay) {
> +                ovn_multicast_add(mcast_groups, &mc_mrouter_flood, op);
> +            }
>          }
>      }
>
> @@ -8628,10 +8836,61 @@ build_mcast_groups(struct northd_context *ctx,
>              continue;
>          }
>
> +        struct in6_addr group_address;
> +        if (!ovn_igmp_group_get_address(sb_igmp, &group_address)) {
> +            static struct vlog_rate_limit rl = VLOG_RATE_LIMIT_INIT(1, 1);
> +            VLOG_WARN_RL(&rl, "invalid IGMP group address: %s",
> +                         sb_igmp->address);
> +            continue;
> +        }
> +
>          /* Add the IGMP group entry. Will also try to allocate an ID for
> it
>           * if the multicast group already exists.
>           */
> -        ovn_igmp_group_add(ctx, igmp_groups, od, sb_igmp);
> +        struct ovn_igmp_group *igmp_group =
> +            ovn_igmp_group_add(ctx, igmp_groups, od, &group_address,
> +                               sb_igmp->address);
> +
> +        /* Extract the IGMP group ports from the SB entry and store them
> +         * in the IGMP group.
> +         */
> +        size_t n_igmp_ports;
> +        struct ovn_port **igmp_ports =
> +            ovn_igmp_group_get_ports(sb_igmp, &n_igmp_ports, ports);
> +        ovn_igmp_group_add_entry(igmp_group, igmp_ports, n_igmp_ports);
> +    }
> +
> +    /* Build IGMP groups for multicast routers with relay enabled. The
> router
> +     * IGMP groups are based on the groups learnt by their multicast
> enabled
> +     * peers.
> +     */
> +    struct ovn_datapath *od;
> +    HMAP_FOR_EACH (od, key_node, datapaths) {
> +
> +        if (ovs_list_is_empty(&od->mcast_info.groups)) {
> +            continue;
> +        }
> +
> +        for (size_t i = 0; i < od->n_router_ports; i++) {
> +            struct ovn_port *router_port = od->router_ports[i]->peer;
> +
> +            if (!router_port || !router_port->od ||
> +                    !router_port->od->mcast_info.rtr.relay) {
> +                continue;
> +            }
> +
> +            struct ovn_igmp_group *igmp_group;
> +            LIST_FOR_EACH (igmp_group, list_node, &od->mcast_info.groups)
> {
> +                struct ovn_igmp_group *igmp_group_rtr =
> +                    ovn_igmp_group_add(ctx, igmp_groups, router_port->od,
> +                                       &igmp_group->address,
> +                                       igmp_group->mcgroup.name);
> +                struct ovn_port **router_igmp_ports =
> +                    xmalloc(sizeof *router_igmp_ports);
> +                router_igmp_ports[0] = router_port;
> +                ovn_igmp_group_add_entry(igmp_group_rtr,
> router_igmp_ports, 1);
> +            }
> +        }
>      }
>
>      /* Walk the aggregated IGMP groups and allocate IDs for new entries.
> @@ -8639,21 +8898,17 @@ build_mcast_groups(struct northd_context *ctx,
>       */
>      struct ovn_igmp_group *igmp_group, *igmp_group_next;
>      HMAP_FOR_EACH_SAFE (igmp_group, igmp_group_next, hmap_node,
> igmp_groups) {
> -        if (igmp_group->mcgroup.key == 0) {
> -            struct mcast_info *mcast_info =
> &igmp_group->datapath->mcast_info;
> -            igmp_group->mcgroup.key =
> ovn_mcast_group_allocate_key(mcast_info);
> -        }
>
> -        /* If we ran out of keys just destroy the entry. */
> -        if (igmp_group->mcgroup.key == 0) {
> +        if (!ovn_igmp_group_allocate_id(igmp_group)) {
> +            /* If we ran out of keys just destroy the entry. */
>              ovn_igmp_group_destroy(igmp_groups, igmp_group);
>              continue;
>          }
>
> -        /* Aggregate the ports from all SB entries corresponding to this
> +        /* Aggregate the ports from all entries corresponding to this
>           * group.
>           */
> -        ovn_igmp_group_aggregate_ports(igmp_group, ports, mcast_groups);
> +        ovn_igmp_group_aggregate_ports(igmp_group, mcast_groups);
>      }
>  }
>
> diff --git a/ovn-nb.xml b/ovn-nb.xml
> index e166190..1f8d751 100644
> --- a/ovn-nb.xml
> +++ b/ovn-nb.xml
> @@ -1527,6 +1527,12 @@
>            address.
>          </p>
>        </column>
> +      <column name="options" key="mcast_relay" type'{"type": "boolean"}'>
> +        <p>
> +          Enables/disables IP multicast relay between logical switches
> +          connected to the logical router. Default: False.
> +        </p>
> +      </column>
>      </group>
>
>      <group title="Common Columns">
> diff --git a/ovn-sb.xml b/ovn-sb.xml
> index 17c45bb..02691bb 100644
> --- a/ovn-sb.xml
> +++ b/ovn-sb.xml
> @@ -1017,6 +1017,8 @@
>          <li><code>eth.mcast</code> expands to
> <code>eth.dst[40]</code></li>
>          <li><code>vlan.present</code> expands to
> <code>vlan.tci[12]</code></li>
>          <li><code>ip4</code> expands to <code>eth.type ==
> 0x800</code></li>
> +        <li><code>ip4.src_mcast</code> expands to
> +        <code>ip4.src[28..31] == 0xe</code></li>
>          <li><code>ip4.mcast</code> expands to <code>ip4.dst[28..31] ==
> 0xe</code></li>
>          <li><code>ip6</code> expands to <code>eth.type ==
> 0x86dd</code></li>
>          <li><code>ip</code> expands to <code>ip4 || ip6</code></li>
> diff --git a/tests/ovn.at b/tests/ovn.at
> index 71eb390..52c044c 100644
> --- a/tests/ovn.at
> +++ b/tests/ovn.at
> @@ -14721,12 +14721,12 @@ AT_CHECK([ovn-sbctl get controller_event $uuid
> seq_num], [0], [dnl
>  OVN_CLEANUP([hv1], [hv2])
>  AT_CLEANUP
>
> -AT_SETUP([ovn -- IGMP snoop/querier])
> +AT_SETUP([ovn -- IGMP snoop/querier/relay])
>  AT_SKIP_IF([test $HAVE_PYTHON = no])
>  ovn_start
>
>  # Logical network:
> -# Two independent logical switches (sw1 and sw2).
> +# Three logical switches (sw1-sw3) connected to a logical router (rtr).
>  # sw1:
>  #   - subnet 10.0.0.0/8
>  #   - 2 ports bound on hv1 (sw1-p11, sw1-p12)
> @@ -14736,6 +14736,10 @@ ovn_start
>  #   - 1 port bound on hv1 (sw2-p1)
>  #   - 1 port bound on hv2 (sw2-p2)
>  #   - IGMP Querier from 20.0.0.254
> +# sw3:
> +#   - subnet 30.0.0.0/8
> +#   - 1 port bound on hv1 (sw3-p1)
> +#   - 1 port bound on hv2 (sw3-p2)
>
>  reset_pcap_file() {
>      local iface=$1
> @@ -14812,29 +14816,47 @@ store_igmp_v3_query() {
>  }
>
>  #
> -# send_ip_multicast_pkt INPORT HV ETH_SRC ETH_DST IP_SRC IP_DST IP_LEN
> -#    IP_PROTO DATA OUTFILE
> +# send_ip_multicast_pkt INPORT HV ETH_SRC ETH_DST IP_SRC IP_DST IP_LEN TTL
> +#    IP_CHKSUM IP_PROTO DATA
>  #
>  # This shell function causes an IP multicast packet to be received on
> INPORT
>  # of HV.
>  # The hexdump of the packet is stored in OUTFILE.
>  #
>  send_ip_multicast_pkt() {
> -    local inport=$1 hv=$2 eth_src=$3 eth_dst=$4 ip_src=$5 ip_dst=$6
> -    local ip_len=$7 ip_chksum=$8 proto=$9 data=${10} outfile=${11}
> -
> -    local ip_ttl=20
> +    local inport=$1 hv=$2 eth_src=$3 eth_dst=$4
> +    local ip_src=$5 ip_dst=$6 ip_len=$7 ip_ttl=$8 ip_chksum=$9 proto=${10}
> +    local data=${11}
>
>      local eth=${eth_dst}${eth_src}0800
>      local
> ip=450000${ip_len}95f14000${ip_ttl}${proto}${ip_chksum}${ip_src}${ip_dst}
>      local packet=${eth}${ip}${data}
>
>      as $hv ovs-appctl netdev-dummy/receive ${inport} ${packet}
> +}
> +
> +#
> +# store_ip_multicast_pkt ETH_SRC ETH_DST IP_SRC IP_DST IP_LEN TTL
> +#    IP_CHKSUM IP_PROTO DATA OUTFILE
> +#
> +# This shell function builds an IP multicast packet and stores the
> hexdump of
> +# the packet in OUTFILE.
> +#
> +store_ip_multicast_pkt() {
> +    local eth_src=$1 eth_dst=$2
> +    local ip_src=$3 ip_dst=$4 ip_len=$5 ip_ttl=$6 ip_chksum=$7 proto=$8
> +    local data=$9 outfile=${10}
> +
> +    local eth=${eth_dst}${eth_src}0800
> +    local
> ip=450000${ip_len}95f14000${ip_ttl}${proto}${ip_chksum}${ip_src}${ip_dst}
> +    local packet=${eth}${ip}${data}
> +
>      echo ${packet} >> ${outfile}
>  }
>
>  ovn-nbctl ls-add sw1
>  ovn-nbctl ls-add sw2
> +ovn-nbctl ls-add sw3
>
>  ovn-nbctl lsp-add sw1 sw1-p11
>  ovn-nbctl lsp-add sw1 sw1-p12
> @@ -14842,6 +14864,26 @@ ovn-nbctl lsp-add sw1 sw1-p21
>  ovn-nbctl lsp-add sw1 sw1-p22
>  ovn-nbctl lsp-add sw2 sw2-p1
>  ovn-nbctl lsp-add sw2 sw2-p2
> +ovn-nbctl lsp-add sw3 sw3-p1
> +ovn-nbctl lsp-add sw3 sw3-p2
> +
> +ovn-nbctl lr-add rtr
> +ovn-nbctl lrp-add rtr rtr-sw1 00:00:00:00:01:00 10.0.0.254/24
> +ovn-nbctl lrp-add rtr rtr-sw2 00:00:00:00:02:00 20.0.0.254/24
> +ovn-nbctl lrp-add rtr rtr-sw3 00:00:00:00:03:00 30.0.0.254/24
> +
> +ovn-nbctl lsp-add sw1 sw1-rtr                      \
> +    -- lsp-set-type sw1-rtr router                 \
> +    -- lsp-set-addresses sw1-rtr 00:00:00:00:01:00 \
> +    -- lsp-set-options sw1-rtr router-port=rtr-sw1
> +ovn-nbctl lsp-add sw2 sw2-rtr                      \
> +    -- lsp-set-type sw2-rtr router                 \
> +    -- lsp-set-addresses sw2-rtr 00:00:00:00:02:00 \
> +    -- lsp-set-options sw2-rtr router-port=rtr-sw2
> +ovn-nbctl lsp-add sw3 sw3-rtr                      \
> +    -- lsp-set-type sw3-rtr router                 \
> +    -- lsp-set-addresses sw3-rtr 00:00:00:00:03:00 \
> +    -- lsp-set-options sw3-rtr router-port=rtr-sw3
>
>  net_add n1
>  sim_add hv1
> @@ -14863,6 +14905,11 @@ ovs-vsctl -- add-port br-int hv1-vif3 -- \
>      options:tx_pcap=hv1/vif3-tx.pcap \
>      options:rxq_pcap=hv1/vif3-rx.pcap \
>      ofport-request=1
> +ovs-vsctl -- add-port br-int hv1-vif4 -- \
> +    set interface hv1-vif4 external-ids:iface-id=sw3-p1 \
> +    options:tx_pcap=hv1/vif4-tx.pcap \
> +    options:rxq_pcap=hv1/vif4-rx.pcap \
> +    ofport-request=1
>
>  sim_add hv2
>  as hv2
> @@ -14883,12 +14930,18 @@ ovs-vsctl -- add-port br-int hv2-vif3 -- \
>      options:tx_pcap=hv2/vif3-tx.pcap \
>      options:rxq_pcap=hv2/vif3-rx.pcap \
>      ofport-request=1
> +ovs-vsctl -- add-port br-int hv2-vif4 -- \
> +    set interface hv2-vif4 external-ids:iface-id=sw3-p2 \
> +    options:tx_pcap=hv2/vif4-tx.pcap \
> +    options:rxq_pcap=hv2/vif4-rx.pcap \
> +    ofport-request=1
>
>  OVN_POPULATE_ARP
>
>  # Enable IGMP snooping on sw1.
> -ovn-nbctl set Logical_Switch sw1 other_config:mcast_querier="false"
> -ovn-nbctl set Logical_Switch sw1 other_config:mcast_snoop="true"
> +ovn-nbctl set Logical_Switch sw1       \
> +    other_config:mcast_querier="false" \
> +    other_config:mcast_snoop="true"
>
>  # No IGMP query should be generated by sw1 (mcast_querier="false").
>  truncate -s 0 expected
> @@ -14921,9 +14974,12 @@ truncate -s 0 expected
>  truncate -s 0 expected_empty
>  send_ip_multicast_pkt hv1-vif2 hv1 \
>      000000000001 01005e000144 \
> -    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e ca70 11 \
> -    e518e518000a3b3a0000 \
> -    expected
> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
> +    e518e518000a3b3a0000
> +store_ip_multicast_pkt \
> +    000000000001 01005e000144 \
> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
> +    e518e518000a3b3a0000 expected
>
>  OVN_CHECK_PACKETS([hv1/vif1-tx.pcap], [expected])
>  OVN_CHECK_PACKETS([hv2/vif1-tx.pcap], [expected])
> @@ -14944,17 +15000,19 @@ OVS_WAIT_UNTIL([
>      test "${total_entries}" = "1"
>  ])
>
> -# Send traffic traffic and make sure it gets forwarded only on the port
> that
> -# joined.
> +# Send traffic and make sure it gets forwarded only on the port that
> joined.
>  as hv1 reset_pcap_file hv1-vif1 hv1/vif1
>  as hv2 reset_pcap_file hv2-vif1 hv2/vif1
>  truncate -s 0 expected
>  truncate -s 0 expected_empty
>  send_ip_multicast_pkt hv1-vif2 hv1 \
>      000000000001 01005e000144 \
> -    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e ca70 11 \
> -    e518e518000a3b3a0000 \
> -    expected
> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
> +    e518e518000a3b3a0000
> +store_ip_multicast_pkt \
> +    000000000001 01005e000144 \
> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
> +    e518e518000a3b3a0000 expected
>
>  OVN_CHECK_PACKETS([hv1/vif1-tx.pcap], [expected_empty])
>  OVN_CHECK_PACKETS([hv2/vif1-tx.pcap], [expected])
> @@ -14988,6 +15046,111 @@ sleep 1
>  OVN_CHECK_PACKETS([hv1/vif3-tx.pcap], [expected])
>  OVN_CHECK_PACKETS([hv2/vif3-tx.pcap], [expected])
>
> +# Dissable IGMP querier on sw2.
> +ovn-nbctl set Logical_Switch sw2 \
> +    other_config:mcast_querier="false"
> +
> +# Enable IGMP snooping on sw3.
> +ovn-nbctl set Logical_Switch sw3       \
> +    other_config:mcast_querier="false" \
> +    other_config:mcast_snoop="true"
> +
> +# Send traffic from sw3 and make sure rtr doesn't relay it.
> +truncate -s 0 expected_empty
> +
> +as hv1 reset_pcap_file hv1-vif1 hv1/vif1
> +as hv1 reset_pcap_file hv1-vif2 hv1/vif2
> +as hv1 reset_pcap_file hv1-vif3 hv1/vif3
> +as hv1 reset_pcap_file hv1-vif4 hv1/vif4
> +as hv2 reset_pcap_file hv2-vif1 hv2/vif1
> +as hv2 reset_pcap_file hv2-vif2 hv2/vif2
> +as hv2 reset_pcap_file hv2-vif3 hv2/vif3
> +as hv2 reset_pcap_file hv2-vif4 hv2/vif4
> +
> +send_ip_multicast_pkt hv2-vif4 hv2 \
> +    000000000001 01005e000144 \
> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
> +    e518e518000a3b3a0000
> +
> +# Sleep a bit to make sure no traffic is received and then check.
> +sleep 1
> +OVN_CHECK_PACKETS([hv1/vif1-tx.pcap], [expected_empty])
> +OVN_CHECK_PACKETS([hv2/vif3-tx.pcap], [expected_empty])
> +OVN_CHECK_PACKETS([hv1/vif4-tx.pcap], [expected_empty])
> +OVN_CHECK_PACKETS([hv1/vif2-tx.pcap], [expected_empty])
> +OVN_CHECK_PACKETS([hv1/vif3-tx.pcap], [expected_empty])
> +OVN_CHECK_PACKETS([hv2/vif1-tx.pcap], [expected_empty])
> +OVN_CHECK_PACKETS([hv2/vif2-tx.pcap], [expected_empty])
> +OVN_CHECK_PACKETS([hv2/vif4-tx.pcap], [expected_empty])
> +
> +# Enable IGMP relay on rtr
> +ovn-nbctl set logical_router rtr \
> +    options:mcast_relay="true"
> +
> +# Inject IGMP Join for 239.0.1.68 on sw1-p11.
> +send_igmp_v3_report hv1-vif1 hv1 \
> +    000000000001 $(ip_to_hex 10 0 0 1) f9f8 \
> +    $(ip_to_hex 239 0 1 68) 04 e9b9 \
> +    /dev/null
> +# Inject IGMP Join for 239.0.1.68 on sw2-p2.
> +send_igmp_v3_report hv2-vif3 hv2 \
> +    000000000001 $(ip_to_hex 10 0 0 1) f9f8 \
> +    $(ip_to_hex 239 0 1 68) 04 e9b9 \
> +    /dev/null
> +# Inject IGMP Join for 239.0.1.68 on sw3-p1.
> +send_igmp_v3_report hv1-vif4 hv1 \
> +    000000000001 $(ip_to_hex 10 0 0 1) f9f8 \
> +    $(ip_to_hex 239 0 1 68) 04 e9b9 \
> +    /dev/null
> +
> +# Check that the IGMP Group is learned by all switches.
> +OVS_WAIT_UNTIL([
> +    total_entries=`ovn-sbctl find IGMP_Group | grep "239.0.1.68" | wc -l`
> +    test "${total_entries}" = "3"
> +])
> +
> +# Send traffic from sw3 and make sure it is relayed by rtr.
> +# and ports that joined.
> +truncate -s 0 expected_routed_sw1
> +truncate -s 0 expected_routed_sw2
> +truncate -s 0 expected_switched
> +truncate -s 0 expected_empty
> +
> +as hv1 reset_pcap_file hv1-vif1 hv1/vif1
> +as hv1 reset_pcap_file hv1-vif2 hv1/vif2
> +as hv1 reset_pcap_file hv1-vif3 hv1/vif3
> +as hv1 reset_pcap_file hv1-vif4 hv1/vif4
> +as hv2 reset_pcap_file hv2-vif1 hv2/vif1
> +as hv2 reset_pcap_file hv2-vif2 hv2/vif2
> +as hv2 reset_pcap_file hv2-vif3 hv2/vif3
> +as hv2 reset_pcap_file hv2-vif4 hv2/vif4
> +
> +send_ip_multicast_pkt hv2-vif4 hv2 \
> +    000000000001 01005e000144 \
> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
> +    e518e518000a3b3a0000
> +store_ip_multicast_pkt \
> +    000000000100 01005e000144 \
> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 1f cb70 11 \
> +    e518e518000a3b3a0000 expected_routed_sw1
> +store_ip_multicast_pkt \
> +    000000000200 01005e000144 \
> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 1f cb70 11 \
> +    e518e518000a3b3a0000 expected_routed_sw2
> +store_ip_multicast_pkt \
> +    000000000001 01005e000144 \
> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
> +    e518e518000a3b3a0000 expected_switched
> +
> +OVN_CHECK_PACKETS([hv1/vif1-tx.pcap], [expected_routed_sw1])
> +OVN_CHECK_PACKETS([hv2/vif3-tx.pcap], [expected_routed_sw2])
> +OVN_CHECK_PACKETS([hv1/vif4-tx.pcap], [expected_switched])
> +OVN_CHECK_PACKETS([hv1/vif2-tx.pcap], [expected_empty])
> +OVN_CHECK_PACKETS([hv1/vif3-tx.pcap], [expected_empty])
> +OVN_CHECK_PACKETS([hv2/vif1-tx.pcap], [expected_empty])
> +OVN_CHECK_PACKETS([hv2/vif2-tx.pcap], [expected_empty])
> +OVN_CHECK_PACKETS([hv2/vif4-tx.pcap], [expected_empty])
> +
>  OVN_CLEANUP([hv1], [hv2])
>  AT_CLEANUP
>
> --
> 1.8.3.1
>
> _______________________________________________
> dev mailing list
> dev@openvswitch.org
> https://mail.openvswitch.org/mailman/listinfo/ovs-dev
>
Numan Siddique Aug. 16, 2019, 1:04 p.m. UTC | #2
On Fri, Aug 16, 2019 at 6:23 PM Numan Siddique <nusiddiq@redhat.com> wrote:

>
>
> On Fri, Aug 16, 2019 at 5:51 PM Dumitru Ceara <dceara@redhat.com> wrote:
>
>> Add a new configuration option 'mcast_relay' to the Logical_Router:options
>> in the OVN Northbound database.
>>
>> If a router is configured with 'mcast_relay' enabled then ovn-northd
>> will install Logical_Flows to allow IP multicast traffic to be routed
>> between Logical_Switches. The logical router will aggregate all IGMP
>> groups from attached logical switches and modify the routing pipeline in
>> the following way:
>> - Table S_ROUTER_IN_IP_INPUT: add flow allowing IP multicast traffic
>>   if mcast_relay is enabled on the datapath.
>> - Table S_ROUTER_IN_IP_ROUTING: add flow matching the group address,
>>   update TTL and set outport="<Multicast_Group> associated with the
>>   IGMP group". Continue to next table.
>> - Table S_ROUTER_IN_ARP_RESOLVE: bypass ARP resolve for IP multicast
>>   traffic and continue to next table.
>> - Table S_ROUTER_OUT_DELIVERY: add flow matching IP multicast traffic
>>   and set ETH.SRC to the MAC address of the logical port on which
>>   traffic is forwarded.
>>
>> Signed-off-by: Dumitru Ceara <dceara@redhat.com>
>> Acked-by: Mark Michelson <mmichels@redhat.com>
>>
>
> Acked-by: Numan Siddique <nusiddiq@redhat.com>
>
>
I went ahead and applied to master with the below changes

****************
diff --git a/northd/ovn-northd.c b/northd/ovn-northd.c
index 592d3b9d4..e29b0fff4 100644
--- a/northd/ovn-northd.c
+++ b/northd/ovn-northd.c
@@ -6355,14 +6355,8 @@ build_lrouter_flows(struct hmap *datapaths, struct
hmap *ports,
                       "drop;");

         /* Allow multicast if relay enabled (priority 95). */
-        ds_clear(&actions);
-        if (od->mcast_info.rtr.relay) {
-            ds_put_cstr(&actions, "next;");
-        } else {
-            ds_put_cstr(&actions, "drop;");
-        }
-        ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_INPUT, 95,
-                      "ip4.mcast", ds_cstr(&actions));
+        ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_INPUT, 95, "ip4.mcast",
+                      od->mcast_info.rtr.relay ? "next;" : "drop;");

         /* ARP reply handling.  Use ARP replies to populate the logical
          * router's ARP table. */
**********

Thanks
Numan


> Just one comment inline.
>
> Thanks
> Numan
>
>
>>
>> ---
>> v4:
>> - Address Numan's comment and use the predicate for ip4.mcast and
>>   add a new ip4.src_mcast predicate.
>> v3:
>> - Address Mark's comment and move setting of the outport in the IP
>>   Routing stage.
>> - Update commit message.
>> - Fix some typos.
>> v2:
>> - Optimize flooding to multicast router ports.
>> - Fix check for source IP multicast in router pipeline.
>> - Use an enum for OVN_MCAST_*_KEY definitions to avoid hard to debug
>>   errors due to typos when adding new OVN_MCAST_*_KEY values.
>> - Fix ovn-northd man page for IGMP.
>> ---
>>  NEWS                    |   1 +
>>  lib/logical-fields.c    |   2 +
>>  lib/mcast-group-index.h |  13 +-
>>  northd/ovn-northd.8.xml |  79 +++++++-
>>  northd/ovn-northd.c     | 505
>> ++++++++++++++++++++++++++++++++++++------------
>>  ovn-nb.xml              |   6 +
>>  ovn-sb.xml              |   2 +
>>  tests/ovn.at            | 199 +++++++++++++++++--
>>  8 files changed, 656 insertions(+), 151 deletions(-)
>>
>> diff --git a/NEWS b/NEWS
>> index f476984..73045d6 100644
>> --- a/NEWS
>> +++ b/NEWS
>> @@ -39,6 +39,7 @@ Post-v2.11.0
>>         logical groups which results in tunnels only been formed between
>>         members of the same transport zone(s).
>>       * Support for new logical switch port type - 'virtual'.
>> +     * Support for IGMP Snooping/Querier and Relay.
>>     - New QoS type "linux-netem" on Linux.
>>     - Added support for TLS Server Name Indication (SNI).
>>
>> diff --git a/lib/logical-fields.c b/lib/logical-fields.c
>> index 4ad5bf4..8fb591c 100644
>> --- a/lib/logical-fields.c
>> +++ b/lib/logical-fields.c
>> @@ -156,6 +156,8 @@ ovn_init_symtab(struct shash *symtab)
>>
>>      expr_symtab_add_field(symtab, "ip4.src", MFF_IPV4_SRC, "ip4", false);
>>      expr_symtab_add_field(symtab, "ip4.dst", MFF_IPV4_DST, "ip4", false);
>> +    expr_symtab_add_predicate(symtab, "ip4.src_mcast",
>> +                              "ip4.src[28..31] == 0xe");
>>      expr_symtab_add_predicate(symtab, "ip4.mcast", "ip4.dst[28..31] ==
>> 0xe");
>>
>>      expr_symtab_add_predicate(symtab, "icmp4", "ip4 && ip.proto == 1");
>> diff --git a/lib/mcast-group-index.h b/lib/mcast-group-index.h
>> index 15a1592..cb49ad7 100644
>> --- a/lib/mcast-group-index.h
>> +++ b/lib/mcast-group-index.h
>> @@ -20,8 +20,17 @@ struct ovsdb_idl;
>>
>>  struct sbrec_datapath_binding;
>>
>> -#define OVN_MCAST_FLOOD_TUNNEL_KEY   65535
>> -#define OVN_MCAST_UNKNOWN_TUNNEL_KEY (OVN_MCAST_FLOOD_TUNNEL_KEY - 1)
>> +#define OVN_MIN_MULTICAST 32768
>> +#define OVN_MAX_MULTICAST 65535
>> +
>> +enum ovn_mcast_tunnel_keys {
>> +
>> +    OVN_MCAST_FLOOD_TUNNEL_KEY = OVN_MIN_MULTICAST,
>> +    OVN_MCAST_UNKNOWN_TUNNEL_KEY,
>> +    OVN_MCAST_MROUTER_FLOOD_TUNNEL_KEY,
>> +    OVN_MIN_IP_MULTICAST,
>> +    OVN_MAX_IP_MULTICAST = OVN_MAX_MULTICAST,
>> +};
>>
>>  struct ovsdb_idl_index *mcast_group_index_create(struct ovsdb_idl *);
>>  const struct sbrec_multicast_group *
>> diff --git a/northd/ovn-northd.8.xml b/northd/ovn-northd.8.xml
>> index 6d2fbe3..d45bb15 100644
>> --- a/northd/ovn-northd.8.xml
>> +++ b/northd/ovn-northd.8.xml
>> @@ -947,10 +947,40 @@ output;
>>
>>      <ul>
>>        <li>
>> -        A priority-100 flow that outputs all packets with an Ethernet
>> broadcast
>> +        A priority-100 flow that punts all IGMP packets to
>> +        <code>ovn-controller</code> if IGMP snooping is enabled on the
>> +        logical switch.
>> +      </li>
>> +
>> +      <li>
>> +        Priority-90 flows that forward registered IP multicast traffic to
>> +        their corresponding multicast group, which
>> <code>ovn-northd</code>
>> +        creates based on learnt <ref table="IGMP_Group"
>> db="OVN_Southbound"/>
>> +        entries.  The flows also forward packets to the
>> +        <code>MC_MROUTER_FLOOD</code> multicast group, which
>> +        <code>ovn-nortdh</code> populates with all the logical ports that
>> +        are connected to logical routers with
>> +        <ref column="options"
>> table="Logical_Router"/>:mcast_relay='true'.
>> +      </li>
>> +
>> +      <li>
>> +        A priority-85 flow that forwards all IP multicast traffic
>> destined to
>> +        224.0.0.X to the <code>MC_FLOOD</code> multicast group, which
>> +        <code>ovn-northd</code> populates with all enabled logical ports.
>> +      </li>
>> +
>> +      <li>
>> +        A priority-80 flow that forwards all unregistered IP multicast
>> traffic
>> +        to the <code>MC_MROUTER_FLOOD</code> multicast group, if any.
>> +        Otherwise the flow drops all unregistered IP multicast packets.
>> This
>> +        flow is added only if <ref column="other_config"
>> +        table="Logical_Switch"/>:mcast_flood_unregistered='false'.
>> +      </li>
>> +
>> +      <li>
>> +        A priority-70 flow that outputs all packets with an Ethernet
>> broadcast
>>          or multicast <code>eth.dst</code> to the <code>MC_FLOOD</code>
>> -        multicast group, which <code>ovn-northd</code> populates with all
>> -        enabled logical ports.
>> +        multicast group.
>>        </li>
>>
>>        <li>
>> @@ -1228,6 +1258,14 @@ output;
>>
>>        <li>
>>          <p>
>> +          A priority-95 flow allows IP multicast traffic if
>> +          <ref column="options"
>> table="Logical_Router"/>:mcast_relay='true',
>> +          otherwise drops it.
>> +        </p>
>> +      </li>
>> +
>> +      <li>
>> +        <p>
>>            ICMP echo reply.  These flows reply to ICMP echo requests
>> received
>>            for the router's IP address.  Let <var>A</var> be an IP address
>>            owned by a router port.  Then, for each <var>A</var> that is
>> @@ -1941,6 +1979,16 @@ output;
>>      <ul>
>>        <li>
>>          <p>
>> +          Priority-500 flows that match IP multicast traffic destined to
>> +          groups registered on any of the attached switches and sets
>> +          <code>outport</code> to the associated multicast group that
>> will
>> +          eventually flood the traffic to all interested attached logical
>> +          switches. The flows also decrement TTL.
>> +        </p>
>> +      </li>
>> +
>> +      <li>
>> +        <p>
>>            For distributed logical routers where one of the logical router
>>            ports specifies a <code>redirect-chassis</code>, a priority-400
>>            logical flow for each ip source/destination couple that
>> matches the
>> @@ -2074,6 +2122,15 @@ next;
>>      <ul>
>>        <li>
>>          <p>
>> +          A priority-500 flow that matches IP multicast traffic that was
>> +          allowed in the routing pipeline. For this kind of traffic the
>> +          <code>outport</code> was already set so the flow just advances
>> to
>> +          the next table.
>> +        </p>
>> +      </li>
>> +
>> +      <li>
>> +        <p>
>>            For distributed logical routers where one of the logical router
>>            ports specifies a <code>redirect-chassis</code>, a priority-400
>>            logical flow with match <code>REGBIT_DISTRIBUTED_NAT ==
>> 1</code>
>> @@ -2641,9 +2698,19 @@ clone {
>>      <h3>Egress Table 3: Delivery</h3>
>>
>>      <p>
>> -      Packets that reach this table are ready for delivery.  It contains
>> -      priority-100 logical flows that match packets on each enabled
>> logical
>> -      router port, with action <code>output;</code>.
>> +      Packets that reach this table are ready for delivery.  It contains:
>> +      <ul>
>> +        <li>
>> +          Priority-110 logical flows that match IP multicast packets on
>> each
>> +          enabled logical router port and modify the Ethernet source
>> address
>> +          of the packets to the Ethernet address of the port and then
>> execute
>> +          action <code>output;</code>.
>> +        </li>
>> +        <li>
>> +          Priority-100 logical flows that match packets on each enabled
>> +          logical router port, with action <code>output;</code>.
>> +        </li>
>> +      </ul>
>>      </p>
>>
>>  </manpage>
>> diff --git a/northd/ovn-northd.c b/northd/ovn-northd.c
>> index e861344..592d3b9 100644
>> --- a/northd/ovn-northd.c
>> +++ b/northd/ovn-northd.c
>> @@ -433,32 +433,52 @@ struct ipam_info {
>>      bool mac_only;
>>  };
>>
>> -#define OVN_MIN_MULTICAST 32768
>> -#define OVN_MAX_MULTICAST OVN_MCAST_FLOOD_TUNNEL_KEY
>> -BUILD_ASSERT_DECL(OVN_MIN_MULTICAST < OVN_MAX_MULTICAST);
>> -
>> -#define OVN_MIN_IP_MULTICAST OVN_MIN_MULTICAST
>> -#define OVN_MAX_IP_MULTICAST (OVN_MCAST_UNKNOWN_TUNNEL_KEY - 1)
>> -BUILD_ASSERT_DECL(OVN_MAX_IP_MULTICAST >= OVN_MIN_MULTICAST);
>> -
>>  /*
>>   * Multicast snooping and querier per datapath configuration.
>>   */
>> +struct mcast_switch_info {
>> +
>> +    bool enabled;               /* True if snooping enabled. */
>> +    bool querier;               /* True if querier enabled. */
>> +    bool flood_unregistered;    /* True if unregistered multicast should
>> be
>> +                                 * flooded.
>> +                                 */
>> +    bool flood_relay;           /* True if the switch is connected to a
>> +                                 * multicast router and unregistered
>> multicast
>> +                                 * should be flooded to the mrouter. Only
>> +                                 * applicable if flood_unregistered ==
>> false.
>> +                                 */
>> +
>> +    int64_t table_size;         /* Max number of IP multicast groups. */
>> +    int64_t idle_timeout;       /* Timeout after which an idle group is
>> +                                 * flushed.
>> +                                 */
>> +    int64_t query_interval;     /* Interval between multicast queries. */
>> +    char *eth_src;              /* ETH src address of the multicast
>> queries. */
>> +    char *ipv4_src;             /* IP src address of the multicast
>> queries. */
>> +    int64_t query_max_response; /* Expected time after which reports
>> should
>> +                                 * be received for queries that were
>> sent out.
>> +                                 */
>> +
>> +    uint32_t active_flows;      /* Current number of active IP multicast
>> +                                 * flows.
>> +                                 */
>> +};
>> +
>> +struct mcast_router_info {
>> +    bool relay; /* True if the router should relay IP multicast. */
>> +};
>> +
>>  struct mcast_info {
>> -    bool enabled;
>> -    bool querier;
>> -    bool flood_unregistered;
>> -
>> -    int64_t table_size;
>> -    int64_t idle_timeout;
>> -    int64_t query_interval;
>> -    char *eth_src;
>> -    char *ipv4_src;
>> -    int64_t  query_max_response;
>> -
>> -    struct hmap group_tnlids;
>> -    uint32_t group_tnlid_hint;
>> -    uint32_t active_flows;
>> +
>> +    struct hmap group_tnlids;  /* Group tunnel IDs in use on this DP. */
>> +    uint32_t group_tnlid_hint; /* Hint for allocating next group tunnel
>> ID. */
>> +    struct ovs_list groups;    /* List of groups learnt on this DP. */
>> +
>> +    union {
>> +        struct mcast_switch_info sw;  /* Switch specific multicast info.
>> */
>> +        struct mcast_router_info rtr; /* Router specific multicast info.
>> */
>> +    };
>>  };
>>
>>  static uint32_t
>> @@ -559,6 +579,7 @@ ovn_datapath_create(struct hmap *datapaths, const
>> struct uuid *key,
>>  }
>>
>>  static void ovn_ls_port_group_destroy(struct hmap *nb_pgs);
>> +static void destroy_mcast_info_for_datapath(struct ovn_datapath *od);
>>
>>  static void
>>  ovn_datapath_destroy(struct hmap *datapaths, struct ovn_datapath *od)
>> @@ -572,12 +593,7 @@ ovn_datapath_destroy(struct hmap *datapaths, struct
>> ovn_datapath *od)
>>          bitmap_free(od->ipam_info.allocated_ipv4s);
>>          free(od->router_ports);
>>          ovn_ls_port_group_destroy(&od->nb_pgs);
>> -
>> -        if (od->nbs) {
>> -            free(od->mcast_info.eth_src);
>> -            free(od->mcast_info.ipv4_src);
>> -            destroy_tnlids(&od->mcast_info.group_tnlids);
>> -        }
>> +        destroy_mcast_info_for_datapath(od);
>>
>>          free(od);
>>      }
>> @@ -714,23 +730,28 @@ init_ipam_info_for_datapath(struct ovn_datapath *od)
>>  }
>>
>>  static void
>> -init_mcast_info_for_datapath(struct ovn_datapath *od)
>> +init_mcast_info_for_router_datapath(struct ovn_datapath *od)
>>  {
>> -    if (!od->nbs) {
>> -        return;
>> -    }
>> +    struct mcast_router_info *mcast_rtr_info = &od->mcast_info.rtr;
>>
>> -    struct mcast_info *mcast_info = &od->mcast_info;
>> +    mcast_rtr_info->relay = smap_get_bool(&od->nbr->options,
>> "mcast_relay",
>> +                                          false);
>> +}
>>
>> -    mcast_info->enabled =
>> +static void
>> +init_mcast_info_for_switch_datapath(struct ovn_datapath *od)
>> +{
>> +    struct mcast_switch_info *mcast_sw_info = &od->mcast_info.sw;
>> +
>> +    mcast_sw_info->enabled =
>>          smap_get_bool(&od->nbs->other_config, "mcast_snoop", false);
>> -    mcast_info->querier =
>> +    mcast_sw_info->querier =
>>          smap_get_bool(&od->nbs->other_config, "mcast_querier", true);
>> -    mcast_info->flood_unregistered =
>> +    mcast_sw_info->flood_unregistered =
>>          smap_get_bool(&od->nbs->other_config, "mcast_flood_unregistered",
>>                        false);
>>
>> -    mcast_info->table_size =
>> +    mcast_sw_info->table_size =
>>          smap_get_ullong(&od->nbs->other_config, "mcast_table_size",
>>                          OVN_MCAST_DEFAULT_MAX_ENTRIES);
>>
>> @@ -742,54 +763,94 @@ init_mcast_info_for_datapath(struct ovn_datapath
>> *od)
>>      } else if (idle_timeout > OVN_MCAST_MAX_IDLE_TIMEOUT_S) {
>>          idle_timeout = OVN_MCAST_MAX_IDLE_TIMEOUT_S;
>>      }
>> -    mcast_info->idle_timeout = idle_timeout;
>> +    mcast_sw_info->idle_timeout = idle_timeout;
>>
>>      uint32_t query_interval =
>>          smap_get_ullong(&od->nbs->other_config, "mcast_query_interval",
>> -                        mcast_info->idle_timeout / 2);
>> +                        mcast_sw_info->idle_timeout / 2);
>>      if (query_interval < OVN_MCAST_MIN_QUERY_INTERVAL_S) {
>>          query_interval = OVN_MCAST_MIN_QUERY_INTERVAL_S;
>>      } else if (query_interval > OVN_MCAST_MAX_QUERY_INTERVAL_S) {
>>          query_interval = OVN_MCAST_MAX_QUERY_INTERVAL_S;
>>      }
>> -    mcast_info->query_interval = query_interval;
>> +    mcast_sw_info->query_interval = query_interval;
>>
>> -    mcast_info->eth_src =
>> +    mcast_sw_info->eth_src =
>>          nullable_xstrdup(smap_get(&od->nbs->other_config,
>> "mcast_eth_src"));
>> -    mcast_info->ipv4_src =
>> +    mcast_sw_info->ipv4_src =
>>          nullable_xstrdup(smap_get(&od->nbs->other_config,
>> "mcast_ip4_src"));
>>
>> -    mcast_info->query_max_response =
>> +    mcast_sw_info->query_max_response =
>>          smap_get_ullong(&od->nbs->other_config,
>> "mcast_query_max_response",
>>                          OVN_MCAST_DEFAULT_QUERY_MAX_RESPONSE_S);
>>
>> -    hmap_init(&mcast_info->group_tnlids);
>> -    mcast_info->group_tnlid_hint = OVN_MIN_IP_MULTICAST;
>> -    mcast_info->active_flows = 0;
>> +    mcast_sw_info->active_flows = 0;
>> +}
>> +
>> +static void
>> +init_mcast_info_for_datapath(struct ovn_datapath *od)
>> +{
>> +    if (!od->nbr && !od->nbs) {
>> +        return;
>> +    }
>> +
>> +    hmap_init(&od->mcast_info.group_tnlids);
>> +    od->mcast_info.group_tnlid_hint = OVN_MIN_IP_MULTICAST;
>> +    ovs_list_init(&od->mcast_info.groups);
>> +
>> +    if (od->nbs) {
>> +        init_mcast_info_for_switch_datapath(od);
>> +    } else {
>> +        init_mcast_info_for_router_datapath(od);
>> +    }
>> +}
>> +
>> +static void
>> +destroy_mcast_info_for_switch_datapath(struct ovn_datapath *od)
>> +{
>> +    struct mcast_switch_info *mcast_sw_info = &od->mcast_info.sw;
>> +
>> +    free(mcast_sw_info->eth_src);
>> +    free(mcast_sw_info->ipv4_src);
>> +}
>> +
>> +static void
>> +destroy_mcast_info_for_datapath(struct ovn_datapath *od)
>> +{
>> +    if (!od->nbr && !od->nbs) {
>> +        return;
>> +    }
>> +
>> +    if (od->nbs) {
>> +        destroy_mcast_info_for_switch_datapath(od);
>> +    }
>> +
>> +    destroy_tnlids(&od->mcast_info.group_tnlids);
>>  }
>>
>>  static void
>> -store_mcast_info_for_datapath(const struct sbrec_ip_multicast *sb,
>> -                              struct ovn_datapath *od)
>> +store_mcast_info_for_switch_datapath(const struct sbrec_ip_multicast *sb,
>> +                                     struct ovn_datapath *od)
>>  {
>> -    struct mcast_info *mcast_info = &od->mcast_info;
>> +    struct mcast_switch_info *mcast_sw_info = &od->mcast_info.sw;
>>
>>      sbrec_ip_multicast_set_datapath(sb, od->sb);
>> -    sbrec_ip_multicast_set_enabled(sb, &mcast_info->enabled, 1);
>> -    sbrec_ip_multicast_set_querier(sb, &mcast_info->querier, 1);
>> -    sbrec_ip_multicast_set_table_size(sb, &mcast_info->table_size, 1);
>> -    sbrec_ip_multicast_set_idle_timeout(sb, &mcast_info->idle_timeout,
>> 1);
>> +    sbrec_ip_multicast_set_enabled(sb, &mcast_sw_info->enabled, 1);
>> +    sbrec_ip_multicast_set_querier(sb, &mcast_sw_info->querier, 1);
>> +    sbrec_ip_multicast_set_table_size(sb, &mcast_sw_info->table_size, 1);
>> +    sbrec_ip_multicast_set_idle_timeout(sb,
>> &mcast_sw_info->idle_timeout, 1);
>>      sbrec_ip_multicast_set_query_interval(sb,
>> -                                          &mcast_info->query_interval,
>> 1);
>> +
>> &mcast_sw_info->query_interval, 1);
>>      sbrec_ip_multicast_set_query_max_resp(sb,
>> -
>> &mcast_info->query_max_response, 1);
>> +
>> &mcast_sw_info->query_max_response,
>> +                                          1);
>>
>> -    if (mcast_info->eth_src) {
>> -        sbrec_ip_multicast_set_eth_src(sb, mcast_info->eth_src);
>> +    if (mcast_sw_info->eth_src) {
>> +        sbrec_ip_multicast_set_eth_src(sb, mcast_sw_info->eth_src);
>>      }
>>
>> -    if (mcast_info->ipv4_src) {
>> -        sbrec_ip_multicast_set_ip4_src(sb, mcast_info->ipv4_src);
>> +    if (mcast_sw_info->ipv4_src) {
>> +        sbrec_ip_multicast_set_ip4_src(sb, mcast_sw_info->ipv4_src);
>>      }
>>  }
>>
>> @@ -906,6 +967,7 @@ join_datapaths(struct northd_context *ctx, struct
>> hmap *datapaths,
>>                                       NULL, nbr, NULL);
>>              ovs_list_push_back(nb_only, &od->list);
>>          }
>> +        init_mcast_info_for_datapath(od);
>>          ovs_list_push_back(lr_list, &od->lr_list);
>>      }
>>  }
>> @@ -1999,6 +2061,13 @@ join_logical_ports(struct northd_context *ctx,
>>                      break;
>>                  }
>>              }
>> +
>> +            /* If the router is multicast enabled then set relay on the
>> switch
>> +             * datapath.
>> +             */
>> +            if (peer->od && peer->od->mcast_info.rtr.relay) {
>> +                op->od->mcast_info.sw.flood_relay = true;
>> +            }
>>          } else if (op->nbrp && op->nbrp->peer && !op->derived) {
>>              struct ovn_port *peer = ovn_port_find(ports, op->nbrp->peer);
>>              if (peer) {
>> @@ -2846,6 +2915,10 @@ struct multicast_group {
>>  static const struct multicast_group mc_flood =
>>      { MC_FLOOD, OVN_MCAST_FLOOD_TUNNEL_KEY };
>>
>> +#define MC_MROUTER_FLOOD "_MC_mrouter_flood"
>> +static const struct multicast_group mc_mrouter_flood =
>> +    { MC_MROUTER_FLOOD, OVN_MCAST_MROUTER_FLOOD_TUNNEL_KEY };
>> +
>>  #define MC_UNKNOWN "_MC_unknown"
>>  static const struct multicast_group mc_unknown =
>>      { MC_UNKNOWN, OVN_MCAST_UNKNOWN_TUNNEL_KEY };
>> @@ -2955,7 +3028,8 @@ ovn_multicast_update_sbrec(const struct
>> ovn_multicast *mc,
>>   */
>>  struct ovn_igmp_group_entry {
>>      struct ovs_list list_node; /* Linkage in the list of entries. */
>> -    const struct sbrec_igmp_group *sb;
>> +    size_t n_ports;
>> +    struct ovn_port **ports;
>>  };
>>
>>  /*
>> @@ -2964,12 +3038,13 @@ struct ovn_igmp_group_entry {
>>   */
>>  struct ovn_igmp_group {
>>      struct hmap_node hmap_node; /* Index on 'datapath' and 'address'. */
>> +    struct ovs_list list_node;  /* Linkage in the per-dp igmp group
>> list. */
>>
>>      struct ovn_datapath *datapath;
>>      struct in6_addr address; /* Multicast IPv6-mapped-IPv4 or IPv4
>> address. */
>>      struct multicast_group mcgroup;
>>
>> -    struct ovs_list sb_entries; /* List of SB entries for this group. */
>> +    struct ovs_list entries; /* List of SB entries for this group. */
>>  };
>>
>>  static uint32_t
>> @@ -2997,77 +3072,120 @@ ovn_igmp_group_find(struct hmap *igmp_groups,
>>      return NULL;
>>  }
>>
>> -static void
>> +static struct ovn_igmp_group *
>>  ovn_igmp_group_add(struct northd_context *ctx, struct hmap *igmp_groups,
>>                     struct ovn_datapath *datapath,
>> -                   const struct sbrec_igmp_group *sb_igmp_group)
>> +                   const struct in6_addr *address,
>> +                   const char *address_s)
>>  {
>> -    struct in6_addr group_address;
>> -    ovs_be32 ipv4;
>> -
>> -    if (ip_parse(sb_igmp_group->address, &ipv4)) {
>> -        group_address = in6_addr_mapped_ipv4(ipv4);
>> -    } else if (!ipv6_parse(sb_igmp_group->address, &group_address)) {
>> -        static struct vlog_rate_limit rl = VLOG_RATE_LIMIT_INIT(1, 1);
>> -        VLOG_WARN_RL(&rl, "invalid IGMP group address: %s",
>> -                     sb_igmp_group->address);
>> -        return;
>> -    }
>> -
>>      struct ovn_igmp_group *igmp_group =
>> -        ovn_igmp_group_find(igmp_groups, datapath, &group_address);
>> +        ovn_igmp_group_find(igmp_groups, datapath, address);
>>
>>      if (!igmp_group) {
>>          igmp_group = xmalloc(sizeof *igmp_group);
>>
>>          const struct sbrec_multicast_group *mcgroup =
>> -            mcast_group_lookup(ctx->sbrec_mcast_group_by_name_dp,
>> -                               sb_igmp_group->address, datapath->sb);
>> +            mcast_group_lookup(ctx->sbrec_mcast_group_by_name_dp,
>> address_s,
>> +                               datapath->sb);
>>
>>          igmp_group->datapath = datapath;
>> -        igmp_group->address = group_address;
>> +        igmp_group->address = *address;
>>          if (mcgroup) {
>>              igmp_group->mcgroup.key = mcgroup->tunnel_key;
>>              add_tnlid(&datapath->mcast_info.group_tnlids,
>> mcgroup->tunnel_key);
>>          } else {
>>              igmp_group->mcgroup.key = 0;
>>          }
>> -        igmp_group->mcgroup.name = sb_igmp_group->address;
>> -        ovs_list_init(&igmp_group->sb_entries);
>> +        igmp_group->mcgroup.name = address_s;
>> +        ovs_list_init(&igmp_group->entries);
>>
>>          hmap_insert(igmp_groups, &igmp_group->hmap_node,
>> -                    ovn_igmp_group_hash(datapath, &group_address));
>> +                    ovn_igmp_group_hash(datapath, address));
>> +        ovs_list_push_back(&datapath->mcast_info.groups,
>> +                           &igmp_group->list_node);
>> +    }
>> +
>> +    return igmp_group;
>> +}
>> +
>> +static bool
>> +ovn_igmp_group_get_address(const struct sbrec_igmp_group *sb_igmp_group,
>> +                           struct in6_addr *address)
>> +{
>> +    ovs_be32 ipv4;
>> +
>> +    if (ip_parse(sb_igmp_group->address, &ipv4)) {
>> +        *address = in6_addr_mapped_ipv4(ipv4);
>> +        return true;
>> +    }
>> +    if (!ipv6_parse(sb_igmp_group->address, address)) {
>> +        return false;
>>      }
>> +    return true;
>> +}
>>
>> +static struct ovn_port **
>> +ovn_igmp_group_get_ports(const struct sbrec_igmp_group *sb_igmp_group,
>> +                         size_t *n_ports, struct hmap *ovn_ports)
>> +{
>> +    struct ovn_port **ports = xmalloc(sb_igmp_group->n_ports * sizeof
>> *ports);
>> +
>> +     *n_ports = 0;
>> +     for (size_t i = 0; i < sb_igmp_group->n_ports; i++) {
>> +        ports[(*n_ports)] =
>> +            ovn_port_find(ovn_ports,
>> sb_igmp_group->ports[i]->logical_port);
>> +        if (ports[(*n_ports)]) {
>> +            (*n_ports)++;
>> +        }
>> +    }
>> +
>> +    return ports;
>> +}
>> +
>> +static void
>> +ovn_igmp_group_add_entry(struct ovn_igmp_group *igmp_group,
>> +                         struct ovn_port **ports, size_t n_ports)
>> +{
>>      struct ovn_igmp_group_entry *entry = xmalloc(sizeof *entry);
>>
>> -    entry->sb = sb_igmp_group;
>> -    ovs_list_push_back(&igmp_group->sb_entries , &entry->list_node);
>> +    entry->ports = ports;
>> +    entry->n_ports = n_ports;
>> +    ovs_list_push_back(&igmp_group->entries, &entry->list_node);
>> +}
>> +
>> +static void
>> +ovn_igmp_group_destroy_entry(struct ovn_igmp_group_entry *entry)
>> +{
>> +    free(entry->ports);
>> +}
>> +
>> +static bool
>> +ovn_igmp_group_allocate_id(struct ovn_igmp_group *igmp_group)
>> +{
>> +    if (igmp_group->mcgroup.key == 0) {
>> +        struct mcast_info *mcast_info =
>> &igmp_group->datapath->mcast_info;
>> +        igmp_group->mcgroup.key =
>> ovn_mcast_group_allocate_key(mcast_info);
>> +    }
>> +
>> +    if (igmp_group->mcgroup.key == 0) {
>> +        return false;
>> +    }
>> +
>> +    return true;
>>  }
>>
>>  static void
>>  ovn_igmp_group_aggregate_ports(struct ovn_igmp_group *igmp_group,
>> -                               struct hmap *ovn_ports,
>>                                 struct hmap *mcast_groups)
>>  {
>>      struct ovn_igmp_group_entry *entry;
>>
>> -    LIST_FOR_EACH_POP (entry, list_node, &igmp_group->sb_entries) {
>> -        size_t n_oports = 0;
>> -        struct ovn_port **oports =
>> -            xmalloc(entry->sb->n_ports * sizeof *oports);
>> -
>> -        for (size_t i = 0; i < entry->sb->n_ports; i++) {
>> -            oports[n_oports] =
>> -                ovn_port_find(ovn_ports,
>> entry->sb->ports[i]->logical_port);
>> -            if (oports[n_oports]) {
>> -                n_oports++;
>> -            }
>> -        }
>> -
>> +    LIST_FOR_EACH_POP (entry, list_node, &igmp_group->entries) {
>>          ovn_multicast_add_ports(mcast_groups, igmp_group->datapath,
>> -                                &igmp_group->mcgroup, oports, n_oports);
>> -        free(oports);
>> +                                &igmp_group->mcgroup, entry->ports,
>> +                                entry->n_ports);
>> +
>> +        ovn_igmp_group_destroy_entry(entry);
>>          free(entry);
>>      }
>>  }
>> @@ -3079,10 +3197,12 @@ ovn_igmp_group_destroy(struct hmap *igmp_groups,
>>      if (igmp_group) {
>>          struct ovn_igmp_group_entry *entry;
>>
>> -        LIST_FOR_EACH_POP (entry, list_node, &igmp_group->sb_entries) {
>> +        LIST_FOR_EACH_POP (entry, list_node, &igmp_group->entries) {
>> +            ovn_igmp_group_destroy_entry(entry);
>>              free(entry);
>>          }
>>          hmap_remove(igmp_groups, &igmp_group->hmap_node);
>> +        ovs_list_remove(&igmp_group->list_node);
>>          free(igmp_group);
>>      }
>>  }
>> @@ -5286,7 +5406,9 @@ build_lswitch_flows(struct hmap *datapaths, struct
>> hmap *ports,
>>              continue;
>>          }
>>
>> -        if (od->mcast_info.enabled) {
>> +        struct mcast_switch_info *mcast_sw_info = &od->mcast_info.sw;
>> +
>> +        if (mcast_sw_info->enabled) {
>>              /* Punt IGMP traffic to controller. */
>>              ovn_lflow_add(lflows, od, S_SWITCH_IN_L2_LKUP, 100,
>>                            "ip4 && ip.proto == 2", "igmp;");
>> @@ -5299,9 +5421,16 @@ build_lswitch_flows(struct hmap *datapaths, struct
>> hmap *ports,
>>                            "outport = \""MC_FLOOD"\"; output;");
>>
>>              /* Drop unregistered IP multicast if not allowed. */
>> -            if (!od->mcast_info.flood_unregistered) {
>> -                ovn_lflow_add(lflows, od, S_SWITCH_IN_L2_LKUP, 80,
>> -                              "ip4 && ip4.mcast", "drop;");
>> +            if (!mcast_sw_info->flood_unregistered) {
>> +                /* Forward unregistered IP multicast to mrouter (if
>> any). */
>> +                if (mcast_sw_info->flood_relay) {
>> +                    ovn_lflow_add(lflows, od, S_SWITCH_IN_L2_LKUP, 80,
>> +                                  "ip4 && ip4.mcast",
>> +                                  "outport = \""MC_MROUTER_FLOOD"\";
>> output;");
>> +                } else {
>> +                    ovn_lflow_add(lflows, od, S_SWITCH_IN_L2_LKUP, 80,
>> +                                  "ip4 && ip4.mcast", "drop;");
>> +                }
>>              }
>>          }
>>
>> @@ -5318,18 +5447,26 @@ build_lswitch_flows(struct hmap *datapaths,
>> struct hmap *ports,
>>              continue;
>>          }
>>
>> -        struct mcast_info *mcast_info =
>> &igmp_group->datapath->mcast_info;
>> +        struct mcast_switch_info *mcast_sw_info =
>> +            &igmp_group->datapath->mcast_info.sw;
>>
>> -        if (mcast_info->active_flows >= mcast_info->table_size) {
>> +        if (mcast_sw_info->active_flows >= mcast_sw_info->table_size) {
>>              continue;
>>          }
>> -        mcast_info->active_flows++;
>> +        mcast_sw_info->active_flows++;
>>
>>          ds_clear(&match);
>>          ds_clear(&actions);
>>
>>          ds_put_format(&match, "eth.mcast && ip4 && ip4.dst == %s ",
>>                        igmp_group->mcgroup.name);
>> +        /* Also flood traffic to all multicast routers with relay
>> enabled. */
>> +        if (mcast_sw_info->flood_relay) {
>> +            ds_put_cstr(&actions,
>> +                        "clone { "
>> +                            "outport = \""MC_MROUTER_FLOOD "\"; output; "
>> +                        "};");
>> +        }
>>          ds_put_format(&actions, "outport = \"%s\"; output; ",
>>                        igmp_group->mcgroup.name);
>>
>> @@ -6209,7 +6346,7 @@ build_lrouter_flows(struct hmap *datapaths, struct
>> hmap *ports,
>>           * source or destination, and zero network source or destination
>>           * (priority 100). */
>>          ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_INPUT, 100,
>> -                      "ip4.mcast || "
>> +                      "ip4.src_mcast ||"
>>                        "ip4.src == 255.255.255.255 || "
>>                        "ip4.src == 127.0.0.0/8 || "
>>                        "ip4.dst == 127.0.0.0/8 || "
>> @@ -6217,6 +6354,16 @@ build_lrouter_flows(struct hmap *datapaths, struct
>> hmap *ports,
>>                        "ip4.dst == 0.0.0.0/8",
>>                        "drop;");
>>
>> +        /* Allow multicast if relay enabled (priority 95). */
>> +        ds_clear(&actions);
>> +        if (od->mcast_info.rtr.relay) {
>> +            ds_put_cstr(&actions, "next;");
>> +        } else {
>> +            ds_put_cstr(&actions, "drop;");
>> +        }
>> +        ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_INPUT, 95,
>> +                      "ip4.mcast", ds_cstr(&actions));
>> +
>>
>
> Instead of if/else how about
> ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_INPUT, 95, "ip4.mcast",
> od->mcast_info.rtr.relay ? "next": "drop");
>
>
>          /* ARP reply handling.  Use ARP replies to populate the logical
>>           * router's ARP table. */
>>          ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_INPUT, 90, "arp.op ==
>> 2",
>> @@ -7487,6 +7634,27 @@ build_lrouter_flows(struct hmap *datapaths, struct
>> hmap *ports,
>>          }
>>      }
>>
>> +    /* IP Multicast lookup. Here we set the output port, adjust TTL and
>> +     * advance to next table (priority 500).
>> +     */
>> +    HMAP_FOR_EACH (od, key_node, datapaths) {
>> +        if (!od->nbr || !od->mcast_info.rtr.relay) {
>> +            continue;
>> +        }
>> +        struct ovn_igmp_group *igmp_group;
>> +
>> +        LIST_FOR_EACH (igmp_group, list_node, &od->mcast_info.groups) {
>> +            ds_clear(&match);
>> +            ds_clear(&actions);
>> +            ds_put_format(&match, "ip4 && ip4.dst == %s ",
>> +                          igmp_group->mcgroup.name);
>> +            ds_put_format(&actions, "outport = \"%s\"; ip.ttl--; next;",
>> +                          igmp_group->mcgroup.name);
>> +            ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_ROUTING, 500,
>> +                          ds_cstr(&match), ds_cstr(&actions));
>> +        }
>> +    }
>> +
>>      /* Logical router ingress table 8: Policy.
>>       *
>>       * A packet that arrives at this table is an IP packet that should be
>> @@ -7517,10 +7685,24 @@ build_lrouter_flows(struct hmap *datapaths,
>> struct hmap *ports,
>>
>>      /* Local router ingress table 9: ARP Resolution.
>>       *
>> -     * Any packet that reaches this table is an IP packet whose next-hop
>> IP
>> -     * address is in reg0. (ip4.dst is the final destination.) This table
>> -     * resolves the IP address in reg0 into an output port in outport
>> and an
>> -     * Ethernet address in eth.dst. */
>> +     * Multicast packets already have the outport set so just advance to
>> next
>> +     * table (priority 500). */
>> +    HMAP_FOR_EACH (od, key_node, datapaths) {
>> +        if (!od->nbr) {
>> +            continue;
>> +        }
>> +
>> +        ovn_lflow_add(lflows, od, S_ROUTER_IN_ARP_RESOLVE, 500,
>> +                      "ip4.mcast", "next;");
>> +    }
>> +
>> +    /* Local router ingress table 9: ARP Resolution.
>> +     *
>> +     * Any unicast packet that reaches this table is an IP packet whose
>> +     * next-hop IP address is in reg0. (ip4.dst is the final
>> destination.)
>> +     * This table resolves the IP address in reg0 into an output port in
>> +     * outport and an Ethernet address in eth.dst.
>> +     */
>>      HMAP_FOR_EACH (op, key_node, ports) {
>>          if (op->nbsp && !lsp_is_enabled(op->nbsp)) {
>>              continue;
>> @@ -8002,9 +8184,13 @@ build_lrouter_flows(struct hmap *datapaths, struct
>> hmap *ports,
>>          ovn_lflow_add(lflows, od, S_ROUTER_IN_ARP_REQUEST, 0, "1",
>> "output;");
>>      }
>>
>> -    /* Logical router egress table 1: Delivery (priority 100).
>> +    /* Logical router egress table 1: Delivery (priority 100-110).
>>       *
>> -     * Priority 100 rules deliver packets to enabled logical ports. */
>> +     * Priority 100 rules deliver packets to enabled logical ports.
>> +     * Priority 110 rules match multicast packets and update the source
>> +     * mac before delivering to enabled logical ports. IP multicast
>> traffic
>> +     * bypasses S_ROUTER_IN_IP_ROUTING route lookups.
>> +     */
>>      HMAP_FOR_EACH (op, key_node, ports) {
>>          if (!op->nbrp) {
>>              continue;
>> @@ -8024,6 +8210,20 @@ build_lrouter_flows(struct hmap *datapaths, struct
>> hmap *ports,
>>              continue;
>>          }
>>
>> +        /* If multicast relay is enabled then also adjust source mac for
>> IP
>> +         * multicast traffic.
>> +         */
>> +        if (op->od->mcast_info.rtr.relay) {
>> +            ds_clear(&match);
>> +            ds_clear(&actions);
>> +            ds_put_format(&match, "ip4.mcast && outport == %s",
>> +                          op->json_key);
>> +            ds_put_format(&actions, "eth.src = %s; output;",
>> +                          op->lrp_networks.ea_s);
>> +            ovn_lflow_add(lflows, op->od, S_ROUTER_OUT_DELIVERY, 110,
>> +                        ds_cstr(&match), ds_cstr(&actions));
>> +        }
>> +
>>          ds_clear(&match);
>>          ds_put_format(&match, "outport == %s", op->json_key);
>>          ovn_lflow_add(lflows, op->od, S_ROUTER_OUT_DELIVERY, 100,
>> @@ -8574,7 +8774,7 @@ build_ip_mcast(struct northd_context *ctx, struct
>> hmap *datapaths)
>>          if (!ip_mcast) {
>>              ip_mcast = sbrec_ip_multicast_insert(ctx->ovnsb_txn);
>>          }
>> -        store_mcast_info_for_datapath(ip_mcast, od);
>> +        store_mcast_info_for_switch_datapath(ip_mcast, od);
>>      }
>>
>>      /* Delete southbound records without northbound matches. */
>> @@ -8606,6 +8806,14 @@ build_mcast_groups(struct northd_context *ctx,
>>
>>          if (lsp_is_enabled(op->nbsp)) {
>>              ovn_multicast_add(mcast_groups, &mc_flood, op);
>> +
>> +            /* If this port is connected to a multicast router then add
>> it
>> +             * to the MC_MROUTER_FLOOD group.
>> +             */
>> +            if (op->od->mcast_info.sw.flood_relay && op->peer &&
>> +                    op->peer->od && op->peer->od->mcast_info.rtr.relay) {
>> +                ovn_multicast_add(mcast_groups, &mc_mrouter_flood, op);
>> +            }
>>          }
>>      }
>>
>> @@ -8628,10 +8836,61 @@ build_mcast_groups(struct northd_context *ctx,
>>              continue;
>>          }
>>
>> +        struct in6_addr group_address;
>> +        if (!ovn_igmp_group_get_address(sb_igmp, &group_address)) {
>> +            static struct vlog_rate_limit rl = VLOG_RATE_LIMIT_INIT(1,
>> 1);
>> +            VLOG_WARN_RL(&rl, "invalid IGMP group address: %s",
>> +                         sb_igmp->address);
>> +            continue;
>> +        }
>> +
>>          /* Add the IGMP group entry. Will also try to allocate an ID for
>> it
>>           * if the multicast group already exists.
>>           */
>> -        ovn_igmp_group_add(ctx, igmp_groups, od, sb_igmp);
>> +        struct ovn_igmp_group *igmp_group =
>> +            ovn_igmp_group_add(ctx, igmp_groups, od, &group_address,
>> +                               sb_igmp->address);
>> +
>> +        /* Extract the IGMP group ports from the SB entry and store them
>> +         * in the IGMP group.
>> +         */
>> +        size_t n_igmp_ports;
>> +        struct ovn_port **igmp_ports =
>> +            ovn_igmp_group_get_ports(sb_igmp, &n_igmp_ports, ports);
>> +        ovn_igmp_group_add_entry(igmp_group, igmp_ports, n_igmp_ports);
>> +    }
>> +
>> +    /* Build IGMP groups for multicast routers with relay enabled. The
>> router
>> +     * IGMP groups are based on the groups learnt by their multicast
>> enabled
>> +     * peers.
>> +     */
>> +    struct ovn_datapath *od;
>> +    HMAP_FOR_EACH (od, key_node, datapaths) {
>> +
>> +        if (ovs_list_is_empty(&od->mcast_info.groups)) {
>> +            continue;
>> +        }
>> +
>> +        for (size_t i = 0; i < od->n_router_ports; i++) {
>> +            struct ovn_port *router_port = od->router_ports[i]->peer;
>> +
>> +            if (!router_port || !router_port->od ||
>> +                    !router_port->od->mcast_info.rtr.relay) {
>> +                continue;
>> +            }
>> +
>> +            struct ovn_igmp_group *igmp_group;
>> +            LIST_FOR_EACH (igmp_group, list_node,
>> &od->mcast_info.groups) {
>> +                struct ovn_igmp_group *igmp_group_rtr =
>> +                    ovn_igmp_group_add(ctx, igmp_groups, router_port->od,
>> +                                       &igmp_group->address,
>> +                                       igmp_group->mcgroup.name);
>> +                struct ovn_port **router_igmp_ports =
>> +                    xmalloc(sizeof *router_igmp_ports);
>> +                router_igmp_ports[0] = router_port;
>> +                ovn_igmp_group_add_entry(igmp_group_rtr,
>> router_igmp_ports, 1);
>> +            }
>> +        }
>>      }
>>
>>      /* Walk the aggregated IGMP groups and allocate IDs for new entries.
>> @@ -8639,21 +8898,17 @@ build_mcast_groups(struct northd_context *ctx,
>>       */
>>      struct ovn_igmp_group *igmp_group, *igmp_group_next;
>>      HMAP_FOR_EACH_SAFE (igmp_group, igmp_group_next, hmap_node,
>> igmp_groups) {
>> -        if (igmp_group->mcgroup.key == 0) {
>> -            struct mcast_info *mcast_info =
>> &igmp_group->datapath->mcast_info;
>> -            igmp_group->mcgroup.key =
>> ovn_mcast_group_allocate_key(mcast_info);
>> -        }
>>
>> -        /* If we ran out of keys just destroy the entry. */
>> -        if (igmp_group->mcgroup.key == 0) {
>> +        if (!ovn_igmp_group_allocate_id(igmp_group)) {
>> +            /* If we ran out of keys just destroy the entry. */
>>              ovn_igmp_group_destroy(igmp_groups, igmp_group);
>>              continue;
>>          }
>>
>> -        /* Aggregate the ports from all SB entries corresponding to this
>> +        /* Aggregate the ports from all entries corresponding to this
>>           * group.
>>           */
>> -        ovn_igmp_group_aggregate_ports(igmp_group, ports, mcast_groups);
>> +        ovn_igmp_group_aggregate_ports(igmp_group, mcast_groups);
>>      }
>>  }
>>
>> diff --git a/ovn-nb.xml b/ovn-nb.xml
>> index e166190..1f8d751 100644
>> --- a/ovn-nb.xml
>> +++ b/ovn-nb.xml
>> @@ -1527,6 +1527,12 @@
>>            address.
>>          </p>
>>        </column>
>> +      <column name="options" key="mcast_relay" type'{"type": "boolean"}'>
>> +        <p>
>> +          Enables/disables IP multicast relay between logical switches
>> +          connected to the logical router. Default: False.
>> +        </p>
>> +      </column>
>>      </group>
>>
>>      <group title="Common Columns">
>> diff --git a/ovn-sb.xml b/ovn-sb.xml
>> index 17c45bb..02691bb 100644
>> --- a/ovn-sb.xml
>> +++ b/ovn-sb.xml
>> @@ -1017,6 +1017,8 @@
>>          <li><code>eth.mcast</code> expands to
>> <code>eth.dst[40]</code></li>
>>          <li><code>vlan.present</code> expands to
>> <code>vlan.tci[12]</code></li>
>>          <li><code>ip4</code> expands to <code>eth.type ==
>> 0x800</code></li>
>> +        <li><code>ip4.src_mcast</code> expands to
>> +        <code>ip4.src[28..31] == 0xe</code></li>
>>          <li><code>ip4.mcast</code> expands to <code>ip4.dst[28..31] ==
>> 0xe</code></li>
>>          <li><code>ip6</code> expands to <code>eth.type ==
>> 0x86dd</code></li>
>>          <li><code>ip</code> expands to <code>ip4 || ip6</code></li>
>> diff --git a/tests/ovn.at b/tests/ovn.at
>> index 71eb390..52c044c 100644
>> --- a/tests/ovn.at
>> +++ b/tests/ovn.at
>> @@ -14721,12 +14721,12 @@ AT_CHECK([ovn-sbctl get controller_event $uuid
>> seq_num], [0], [dnl
>>  OVN_CLEANUP([hv1], [hv2])
>>  AT_CLEANUP
>>
>> -AT_SETUP([ovn -- IGMP snoop/querier])
>> +AT_SETUP([ovn -- IGMP snoop/querier/relay])
>>  AT_SKIP_IF([test $HAVE_PYTHON = no])
>>  ovn_start
>>
>>  # Logical network:
>> -# Two independent logical switches (sw1 and sw2).
>> +# Three logical switches (sw1-sw3) connected to a logical router (rtr).
>>  # sw1:
>>  #   - subnet 10.0.0.0/8
>>  #   - 2 ports bound on hv1 (sw1-p11, sw1-p12)
>> @@ -14736,6 +14736,10 @@ ovn_start
>>  #   - 1 port bound on hv1 (sw2-p1)
>>  #   - 1 port bound on hv2 (sw2-p2)
>>  #   - IGMP Querier from 20.0.0.254
>> +# sw3:
>> +#   - subnet 30.0.0.0/8
>> +#   - 1 port bound on hv1 (sw3-p1)
>> +#   - 1 port bound on hv2 (sw3-p2)
>>
>>  reset_pcap_file() {
>>      local iface=$1
>> @@ -14812,29 +14816,47 @@ store_igmp_v3_query() {
>>  }
>>
>>  #
>> -# send_ip_multicast_pkt INPORT HV ETH_SRC ETH_DST IP_SRC IP_DST IP_LEN
>> -#    IP_PROTO DATA OUTFILE
>> +# send_ip_multicast_pkt INPORT HV ETH_SRC ETH_DST IP_SRC IP_DST IP_LEN
>> TTL
>> +#    IP_CHKSUM IP_PROTO DATA
>>  #
>>  # This shell function causes an IP multicast packet to be received on
>> INPORT
>>  # of HV.
>>  # The hexdump of the packet is stored in OUTFILE.
>>  #
>>  send_ip_multicast_pkt() {
>> -    local inport=$1 hv=$2 eth_src=$3 eth_dst=$4 ip_src=$5 ip_dst=$6
>> -    local ip_len=$7 ip_chksum=$8 proto=$9 data=${10} outfile=${11}
>> -
>> -    local ip_ttl=20
>> +    local inport=$1 hv=$2 eth_src=$3 eth_dst=$4
>> +    local ip_src=$5 ip_dst=$6 ip_len=$7 ip_ttl=$8 ip_chksum=$9
>> proto=${10}
>> +    local data=${11}
>>
>>      local eth=${eth_dst}${eth_src}0800
>>      local
>> ip=450000${ip_len}95f14000${ip_ttl}${proto}${ip_chksum}${ip_src}${ip_dst}
>>      local packet=${eth}${ip}${data}
>>
>>      as $hv ovs-appctl netdev-dummy/receive ${inport} ${packet}
>> +}
>> +
>> +#
>> +# store_ip_multicast_pkt ETH_SRC ETH_DST IP_SRC IP_DST IP_LEN TTL
>> +#    IP_CHKSUM IP_PROTO DATA OUTFILE
>> +#
>> +# This shell function builds an IP multicast packet and stores the
>> hexdump of
>> +# the packet in OUTFILE.
>> +#
>> +store_ip_multicast_pkt() {
>> +    local eth_src=$1 eth_dst=$2
>> +    local ip_src=$3 ip_dst=$4 ip_len=$5 ip_ttl=$6 ip_chksum=$7 proto=$8
>> +    local data=$9 outfile=${10}
>> +
>> +    local eth=${eth_dst}${eth_src}0800
>> +    local
>> ip=450000${ip_len}95f14000${ip_ttl}${proto}${ip_chksum}${ip_src}${ip_dst}
>> +    local packet=${eth}${ip}${data}
>> +
>>      echo ${packet} >> ${outfile}
>>  }
>>
>>  ovn-nbctl ls-add sw1
>>  ovn-nbctl ls-add sw2
>> +ovn-nbctl ls-add sw3
>>
>>  ovn-nbctl lsp-add sw1 sw1-p11
>>  ovn-nbctl lsp-add sw1 sw1-p12
>> @@ -14842,6 +14864,26 @@ ovn-nbctl lsp-add sw1 sw1-p21
>>  ovn-nbctl lsp-add sw1 sw1-p22
>>  ovn-nbctl lsp-add sw2 sw2-p1
>>  ovn-nbctl lsp-add sw2 sw2-p2
>> +ovn-nbctl lsp-add sw3 sw3-p1
>> +ovn-nbctl lsp-add sw3 sw3-p2
>> +
>> +ovn-nbctl lr-add rtr
>> +ovn-nbctl lrp-add rtr rtr-sw1 00:00:00:00:01:00 10.0.0.254/24
>> +ovn-nbctl lrp-add rtr rtr-sw2 00:00:00:00:02:00 20.0.0.254/24
>> +ovn-nbctl lrp-add rtr rtr-sw3 00:00:00:00:03:00 30.0.0.254/24
>> +
>> +ovn-nbctl lsp-add sw1 sw1-rtr                      \
>> +    -- lsp-set-type sw1-rtr router                 \
>> +    -- lsp-set-addresses sw1-rtr 00:00:00:00:01:00 \
>> +    -- lsp-set-options sw1-rtr router-port=rtr-sw1
>> +ovn-nbctl lsp-add sw2 sw2-rtr                      \
>> +    -- lsp-set-type sw2-rtr router                 \
>> +    -- lsp-set-addresses sw2-rtr 00:00:00:00:02:00 \
>> +    -- lsp-set-options sw2-rtr router-port=rtr-sw2
>> +ovn-nbctl lsp-add sw3 sw3-rtr                      \
>> +    -- lsp-set-type sw3-rtr router                 \
>> +    -- lsp-set-addresses sw3-rtr 00:00:00:00:03:00 \
>> +    -- lsp-set-options sw3-rtr router-port=rtr-sw3
>>
>>  net_add n1
>>  sim_add hv1
>> @@ -14863,6 +14905,11 @@ ovs-vsctl -- add-port br-int hv1-vif3 -- \
>>      options:tx_pcap=hv1/vif3-tx.pcap \
>>      options:rxq_pcap=hv1/vif3-rx.pcap \
>>      ofport-request=1
>> +ovs-vsctl -- add-port br-int hv1-vif4 -- \
>> +    set interface hv1-vif4 external-ids:iface-id=sw3-p1 \
>> +    options:tx_pcap=hv1/vif4-tx.pcap \
>> +    options:rxq_pcap=hv1/vif4-rx.pcap \
>> +    ofport-request=1
>>
>>  sim_add hv2
>>  as hv2
>> @@ -14883,12 +14930,18 @@ ovs-vsctl -- add-port br-int hv2-vif3 -- \
>>      options:tx_pcap=hv2/vif3-tx.pcap \
>>      options:rxq_pcap=hv2/vif3-rx.pcap \
>>      ofport-request=1
>> +ovs-vsctl -- add-port br-int hv2-vif4 -- \
>> +    set interface hv2-vif4 external-ids:iface-id=sw3-p2 \
>> +    options:tx_pcap=hv2/vif4-tx.pcap \
>> +    options:rxq_pcap=hv2/vif4-rx.pcap \
>> +    ofport-request=1
>>
>>  OVN_POPULATE_ARP
>>
>>  # Enable IGMP snooping on sw1.
>> -ovn-nbctl set Logical_Switch sw1 other_config:mcast_querier="false"
>> -ovn-nbctl set Logical_Switch sw1 other_config:mcast_snoop="true"
>> +ovn-nbctl set Logical_Switch sw1       \
>> +    other_config:mcast_querier="false" \
>> +    other_config:mcast_snoop="true"
>>
>>  # No IGMP query should be generated by sw1 (mcast_querier="false").
>>  truncate -s 0 expected
>> @@ -14921,9 +14974,12 @@ truncate -s 0 expected
>>  truncate -s 0 expected_empty
>>  send_ip_multicast_pkt hv1-vif2 hv1 \
>>      000000000001 01005e000144 \
>> -    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e ca70 11 \
>> -    e518e518000a3b3a0000 \
>> -    expected
>> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
>> +    e518e518000a3b3a0000
>> +store_ip_multicast_pkt \
>> +    000000000001 01005e000144 \
>> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
>> +    e518e518000a3b3a0000 expected
>>
>>  OVN_CHECK_PACKETS([hv1/vif1-tx.pcap], [expected])
>>  OVN_CHECK_PACKETS([hv2/vif1-tx.pcap], [expected])
>> @@ -14944,17 +15000,19 @@ OVS_WAIT_UNTIL([
>>      test "${total_entries}" = "1"
>>  ])
>>
>> -# Send traffic traffic and make sure it gets forwarded only on the port
>> that
>> -# joined.
>> +# Send traffic and make sure it gets forwarded only on the port that
>> joined.
>>  as hv1 reset_pcap_file hv1-vif1 hv1/vif1
>>  as hv2 reset_pcap_file hv2-vif1 hv2/vif1
>>  truncate -s 0 expected
>>  truncate -s 0 expected_empty
>>  send_ip_multicast_pkt hv1-vif2 hv1 \
>>      000000000001 01005e000144 \
>> -    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e ca70 11 \
>> -    e518e518000a3b3a0000 \
>> -    expected
>> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
>> +    e518e518000a3b3a0000
>> +store_ip_multicast_pkt \
>> +    000000000001 01005e000144 \
>> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
>> +    e518e518000a3b3a0000 expected
>>
>>  OVN_CHECK_PACKETS([hv1/vif1-tx.pcap], [expected_empty])
>>  OVN_CHECK_PACKETS([hv2/vif1-tx.pcap], [expected])
>> @@ -14988,6 +15046,111 @@ sleep 1
>>  OVN_CHECK_PACKETS([hv1/vif3-tx.pcap], [expected])
>>  OVN_CHECK_PACKETS([hv2/vif3-tx.pcap], [expected])
>>
>> +# Dissable IGMP querier on sw2.
>> +ovn-nbctl set Logical_Switch sw2 \
>> +    other_config:mcast_querier="false"
>> +
>> +# Enable IGMP snooping on sw3.
>> +ovn-nbctl set Logical_Switch sw3       \
>> +    other_config:mcast_querier="false" \
>> +    other_config:mcast_snoop="true"
>> +
>> +# Send traffic from sw3 and make sure rtr doesn't relay it.
>> +truncate -s 0 expected_empty
>> +
>> +as hv1 reset_pcap_file hv1-vif1 hv1/vif1
>> +as hv1 reset_pcap_file hv1-vif2 hv1/vif2
>> +as hv1 reset_pcap_file hv1-vif3 hv1/vif3
>> +as hv1 reset_pcap_file hv1-vif4 hv1/vif4
>> +as hv2 reset_pcap_file hv2-vif1 hv2/vif1
>> +as hv2 reset_pcap_file hv2-vif2 hv2/vif2
>> +as hv2 reset_pcap_file hv2-vif3 hv2/vif3
>> +as hv2 reset_pcap_file hv2-vif4 hv2/vif4
>> +
>> +send_ip_multicast_pkt hv2-vif4 hv2 \
>> +    000000000001 01005e000144 \
>> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
>> +    e518e518000a3b3a0000
>> +
>> +# Sleep a bit to make sure no traffic is received and then check.
>> +sleep 1
>> +OVN_CHECK_PACKETS([hv1/vif1-tx.pcap], [expected_empty])
>> +OVN_CHECK_PACKETS([hv2/vif3-tx.pcap], [expected_empty])
>> +OVN_CHECK_PACKETS([hv1/vif4-tx.pcap], [expected_empty])
>> +OVN_CHECK_PACKETS([hv1/vif2-tx.pcap], [expected_empty])
>> +OVN_CHECK_PACKETS([hv1/vif3-tx.pcap], [expected_empty])
>> +OVN_CHECK_PACKETS([hv2/vif1-tx.pcap], [expected_empty])
>> +OVN_CHECK_PACKETS([hv2/vif2-tx.pcap], [expected_empty])
>> +OVN_CHECK_PACKETS([hv2/vif4-tx.pcap], [expected_empty])
>> +
>> +# Enable IGMP relay on rtr
>> +ovn-nbctl set logical_router rtr \
>> +    options:mcast_relay="true"
>> +
>> +# Inject IGMP Join for 239.0.1.68 on sw1-p11.
>> +send_igmp_v3_report hv1-vif1 hv1 \
>> +    000000000001 $(ip_to_hex 10 0 0 1) f9f8 \
>> +    $(ip_to_hex 239 0 1 68) 04 e9b9 \
>> +    /dev/null
>> +# Inject IGMP Join for 239.0.1.68 on sw2-p2.
>> +send_igmp_v3_report hv2-vif3 hv2 \
>> +    000000000001 $(ip_to_hex 10 0 0 1) f9f8 \
>> +    $(ip_to_hex 239 0 1 68) 04 e9b9 \
>> +    /dev/null
>> +# Inject IGMP Join for 239.0.1.68 on sw3-p1.
>> +send_igmp_v3_report hv1-vif4 hv1 \
>> +    000000000001 $(ip_to_hex 10 0 0 1) f9f8 \
>> +    $(ip_to_hex 239 0 1 68) 04 e9b9 \
>> +    /dev/null
>> +
>> +# Check that the IGMP Group is learned by all switches.
>> +OVS_WAIT_UNTIL([
>> +    total_entries=`ovn-sbctl find IGMP_Group | grep "239.0.1.68" | wc -l`
>> +    test "${total_entries}" = "3"
>> +])
>> +
>> +# Send traffic from sw3 and make sure it is relayed by rtr.
>> +# and ports that joined.
>> +truncate -s 0 expected_routed_sw1
>> +truncate -s 0 expected_routed_sw2
>> +truncate -s 0 expected_switched
>> +truncate -s 0 expected_empty
>> +
>> +as hv1 reset_pcap_file hv1-vif1 hv1/vif1
>> +as hv1 reset_pcap_file hv1-vif2 hv1/vif2
>> +as hv1 reset_pcap_file hv1-vif3 hv1/vif3
>> +as hv1 reset_pcap_file hv1-vif4 hv1/vif4
>> +as hv2 reset_pcap_file hv2-vif1 hv2/vif1
>> +as hv2 reset_pcap_file hv2-vif2 hv2/vif2
>> +as hv2 reset_pcap_file hv2-vif3 hv2/vif3
>> +as hv2 reset_pcap_file hv2-vif4 hv2/vif4
>> +
>> +send_ip_multicast_pkt hv2-vif4 hv2 \
>> +    000000000001 01005e000144 \
>> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
>> +    e518e518000a3b3a0000
>> +store_ip_multicast_pkt \
>> +    000000000100 01005e000144 \
>> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 1f cb70 11 \
>> +    e518e518000a3b3a0000 expected_routed_sw1
>> +store_ip_multicast_pkt \
>> +    000000000200 01005e000144 \
>> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 1f cb70 11 \
>> +    e518e518000a3b3a0000 expected_routed_sw2
>> +store_ip_multicast_pkt \
>> +    000000000001 01005e000144 \
>> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
>> +    e518e518000a3b3a0000 expected_switched
>> +
>> +OVN_CHECK_PACKETS([hv1/vif1-tx.pcap], [expected_routed_sw1])
>> +OVN_CHECK_PACKETS([hv2/vif3-tx.pcap], [expected_routed_sw2])
>> +OVN_CHECK_PACKETS([hv1/vif4-tx.pcap], [expected_switched])
>> +OVN_CHECK_PACKETS([hv1/vif2-tx.pcap], [expected_empty])
>> +OVN_CHECK_PACKETS([hv1/vif3-tx.pcap], [expected_empty])
>> +OVN_CHECK_PACKETS([hv2/vif1-tx.pcap], [expected_empty])
>> +OVN_CHECK_PACKETS([hv2/vif2-tx.pcap], [expected_empty])
>> +OVN_CHECK_PACKETS([hv2/vif4-tx.pcap], [expected_empty])
>> +
>>  OVN_CLEANUP([hv1], [hv2])
>>  AT_CLEANUP
>>
>> --
>> 1.8.3.1
>>
>> _______________________________________________
>> dev mailing list
>> dev@openvswitch.org
>> https://mail.openvswitch.org/mailman/listinfo/ovs-dev
>>
>
Dumitru Ceara Aug. 16, 2019, 1:53 p.m. UTC | #3
On Fri, Aug 16, 2019 at 3:04 PM Numan Siddique <nusiddiq@redhat.com> wrote:
>
>
>
> On Fri, Aug 16, 2019 at 6:23 PM Numan Siddique <nusiddiq@redhat.com> wrote:
>>
>>
>>
>> On Fri, Aug 16, 2019 at 5:51 PM Dumitru Ceara <dceara@redhat.com> wrote:
>>>
>>> Add a new configuration option 'mcast_relay' to the Logical_Router:options
>>> in the OVN Northbound database.
>>>
>>> If a router is configured with 'mcast_relay' enabled then ovn-northd
>>> will install Logical_Flows to allow IP multicast traffic to be routed
>>> between Logical_Switches. The logical router will aggregate all IGMP
>>> groups from attached logical switches and modify the routing pipeline in
>>> the following way:
>>> - Table S_ROUTER_IN_IP_INPUT: add flow allowing IP multicast traffic
>>>   if mcast_relay is enabled on the datapath.
>>> - Table S_ROUTER_IN_IP_ROUTING: add flow matching the group address,
>>>   update TTL and set outport="<Multicast_Group> associated with the
>>>   IGMP group". Continue to next table.
>>> - Table S_ROUTER_IN_ARP_RESOLVE: bypass ARP resolve for IP multicast
>>>   traffic and continue to next table.
>>> - Table S_ROUTER_OUT_DELIVERY: add flow matching IP multicast traffic
>>>   and set ETH.SRC to the MAC address of the logical port on which
>>>   traffic is forwarded.
>>>
>>> Signed-off-by: Dumitru Ceara <dceara@redhat.com>
>>> Acked-by: Mark Michelson <mmichels@redhat.com>
>>
>>
>> Acked-by: Numan Siddique <nusiddiq@redhat.com>
>>
>
> I went ahead and applied to master with the below changes
>
> ****************
> diff --git a/northd/ovn-northd.c b/northd/ovn-northd.c
> index 592d3b9d4..e29b0fff4 100644
> --- a/northd/ovn-northd.c
> +++ b/northd/ovn-northd.c
> @@ -6355,14 +6355,8 @@ build_lrouter_flows(struct hmap *datapaths, struct hmap *ports,
>                        "drop;");
>
>          /* Allow multicast if relay enabled (priority 95). */
> -        ds_clear(&actions);
> -        if (od->mcast_info.rtr.relay) {
> -            ds_put_cstr(&actions, "next;");
> -        } else {
> -            ds_put_cstr(&actions, "drop;");
> -        }
> -        ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_INPUT, 95,
> -                      "ip4.mcast", ds_cstr(&actions));
> +        ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_INPUT, 95, "ip4.mcast",
> +                      od->mcast_info.rtr.relay ? "next;" : "drop;");
>
>          /* ARP reply handling.  Use ARP replies to populate the logical
>           * router's ARP table. */
> **********
>
> Thanks
> Numan

Great, thanks Numan!

>
>>
>> Just one comment inline.
>>
>> Thanks
>> Numan
>>
>>>
>>>
>>> ---
>>> v4:
>>> - Address Numan's comment and use the predicate for ip4.mcast and
>>>   add a new ip4.src_mcast predicate.
>>> v3:
>>> - Address Mark's comment and move setting of the outport in the IP
>>>   Routing stage.
>>> - Update commit message.
>>> - Fix some typos.
>>> v2:
>>> - Optimize flooding to multicast router ports.
>>> - Fix check for source IP multicast in router pipeline.
>>> - Use an enum for OVN_MCAST_*_KEY definitions to avoid hard to debug
>>>   errors due to typos when adding new OVN_MCAST_*_KEY values.
>>> - Fix ovn-northd man page for IGMP.
>>> ---
>>>  NEWS                    |   1 +
>>>  lib/logical-fields.c    |   2 +
>>>  lib/mcast-group-index.h |  13 +-
>>>  northd/ovn-northd.8.xml |  79 +++++++-
>>>  northd/ovn-northd.c     | 505 ++++++++++++++++++++++++++++++++++++------------
>>>  ovn-nb.xml              |   6 +
>>>  ovn-sb.xml              |   2 +
>>>  tests/ovn.at            | 199 +++++++++++++++++--
>>>  8 files changed, 656 insertions(+), 151 deletions(-)
>>>
>>> diff --git a/NEWS b/NEWS
>>> index f476984..73045d6 100644
>>> --- a/NEWS
>>> +++ b/NEWS
>>> @@ -39,6 +39,7 @@ Post-v2.11.0
>>>         logical groups which results in tunnels only been formed between
>>>         members of the same transport zone(s).
>>>       * Support for new logical switch port type - 'virtual'.
>>> +     * Support for IGMP Snooping/Querier and Relay.
>>>     - New QoS type "linux-netem" on Linux.
>>>     - Added support for TLS Server Name Indication (SNI).
>>>
>>> diff --git a/lib/logical-fields.c b/lib/logical-fields.c
>>> index 4ad5bf4..8fb591c 100644
>>> --- a/lib/logical-fields.c
>>> +++ b/lib/logical-fields.c
>>> @@ -156,6 +156,8 @@ ovn_init_symtab(struct shash *symtab)
>>>
>>>      expr_symtab_add_field(symtab, "ip4.src", MFF_IPV4_SRC, "ip4", false);
>>>      expr_symtab_add_field(symtab, "ip4.dst", MFF_IPV4_DST, "ip4", false);
>>> +    expr_symtab_add_predicate(symtab, "ip4.src_mcast",
>>> +                              "ip4.src[28..31] == 0xe");
>>>      expr_symtab_add_predicate(symtab, "ip4.mcast", "ip4.dst[28..31] == 0xe");
>>>
>>>      expr_symtab_add_predicate(symtab, "icmp4", "ip4 && ip.proto == 1");
>>> diff --git a/lib/mcast-group-index.h b/lib/mcast-group-index.h
>>> index 15a1592..cb49ad7 100644
>>> --- a/lib/mcast-group-index.h
>>> +++ b/lib/mcast-group-index.h
>>> @@ -20,8 +20,17 @@ struct ovsdb_idl;
>>>
>>>  struct sbrec_datapath_binding;
>>>
>>> -#define OVN_MCAST_FLOOD_TUNNEL_KEY   65535
>>> -#define OVN_MCAST_UNKNOWN_TUNNEL_KEY (OVN_MCAST_FLOOD_TUNNEL_KEY - 1)
>>> +#define OVN_MIN_MULTICAST 32768
>>> +#define OVN_MAX_MULTICAST 65535
>>> +
>>> +enum ovn_mcast_tunnel_keys {
>>> +
>>> +    OVN_MCAST_FLOOD_TUNNEL_KEY = OVN_MIN_MULTICAST,
>>> +    OVN_MCAST_UNKNOWN_TUNNEL_KEY,
>>> +    OVN_MCAST_MROUTER_FLOOD_TUNNEL_KEY,
>>> +    OVN_MIN_IP_MULTICAST,
>>> +    OVN_MAX_IP_MULTICAST = OVN_MAX_MULTICAST,
>>> +};
>>>
>>>  struct ovsdb_idl_index *mcast_group_index_create(struct ovsdb_idl *);
>>>  const struct sbrec_multicast_group *
>>> diff --git a/northd/ovn-northd.8.xml b/northd/ovn-northd.8.xml
>>> index 6d2fbe3..d45bb15 100644
>>> --- a/northd/ovn-northd.8.xml
>>> +++ b/northd/ovn-northd.8.xml
>>> @@ -947,10 +947,40 @@ output;
>>>
>>>      <ul>
>>>        <li>
>>> -        A priority-100 flow that outputs all packets with an Ethernet broadcast
>>> +        A priority-100 flow that punts all IGMP packets to
>>> +        <code>ovn-controller</code> if IGMP snooping is enabled on the
>>> +        logical switch.
>>> +      </li>
>>> +
>>> +      <li>
>>> +        Priority-90 flows that forward registered IP multicast traffic to
>>> +        their corresponding multicast group, which <code>ovn-northd</code>
>>> +        creates based on learnt <ref table="IGMP_Group" db="OVN_Southbound"/>
>>> +        entries.  The flows also forward packets to the
>>> +        <code>MC_MROUTER_FLOOD</code> multicast group, which
>>> +        <code>ovn-nortdh</code> populates with all the logical ports that
>>> +        are connected to logical routers with
>>> +        <ref column="options" table="Logical_Router"/>:mcast_relay='true'.
>>> +      </li>
>>> +
>>> +      <li>
>>> +        A priority-85 flow that forwards all IP multicast traffic destined to
>>> +        224.0.0.X to the <code>MC_FLOOD</code> multicast group, which
>>> +        <code>ovn-northd</code> populates with all enabled logical ports.
>>> +      </li>
>>> +
>>> +      <li>
>>> +        A priority-80 flow that forwards all unregistered IP multicast traffic
>>> +        to the <code>MC_MROUTER_FLOOD</code> multicast group, if any.
>>> +        Otherwise the flow drops all unregistered IP multicast packets.  This
>>> +        flow is added only if <ref column="other_config"
>>> +        table="Logical_Switch"/>:mcast_flood_unregistered='false'.
>>> +      </li>
>>> +
>>> +      <li>
>>> +        A priority-70 flow that outputs all packets with an Ethernet broadcast
>>>          or multicast <code>eth.dst</code> to the <code>MC_FLOOD</code>
>>> -        multicast group, which <code>ovn-northd</code> populates with all
>>> -        enabled logical ports.
>>> +        multicast group.
>>>        </li>
>>>
>>>        <li>
>>> @@ -1228,6 +1258,14 @@ output;
>>>
>>>        <li>
>>>          <p>
>>> +          A priority-95 flow allows IP multicast traffic if
>>> +          <ref column="options" table="Logical_Router"/>:mcast_relay='true',
>>> +          otherwise drops it.
>>> +        </p>
>>> +      </li>
>>> +
>>> +      <li>
>>> +        <p>
>>>            ICMP echo reply.  These flows reply to ICMP echo requests received
>>>            for the router's IP address.  Let <var>A</var> be an IP address
>>>            owned by a router port.  Then, for each <var>A</var> that is
>>> @@ -1941,6 +1979,16 @@ output;
>>>      <ul>
>>>        <li>
>>>          <p>
>>> +          Priority-500 flows that match IP multicast traffic destined to
>>> +          groups registered on any of the attached switches and sets
>>> +          <code>outport</code> to the associated multicast group that will
>>> +          eventually flood the traffic to all interested attached logical
>>> +          switches. The flows also decrement TTL.
>>> +        </p>
>>> +      </li>
>>> +
>>> +      <li>
>>> +        <p>
>>>            For distributed logical routers where one of the logical router
>>>            ports specifies a <code>redirect-chassis</code>, a priority-400
>>>            logical flow for each ip source/destination couple that matches the
>>> @@ -2074,6 +2122,15 @@ next;
>>>      <ul>
>>>        <li>
>>>          <p>
>>> +          A priority-500 flow that matches IP multicast traffic that was
>>> +          allowed in the routing pipeline. For this kind of traffic the
>>> +          <code>outport</code> was already set so the flow just advances to
>>> +          the next table.
>>> +        </p>
>>> +      </li>
>>> +
>>> +      <li>
>>> +        <p>
>>>            For distributed logical routers where one of the logical router
>>>            ports specifies a <code>redirect-chassis</code>, a priority-400
>>>            logical flow with match <code>REGBIT_DISTRIBUTED_NAT == 1</code>
>>> @@ -2641,9 +2698,19 @@ clone {
>>>      <h3>Egress Table 3: Delivery</h3>
>>>
>>>      <p>
>>> -      Packets that reach this table are ready for delivery.  It contains
>>> -      priority-100 logical flows that match packets on each enabled logical
>>> -      router port, with action <code>output;</code>.
>>> +      Packets that reach this table are ready for delivery.  It contains:
>>> +      <ul>
>>> +        <li>
>>> +          Priority-110 logical flows that match IP multicast packets on each
>>> +          enabled logical router port and modify the Ethernet source address
>>> +          of the packets to the Ethernet address of the port and then execute
>>> +          action <code>output;</code>.
>>> +        </li>
>>> +        <li>
>>> +          Priority-100 logical flows that match packets on each enabled
>>> +          logical router port, with action <code>output;</code>.
>>> +        </li>
>>> +      </ul>
>>>      </p>
>>>
>>>  </manpage>
>>> diff --git a/northd/ovn-northd.c b/northd/ovn-northd.c
>>> index e861344..592d3b9 100644
>>> --- a/northd/ovn-northd.c
>>> +++ b/northd/ovn-northd.c
>>> @@ -433,32 +433,52 @@ struct ipam_info {
>>>      bool mac_only;
>>>  };
>>>
>>> -#define OVN_MIN_MULTICAST 32768
>>> -#define OVN_MAX_MULTICAST OVN_MCAST_FLOOD_TUNNEL_KEY
>>> -BUILD_ASSERT_DECL(OVN_MIN_MULTICAST < OVN_MAX_MULTICAST);
>>> -
>>> -#define OVN_MIN_IP_MULTICAST OVN_MIN_MULTICAST
>>> -#define OVN_MAX_IP_MULTICAST (OVN_MCAST_UNKNOWN_TUNNEL_KEY - 1)
>>> -BUILD_ASSERT_DECL(OVN_MAX_IP_MULTICAST >= OVN_MIN_MULTICAST);
>>> -
>>>  /*
>>>   * Multicast snooping and querier per datapath configuration.
>>>   */
>>> +struct mcast_switch_info {
>>> +
>>> +    bool enabled;               /* True if snooping enabled. */
>>> +    bool querier;               /* True if querier enabled. */
>>> +    bool flood_unregistered;    /* True if unregistered multicast should be
>>> +                                 * flooded.
>>> +                                 */
>>> +    bool flood_relay;           /* True if the switch is connected to a
>>> +                                 * multicast router and unregistered multicast
>>> +                                 * should be flooded to the mrouter. Only
>>> +                                 * applicable if flood_unregistered == false.
>>> +                                 */
>>> +
>>> +    int64_t table_size;         /* Max number of IP multicast groups. */
>>> +    int64_t idle_timeout;       /* Timeout after which an idle group is
>>> +                                 * flushed.
>>> +                                 */
>>> +    int64_t query_interval;     /* Interval between multicast queries. */
>>> +    char *eth_src;              /* ETH src address of the multicast queries. */
>>> +    char *ipv4_src;             /* IP src address of the multicast queries. */
>>> +    int64_t query_max_response; /* Expected time after which reports should
>>> +                                 * be received for queries that were sent out.
>>> +                                 */
>>> +
>>> +    uint32_t active_flows;      /* Current number of active IP multicast
>>> +                                 * flows.
>>> +                                 */
>>> +};
>>> +
>>> +struct mcast_router_info {
>>> +    bool relay; /* True if the router should relay IP multicast. */
>>> +};
>>> +
>>>  struct mcast_info {
>>> -    bool enabled;
>>> -    bool querier;
>>> -    bool flood_unregistered;
>>> -
>>> -    int64_t table_size;
>>> -    int64_t idle_timeout;
>>> -    int64_t query_interval;
>>> -    char *eth_src;
>>> -    char *ipv4_src;
>>> -    int64_t  query_max_response;
>>> -
>>> -    struct hmap group_tnlids;
>>> -    uint32_t group_tnlid_hint;
>>> -    uint32_t active_flows;
>>> +
>>> +    struct hmap group_tnlids;  /* Group tunnel IDs in use on this DP. */
>>> +    uint32_t group_tnlid_hint; /* Hint for allocating next group tunnel ID. */
>>> +    struct ovs_list groups;    /* List of groups learnt on this DP. */
>>> +
>>> +    union {
>>> +        struct mcast_switch_info sw;  /* Switch specific multicast info. */
>>> +        struct mcast_router_info rtr; /* Router specific multicast info. */
>>> +    };
>>>  };
>>>
>>>  static uint32_t
>>> @@ -559,6 +579,7 @@ ovn_datapath_create(struct hmap *datapaths, const struct uuid *key,
>>>  }
>>>
>>>  static void ovn_ls_port_group_destroy(struct hmap *nb_pgs);
>>> +static void destroy_mcast_info_for_datapath(struct ovn_datapath *od);
>>>
>>>  static void
>>>  ovn_datapath_destroy(struct hmap *datapaths, struct ovn_datapath *od)
>>> @@ -572,12 +593,7 @@ ovn_datapath_destroy(struct hmap *datapaths, struct ovn_datapath *od)
>>>          bitmap_free(od->ipam_info.allocated_ipv4s);
>>>          free(od->router_ports);
>>>          ovn_ls_port_group_destroy(&od->nb_pgs);
>>> -
>>> -        if (od->nbs) {
>>> -            free(od->mcast_info.eth_src);
>>> -            free(od->mcast_info.ipv4_src);
>>> -            destroy_tnlids(&od->mcast_info.group_tnlids);
>>> -        }
>>> +        destroy_mcast_info_for_datapath(od);
>>>
>>>          free(od);
>>>      }
>>> @@ -714,23 +730,28 @@ init_ipam_info_for_datapath(struct ovn_datapath *od)
>>>  }
>>>
>>>  static void
>>> -init_mcast_info_for_datapath(struct ovn_datapath *od)
>>> +init_mcast_info_for_router_datapath(struct ovn_datapath *od)
>>>  {
>>> -    if (!od->nbs) {
>>> -        return;
>>> -    }
>>> +    struct mcast_router_info *mcast_rtr_info = &od->mcast_info.rtr;
>>>
>>> -    struct mcast_info *mcast_info = &od->mcast_info;
>>> +    mcast_rtr_info->relay = smap_get_bool(&od->nbr->options, "mcast_relay",
>>> +                                          false);
>>> +}
>>>
>>> -    mcast_info->enabled =
>>> +static void
>>> +init_mcast_info_for_switch_datapath(struct ovn_datapath *od)
>>> +{
>>> +    struct mcast_switch_info *mcast_sw_info = &od->mcast_info.sw;
>>> +
>>> +    mcast_sw_info->enabled =
>>>          smap_get_bool(&od->nbs->other_config, "mcast_snoop", false);
>>> -    mcast_info->querier =
>>> +    mcast_sw_info->querier =
>>>          smap_get_bool(&od->nbs->other_config, "mcast_querier", true);
>>> -    mcast_info->flood_unregistered =
>>> +    mcast_sw_info->flood_unregistered =
>>>          smap_get_bool(&od->nbs->other_config, "mcast_flood_unregistered",
>>>                        false);
>>>
>>> -    mcast_info->table_size =
>>> +    mcast_sw_info->table_size =
>>>          smap_get_ullong(&od->nbs->other_config, "mcast_table_size",
>>>                          OVN_MCAST_DEFAULT_MAX_ENTRIES);
>>>
>>> @@ -742,54 +763,94 @@ init_mcast_info_for_datapath(struct ovn_datapath *od)
>>>      } else if (idle_timeout > OVN_MCAST_MAX_IDLE_TIMEOUT_S) {
>>>          idle_timeout = OVN_MCAST_MAX_IDLE_TIMEOUT_S;
>>>      }
>>> -    mcast_info->idle_timeout = idle_timeout;
>>> +    mcast_sw_info->idle_timeout = idle_timeout;
>>>
>>>      uint32_t query_interval =
>>>          smap_get_ullong(&od->nbs->other_config, "mcast_query_interval",
>>> -                        mcast_info->idle_timeout / 2);
>>> +                        mcast_sw_info->idle_timeout / 2);
>>>      if (query_interval < OVN_MCAST_MIN_QUERY_INTERVAL_S) {
>>>          query_interval = OVN_MCAST_MIN_QUERY_INTERVAL_S;
>>>      } else if (query_interval > OVN_MCAST_MAX_QUERY_INTERVAL_S) {
>>>          query_interval = OVN_MCAST_MAX_QUERY_INTERVAL_S;
>>>      }
>>> -    mcast_info->query_interval = query_interval;
>>> +    mcast_sw_info->query_interval = query_interval;
>>>
>>> -    mcast_info->eth_src =
>>> +    mcast_sw_info->eth_src =
>>>          nullable_xstrdup(smap_get(&od->nbs->other_config, "mcast_eth_src"));
>>> -    mcast_info->ipv4_src =
>>> +    mcast_sw_info->ipv4_src =
>>>          nullable_xstrdup(smap_get(&od->nbs->other_config, "mcast_ip4_src"));
>>>
>>> -    mcast_info->query_max_response =
>>> +    mcast_sw_info->query_max_response =
>>>          smap_get_ullong(&od->nbs->other_config, "mcast_query_max_response",
>>>                          OVN_MCAST_DEFAULT_QUERY_MAX_RESPONSE_S);
>>>
>>> -    hmap_init(&mcast_info->group_tnlids);
>>> -    mcast_info->group_tnlid_hint = OVN_MIN_IP_MULTICAST;
>>> -    mcast_info->active_flows = 0;
>>> +    mcast_sw_info->active_flows = 0;
>>> +}
>>> +
>>> +static void
>>> +init_mcast_info_for_datapath(struct ovn_datapath *od)
>>> +{
>>> +    if (!od->nbr && !od->nbs) {
>>> +        return;
>>> +    }
>>> +
>>> +    hmap_init(&od->mcast_info.group_tnlids);
>>> +    od->mcast_info.group_tnlid_hint = OVN_MIN_IP_MULTICAST;
>>> +    ovs_list_init(&od->mcast_info.groups);
>>> +
>>> +    if (od->nbs) {
>>> +        init_mcast_info_for_switch_datapath(od);
>>> +    } else {
>>> +        init_mcast_info_for_router_datapath(od);
>>> +    }
>>> +}
>>> +
>>> +static void
>>> +destroy_mcast_info_for_switch_datapath(struct ovn_datapath *od)
>>> +{
>>> +    struct mcast_switch_info *mcast_sw_info = &od->mcast_info.sw;
>>> +
>>> +    free(mcast_sw_info->eth_src);
>>> +    free(mcast_sw_info->ipv4_src);
>>> +}
>>> +
>>> +static void
>>> +destroy_mcast_info_for_datapath(struct ovn_datapath *od)
>>> +{
>>> +    if (!od->nbr && !od->nbs) {
>>> +        return;
>>> +    }
>>> +
>>> +    if (od->nbs) {
>>> +        destroy_mcast_info_for_switch_datapath(od);
>>> +    }
>>> +
>>> +    destroy_tnlids(&od->mcast_info.group_tnlids);
>>>  }
>>>
>>>  static void
>>> -store_mcast_info_for_datapath(const struct sbrec_ip_multicast *sb,
>>> -                              struct ovn_datapath *od)
>>> +store_mcast_info_for_switch_datapath(const struct sbrec_ip_multicast *sb,
>>> +                                     struct ovn_datapath *od)
>>>  {
>>> -    struct mcast_info *mcast_info = &od->mcast_info;
>>> +    struct mcast_switch_info *mcast_sw_info = &od->mcast_info.sw;
>>>
>>>      sbrec_ip_multicast_set_datapath(sb, od->sb);
>>> -    sbrec_ip_multicast_set_enabled(sb, &mcast_info->enabled, 1);
>>> -    sbrec_ip_multicast_set_querier(sb, &mcast_info->querier, 1);
>>> -    sbrec_ip_multicast_set_table_size(sb, &mcast_info->table_size, 1);
>>> -    sbrec_ip_multicast_set_idle_timeout(sb, &mcast_info->idle_timeout, 1);
>>> +    sbrec_ip_multicast_set_enabled(sb, &mcast_sw_info->enabled, 1);
>>> +    sbrec_ip_multicast_set_querier(sb, &mcast_sw_info->querier, 1);
>>> +    sbrec_ip_multicast_set_table_size(sb, &mcast_sw_info->table_size, 1);
>>> +    sbrec_ip_multicast_set_idle_timeout(sb, &mcast_sw_info->idle_timeout, 1);
>>>      sbrec_ip_multicast_set_query_interval(sb,
>>> -                                          &mcast_info->query_interval, 1);
>>> +                                          &mcast_sw_info->query_interval, 1);
>>>      sbrec_ip_multicast_set_query_max_resp(sb,
>>> -                                          &mcast_info->query_max_response, 1);
>>> +                                          &mcast_sw_info->query_max_response,
>>> +                                          1);
>>>
>>> -    if (mcast_info->eth_src) {
>>> -        sbrec_ip_multicast_set_eth_src(sb, mcast_info->eth_src);
>>> +    if (mcast_sw_info->eth_src) {
>>> +        sbrec_ip_multicast_set_eth_src(sb, mcast_sw_info->eth_src);
>>>      }
>>>
>>> -    if (mcast_info->ipv4_src) {
>>> -        sbrec_ip_multicast_set_ip4_src(sb, mcast_info->ipv4_src);
>>> +    if (mcast_sw_info->ipv4_src) {
>>> +        sbrec_ip_multicast_set_ip4_src(sb, mcast_sw_info->ipv4_src);
>>>      }
>>>  }
>>>
>>> @@ -906,6 +967,7 @@ join_datapaths(struct northd_context *ctx, struct hmap *datapaths,
>>>                                       NULL, nbr, NULL);
>>>              ovs_list_push_back(nb_only, &od->list);
>>>          }
>>> +        init_mcast_info_for_datapath(od);
>>>          ovs_list_push_back(lr_list, &od->lr_list);
>>>      }
>>>  }
>>> @@ -1999,6 +2061,13 @@ join_logical_ports(struct northd_context *ctx,
>>>                      break;
>>>                  }
>>>              }
>>> +
>>> +            /* If the router is multicast enabled then set relay on the switch
>>> +             * datapath.
>>> +             */
>>> +            if (peer->od && peer->od->mcast_info.rtr.relay) {
>>> +                op->od->mcast_info.sw.flood_relay = true;
>>> +            }
>>>          } else if (op->nbrp && op->nbrp->peer && !op->derived) {
>>>              struct ovn_port *peer = ovn_port_find(ports, op->nbrp->peer);
>>>              if (peer) {
>>> @@ -2846,6 +2915,10 @@ struct multicast_group {
>>>  static const struct multicast_group mc_flood =
>>>      { MC_FLOOD, OVN_MCAST_FLOOD_TUNNEL_KEY };
>>>
>>> +#define MC_MROUTER_FLOOD "_MC_mrouter_flood"
>>> +static const struct multicast_group mc_mrouter_flood =
>>> +    { MC_MROUTER_FLOOD, OVN_MCAST_MROUTER_FLOOD_TUNNEL_KEY };
>>> +
>>>  #define MC_UNKNOWN "_MC_unknown"
>>>  static const struct multicast_group mc_unknown =
>>>      { MC_UNKNOWN, OVN_MCAST_UNKNOWN_TUNNEL_KEY };
>>> @@ -2955,7 +3028,8 @@ ovn_multicast_update_sbrec(const struct ovn_multicast *mc,
>>>   */
>>>  struct ovn_igmp_group_entry {
>>>      struct ovs_list list_node; /* Linkage in the list of entries. */
>>> -    const struct sbrec_igmp_group *sb;
>>> +    size_t n_ports;
>>> +    struct ovn_port **ports;
>>>  };
>>>
>>>  /*
>>> @@ -2964,12 +3038,13 @@ struct ovn_igmp_group_entry {
>>>   */
>>>  struct ovn_igmp_group {
>>>      struct hmap_node hmap_node; /* Index on 'datapath' and 'address'. */
>>> +    struct ovs_list list_node;  /* Linkage in the per-dp igmp group list. */
>>>
>>>      struct ovn_datapath *datapath;
>>>      struct in6_addr address; /* Multicast IPv6-mapped-IPv4 or IPv4 address. */
>>>      struct multicast_group mcgroup;
>>>
>>> -    struct ovs_list sb_entries; /* List of SB entries for this group. */
>>> +    struct ovs_list entries; /* List of SB entries for this group. */
>>>  };
>>>
>>>  static uint32_t
>>> @@ -2997,77 +3072,120 @@ ovn_igmp_group_find(struct hmap *igmp_groups,
>>>      return NULL;
>>>  }
>>>
>>> -static void
>>> +static struct ovn_igmp_group *
>>>  ovn_igmp_group_add(struct northd_context *ctx, struct hmap *igmp_groups,
>>>                     struct ovn_datapath *datapath,
>>> -                   const struct sbrec_igmp_group *sb_igmp_group)
>>> +                   const struct in6_addr *address,
>>> +                   const char *address_s)
>>>  {
>>> -    struct in6_addr group_address;
>>> -    ovs_be32 ipv4;
>>> -
>>> -    if (ip_parse(sb_igmp_group->address, &ipv4)) {
>>> -        group_address = in6_addr_mapped_ipv4(ipv4);
>>> -    } else if (!ipv6_parse(sb_igmp_group->address, &group_address)) {
>>> -        static struct vlog_rate_limit rl = VLOG_RATE_LIMIT_INIT(1, 1);
>>> -        VLOG_WARN_RL(&rl, "invalid IGMP group address: %s",
>>> -                     sb_igmp_group->address);
>>> -        return;
>>> -    }
>>> -
>>>      struct ovn_igmp_group *igmp_group =
>>> -        ovn_igmp_group_find(igmp_groups, datapath, &group_address);
>>> +        ovn_igmp_group_find(igmp_groups, datapath, address);
>>>
>>>      if (!igmp_group) {
>>>          igmp_group = xmalloc(sizeof *igmp_group);
>>>
>>>          const struct sbrec_multicast_group *mcgroup =
>>> -            mcast_group_lookup(ctx->sbrec_mcast_group_by_name_dp,
>>> -                               sb_igmp_group->address, datapath->sb);
>>> +            mcast_group_lookup(ctx->sbrec_mcast_group_by_name_dp, address_s,
>>> +                               datapath->sb);
>>>
>>>          igmp_group->datapath = datapath;
>>> -        igmp_group->address = group_address;
>>> +        igmp_group->address = *address;
>>>          if (mcgroup) {
>>>              igmp_group->mcgroup.key = mcgroup->tunnel_key;
>>>              add_tnlid(&datapath->mcast_info.group_tnlids, mcgroup->tunnel_key);
>>>          } else {
>>>              igmp_group->mcgroup.key = 0;
>>>          }
>>> -        igmp_group->mcgroup.name = sb_igmp_group->address;
>>> -        ovs_list_init(&igmp_group->sb_entries);
>>> +        igmp_group->mcgroup.name = address_s;
>>> +        ovs_list_init(&igmp_group->entries);
>>>
>>>          hmap_insert(igmp_groups, &igmp_group->hmap_node,
>>> -                    ovn_igmp_group_hash(datapath, &group_address));
>>> +                    ovn_igmp_group_hash(datapath, address));
>>> +        ovs_list_push_back(&datapath->mcast_info.groups,
>>> +                           &igmp_group->list_node);
>>> +    }
>>> +
>>> +    return igmp_group;
>>> +}
>>> +
>>> +static bool
>>> +ovn_igmp_group_get_address(const struct sbrec_igmp_group *sb_igmp_group,
>>> +                           struct in6_addr *address)
>>> +{
>>> +    ovs_be32 ipv4;
>>> +
>>> +    if (ip_parse(sb_igmp_group->address, &ipv4)) {
>>> +        *address = in6_addr_mapped_ipv4(ipv4);
>>> +        return true;
>>> +    }
>>> +    if (!ipv6_parse(sb_igmp_group->address, address)) {
>>> +        return false;
>>>      }
>>> +    return true;
>>> +}
>>>
>>> +static struct ovn_port **
>>> +ovn_igmp_group_get_ports(const struct sbrec_igmp_group *sb_igmp_group,
>>> +                         size_t *n_ports, struct hmap *ovn_ports)
>>> +{
>>> +    struct ovn_port **ports = xmalloc(sb_igmp_group->n_ports * sizeof *ports);
>>> +
>>> +     *n_ports = 0;
>>> +     for (size_t i = 0; i < sb_igmp_group->n_ports; i++) {
>>> +        ports[(*n_ports)] =
>>> +            ovn_port_find(ovn_ports, sb_igmp_group->ports[i]->logical_port);
>>> +        if (ports[(*n_ports)]) {
>>> +            (*n_ports)++;
>>> +        }
>>> +    }
>>> +
>>> +    return ports;
>>> +}
>>> +
>>> +static void
>>> +ovn_igmp_group_add_entry(struct ovn_igmp_group *igmp_group,
>>> +                         struct ovn_port **ports, size_t n_ports)
>>> +{
>>>      struct ovn_igmp_group_entry *entry = xmalloc(sizeof *entry);
>>>
>>> -    entry->sb = sb_igmp_group;
>>> -    ovs_list_push_back(&igmp_group->sb_entries , &entry->list_node);
>>> +    entry->ports = ports;
>>> +    entry->n_ports = n_ports;
>>> +    ovs_list_push_back(&igmp_group->entries, &entry->list_node);
>>> +}
>>> +
>>> +static void
>>> +ovn_igmp_group_destroy_entry(struct ovn_igmp_group_entry *entry)
>>> +{
>>> +    free(entry->ports);
>>> +}
>>> +
>>> +static bool
>>> +ovn_igmp_group_allocate_id(struct ovn_igmp_group *igmp_group)
>>> +{
>>> +    if (igmp_group->mcgroup.key == 0) {
>>> +        struct mcast_info *mcast_info = &igmp_group->datapath->mcast_info;
>>> +        igmp_group->mcgroup.key = ovn_mcast_group_allocate_key(mcast_info);
>>> +    }
>>> +
>>> +    if (igmp_group->mcgroup.key == 0) {
>>> +        return false;
>>> +    }
>>> +
>>> +    return true;
>>>  }
>>>
>>>  static void
>>>  ovn_igmp_group_aggregate_ports(struct ovn_igmp_group *igmp_group,
>>> -                               struct hmap *ovn_ports,
>>>                                 struct hmap *mcast_groups)
>>>  {
>>>      struct ovn_igmp_group_entry *entry;
>>>
>>> -    LIST_FOR_EACH_POP (entry, list_node, &igmp_group->sb_entries) {
>>> -        size_t n_oports = 0;
>>> -        struct ovn_port **oports =
>>> -            xmalloc(entry->sb->n_ports * sizeof *oports);
>>> -
>>> -        for (size_t i = 0; i < entry->sb->n_ports; i++) {
>>> -            oports[n_oports] =
>>> -                ovn_port_find(ovn_ports, entry->sb->ports[i]->logical_port);
>>> -            if (oports[n_oports]) {
>>> -                n_oports++;
>>> -            }
>>> -        }
>>> -
>>> +    LIST_FOR_EACH_POP (entry, list_node, &igmp_group->entries) {
>>>          ovn_multicast_add_ports(mcast_groups, igmp_group->datapath,
>>> -                                &igmp_group->mcgroup, oports, n_oports);
>>> -        free(oports);
>>> +                                &igmp_group->mcgroup, entry->ports,
>>> +                                entry->n_ports);
>>> +
>>> +        ovn_igmp_group_destroy_entry(entry);
>>>          free(entry);
>>>      }
>>>  }
>>> @@ -3079,10 +3197,12 @@ ovn_igmp_group_destroy(struct hmap *igmp_groups,
>>>      if (igmp_group) {
>>>          struct ovn_igmp_group_entry *entry;
>>>
>>> -        LIST_FOR_EACH_POP (entry, list_node, &igmp_group->sb_entries) {
>>> +        LIST_FOR_EACH_POP (entry, list_node, &igmp_group->entries) {
>>> +            ovn_igmp_group_destroy_entry(entry);
>>>              free(entry);
>>>          }
>>>          hmap_remove(igmp_groups, &igmp_group->hmap_node);
>>> +        ovs_list_remove(&igmp_group->list_node);
>>>          free(igmp_group);
>>>      }
>>>  }
>>> @@ -5286,7 +5406,9 @@ build_lswitch_flows(struct hmap *datapaths, struct hmap *ports,
>>>              continue;
>>>          }
>>>
>>> -        if (od->mcast_info.enabled) {
>>> +        struct mcast_switch_info *mcast_sw_info = &od->mcast_info.sw;
>>> +
>>> +        if (mcast_sw_info->enabled) {
>>>              /* Punt IGMP traffic to controller. */
>>>              ovn_lflow_add(lflows, od, S_SWITCH_IN_L2_LKUP, 100,
>>>                            "ip4 && ip.proto == 2", "igmp;");
>>> @@ -5299,9 +5421,16 @@ build_lswitch_flows(struct hmap *datapaths, struct hmap *ports,
>>>                            "outport = \""MC_FLOOD"\"; output;");
>>>
>>>              /* Drop unregistered IP multicast if not allowed. */
>>> -            if (!od->mcast_info.flood_unregistered) {
>>> -                ovn_lflow_add(lflows, od, S_SWITCH_IN_L2_LKUP, 80,
>>> -                              "ip4 && ip4.mcast", "drop;");
>>> +            if (!mcast_sw_info->flood_unregistered) {
>>> +                /* Forward unregistered IP multicast to mrouter (if any). */
>>> +                if (mcast_sw_info->flood_relay) {
>>> +                    ovn_lflow_add(lflows, od, S_SWITCH_IN_L2_LKUP, 80,
>>> +                                  "ip4 && ip4.mcast",
>>> +                                  "outport = \""MC_MROUTER_FLOOD"\"; output;");
>>> +                } else {
>>> +                    ovn_lflow_add(lflows, od, S_SWITCH_IN_L2_LKUP, 80,
>>> +                                  "ip4 && ip4.mcast", "drop;");
>>> +                }
>>>              }
>>>          }
>>>
>>> @@ -5318,18 +5447,26 @@ build_lswitch_flows(struct hmap *datapaths, struct hmap *ports,
>>>              continue;
>>>          }
>>>
>>> -        struct mcast_info *mcast_info = &igmp_group->datapath->mcast_info;
>>> +        struct mcast_switch_info *mcast_sw_info =
>>> +            &igmp_group->datapath->mcast_info.sw;
>>>
>>> -        if (mcast_info->active_flows >= mcast_info->table_size) {
>>> +        if (mcast_sw_info->active_flows >= mcast_sw_info->table_size) {
>>>              continue;
>>>          }
>>> -        mcast_info->active_flows++;
>>> +        mcast_sw_info->active_flows++;
>>>
>>>          ds_clear(&match);
>>>          ds_clear(&actions);
>>>
>>>          ds_put_format(&match, "eth.mcast && ip4 && ip4.dst == %s ",
>>>                        igmp_group->mcgroup.name);
>>> +        /* Also flood traffic to all multicast routers with relay enabled. */
>>> +        if (mcast_sw_info->flood_relay) {
>>> +            ds_put_cstr(&actions,
>>> +                        "clone { "
>>> +                            "outport = \""MC_MROUTER_FLOOD "\"; output; "
>>> +                        "};");
>>> +        }
>>>          ds_put_format(&actions, "outport = \"%s\"; output; ",
>>>                        igmp_group->mcgroup.name);
>>>
>>> @@ -6209,7 +6346,7 @@ build_lrouter_flows(struct hmap *datapaths, struct hmap *ports,
>>>           * source or destination, and zero network source or destination
>>>           * (priority 100). */
>>>          ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_INPUT, 100,
>>> -                      "ip4.mcast || "
>>> +                      "ip4.src_mcast ||"
>>>                        "ip4.src == 255.255.255.255 || "
>>>                        "ip4.src == 127.0.0.0/8 || "
>>>                        "ip4.dst == 127.0.0.0/8 || "
>>> @@ -6217,6 +6354,16 @@ build_lrouter_flows(struct hmap *datapaths, struct hmap *ports,
>>>                        "ip4.dst == 0.0.0.0/8",
>>>                        "drop;");
>>>
>>> +        /* Allow multicast if relay enabled (priority 95). */
>>> +        ds_clear(&actions);
>>> +        if (od->mcast_info.rtr.relay) {
>>> +            ds_put_cstr(&actions, "next;");
>>> +        } else {
>>> +            ds_put_cstr(&actions, "drop;");
>>> +        }
>>> +        ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_INPUT, 95,
>>> +                      "ip4.mcast", ds_cstr(&actions));
>>> +
>>
>>
>> Instead of if/else how about
>> ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_INPUT, 95, "ip4.mcast",  od->mcast_info.rtr.relay ? "next": "drop");
>>
>>
>>>          /* ARP reply handling.  Use ARP replies to populate the logical
>>>           * router's ARP table. */
>>>          ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_INPUT, 90, "arp.op == 2",
>>> @@ -7487,6 +7634,27 @@ build_lrouter_flows(struct hmap *datapaths, struct hmap *ports,
>>>          }
>>>      }
>>>
>>> +    /* IP Multicast lookup. Here we set the output port, adjust TTL and
>>> +     * advance to next table (priority 500).
>>> +     */
>>> +    HMAP_FOR_EACH (od, key_node, datapaths) {
>>> +        if (!od->nbr || !od->mcast_info.rtr.relay) {
>>> +            continue;
>>> +        }
>>> +        struct ovn_igmp_group *igmp_group;
>>> +
>>> +        LIST_FOR_EACH (igmp_group, list_node, &od->mcast_info.groups) {
>>> +            ds_clear(&match);
>>> +            ds_clear(&actions);
>>> +            ds_put_format(&match, "ip4 && ip4.dst == %s ",
>>> +                          igmp_group->mcgroup.name);
>>> +            ds_put_format(&actions, "outport = \"%s\"; ip.ttl--; next;",
>>> +                          igmp_group->mcgroup.name);
>>> +            ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_ROUTING, 500,
>>> +                          ds_cstr(&match), ds_cstr(&actions));
>>> +        }
>>> +    }
>>> +
>>>      /* Logical router ingress table 8: Policy.
>>>       *
>>>       * A packet that arrives at this table is an IP packet that should be
>>> @@ -7517,10 +7685,24 @@ build_lrouter_flows(struct hmap *datapaths, struct hmap *ports,
>>>
>>>      /* Local router ingress table 9: ARP Resolution.
>>>       *
>>> -     * Any packet that reaches this table is an IP packet whose next-hop IP
>>> -     * address is in reg0. (ip4.dst is the final destination.) This table
>>> -     * resolves the IP address in reg0 into an output port in outport and an
>>> -     * Ethernet address in eth.dst. */
>>> +     * Multicast packets already have the outport set so just advance to next
>>> +     * table (priority 500). */
>>> +    HMAP_FOR_EACH (od, key_node, datapaths) {
>>> +        if (!od->nbr) {
>>> +            continue;
>>> +        }
>>> +
>>> +        ovn_lflow_add(lflows, od, S_ROUTER_IN_ARP_RESOLVE, 500,
>>> +                      "ip4.mcast", "next;");
>>> +    }
>>> +
>>> +    /* Local router ingress table 9: ARP Resolution.
>>> +     *
>>> +     * Any unicast packet that reaches this table is an IP packet whose
>>> +     * next-hop IP address is in reg0. (ip4.dst is the final destination.)
>>> +     * This table resolves the IP address in reg0 into an output port in
>>> +     * outport and an Ethernet address in eth.dst.
>>> +     */
>>>      HMAP_FOR_EACH (op, key_node, ports) {
>>>          if (op->nbsp && !lsp_is_enabled(op->nbsp)) {
>>>              continue;
>>> @@ -8002,9 +8184,13 @@ build_lrouter_flows(struct hmap *datapaths, struct hmap *ports,
>>>          ovn_lflow_add(lflows, od, S_ROUTER_IN_ARP_REQUEST, 0, "1", "output;");
>>>      }
>>>
>>> -    /* Logical router egress table 1: Delivery (priority 100).
>>> +    /* Logical router egress table 1: Delivery (priority 100-110).
>>>       *
>>> -     * Priority 100 rules deliver packets to enabled logical ports. */
>>> +     * Priority 100 rules deliver packets to enabled logical ports.
>>> +     * Priority 110 rules match multicast packets and update the source
>>> +     * mac before delivering to enabled logical ports. IP multicast traffic
>>> +     * bypasses S_ROUTER_IN_IP_ROUTING route lookups.
>>> +     */
>>>      HMAP_FOR_EACH (op, key_node, ports) {
>>>          if (!op->nbrp) {
>>>              continue;
>>> @@ -8024,6 +8210,20 @@ build_lrouter_flows(struct hmap *datapaths, struct hmap *ports,
>>>              continue;
>>>          }
>>>
>>> +        /* If multicast relay is enabled then also adjust source mac for IP
>>> +         * multicast traffic.
>>> +         */
>>> +        if (op->od->mcast_info.rtr.relay) {
>>> +            ds_clear(&match);
>>> +            ds_clear(&actions);
>>> +            ds_put_format(&match, "ip4.mcast && outport == %s",
>>> +                          op->json_key);
>>> +            ds_put_format(&actions, "eth.src = %s; output;",
>>> +                          op->lrp_networks.ea_s);
>>> +            ovn_lflow_add(lflows, op->od, S_ROUTER_OUT_DELIVERY, 110,
>>> +                        ds_cstr(&match), ds_cstr(&actions));
>>> +        }
>>> +
>>>          ds_clear(&match);
>>>          ds_put_format(&match, "outport == %s", op->json_key);
>>>          ovn_lflow_add(lflows, op->od, S_ROUTER_OUT_DELIVERY, 100,
>>> @@ -8574,7 +8774,7 @@ build_ip_mcast(struct northd_context *ctx, struct hmap *datapaths)
>>>          if (!ip_mcast) {
>>>              ip_mcast = sbrec_ip_multicast_insert(ctx->ovnsb_txn);
>>>          }
>>> -        store_mcast_info_for_datapath(ip_mcast, od);
>>> +        store_mcast_info_for_switch_datapath(ip_mcast, od);
>>>      }
>>>
>>>      /* Delete southbound records without northbound matches. */
>>> @@ -8606,6 +8806,14 @@ build_mcast_groups(struct northd_context *ctx,
>>>
>>>          if (lsp_is_enabled(op->nbsp)) {
>>>              ovn_multicast_add(mcast_groups, &mc_flood, op);
>>> +
>>> +            /* If this port is connected to a multicast router then add it
>>> +             * to the MC_MROUTER_FLOOD group.
>>> +             */
>>> +            if (op->od->mcast_info.sw.flood_relay && op->peer &&
>>> +                    op->peer->od && op->peer->od->mcast_info.rtr.relay) {
>>> +                ovn_multicast_add(mcast_groups, &mc_mrouter_flood, op);
>>> +            }
>>>          }
>>>      }
>>>
>>> @@ -8628,10 +8836,61 @@ build_mcast_groups(struct northd_context *ctx,
>>>              continue;
>>>          }
>>>
>>> +        struct in6_addr group_address;
>>> +        if (!ovn_igmp_group_get_address(sb_igmp, &group_address)) {
>>> +            static struct vlog_rate_limit rl = VLOG_RATE_LIMIT_INIT(1, 1);
>>> +            VLOG_WARN_RL(&rl, "invalid IGMP group address: %s",
>>> +                         sb_igmp->address);
>>> +            continue;
>>> +        }
>>> +
>>>          /* Add the IGMP group entry. Will also try to allocate an ID for it
>>>           * if the multicast group already exists.
>>>           */
>>> -        ovn_igmp_group_add(ctx, igmp_groups, od, sb_igmp);
>>> +        struct ovn_igmp_group *igmp_group =
>>> +            ovn_igmp_group_add(ctx, igmp_groups, od, &group_address,
>>> +                               sb_igmp->address);
>>> +
>>> +        /* Extract the IGMP group ports from the SB entry and store them
>>> +         * in the IGMP group.
>>> +         */
>>> +        size_t n_igmp_ports;
>>> +        struct ovn_port **igmp_ports =
>>> +            ovn_igmp_group_get_ports(sb_igmp, &n_igmp_ports, ports);
>>> +        ovn_igmp_group_add_entry(igmp_group, igmp_ports, n_igmp_ports);
>>> +    }
>>> +
>>> +    /* Build IGMP groups for multicast routers with relay enabled. The router
>>> +     * IGMP groups are based on the groups learnt by their multicast enabled
>>> +     * peers.
>>> +     */
>>> +    struct ovn_datapath *od;
>>> +    HMAP_FOR_EACH (od, key_node, datapaths) {
>>> +
>>> +        if (ovs_list_is_empty(&od->mcast_info.groups)) {
>>> +            continue;
>>> +        }
>>> +
>>> +        for (size_t i = 0; i < od->n_router_ports; i++) {
>>> +            struct ovn_port *router_port = od->router_ports[i]->peer;
>>> +
>>> +            if (!router_port || !router_port->od ||
>>> +                    !router_port->od->mcast_info.rtr.relay) {
>>> +                continue;
>>> +            }
>>> +
>>> +            struct ovn_igmp_group *igmp_group;
>>> +            LIST_FOR_EACH (igmp_group, list_node, &od->mcast_info.groups) {
>>> +                struct ovn_igmp_group *igmp_group_rtr =
>>> +                    ovn_igmp_group_add(ctx, igmp_groups, router_port->od,
>>> +                                       &igmp_group->address,
>>> +                                       igmp_group->mcgroup.name);
>>> +                struct ovn_port **router_igmp_ports =
>>> +                    xmalloc(sizeof *router_igmp_ports);
>>> +                router_igmp_ports[0] = router_port;
>>> +                ovn_igmp_group_add_entry(igmp_group_rtr, router_igmp_ports, 1);
>>> +            }
>>> +        }
>>>      }
>>>
>>>      /* Walk the aggregated IGMP groups and allocate IDs for new entries.
>>> @@ -8639,21 +8898,17 @@ build_mcast_groups(struct northd_context *ctx,
>>>       */
>>>      struct ovn_igmp_group *igmp_group, *igmp_group_next;
>>>      HMAP_FOR_EACH_SAFE (igmp_group, igmp_group_next, hmap_node, igmp_groups) {
>>> -        if (igmp_group->mcgroup.key == 0) {
>>> -            struct mcast_info *mcast_info = &igmp_group->datapath->mcast_info;
>>> -            igmp_group->mcgroup.key = ovn_mcast_group_allocate_key(mcast_info);
>>> -        }
>>>
>>> -        /* If we ran out of keys just destroy the entry. */
>>> -        if (igmp_group->mcgroup.key == 0) {
>>> +        if (!ovn_igmp_group_allocate_id(igmp_group)) {
>>> +            /* If we ran out of keys just destroy the entry. */
>>>              ovn_igmp_group_destroy(igmp_groups, igmp_group);
>>>              continue;
>>>          }
>>>
>>> -        /* Aggregate the ports from all SB entries corresponding to this
>>> +        /* Aggregate the ports from all entries corresponding to this
>>>           * group.
>>>           */
>>> -        ovn_igmp_group_aggregate_ports(igmp_group, ports, mcast_groups);
>>> +        ovn_igmp_group_aggregate_ports(igmp_group, mcast_groups);
>>>      }
>>>  }
>>>
>>> diff --git a/ovn-nb.xml b/ovn-nb.xml
>>> index e166190..1f8d751 100644
>>> --- a/ovn-nb.xml
>>> +++ b/ovn-nb.xml
>>> @@ -1527,6 +1527,12 @@
>>>            address.
>>>          </p>
>>>        </column>
>>> +      <column name="options" key="mcast_relay" type'{"type": "boolean"}'>
>>> +        <p>
>>> +          Enables/disables IP multicast relay between logical switches
>>> +          connected to the logical router. Default: False.
>>> +        </p>
>>> +      </column>
>>>      </group>
>>>
>>>      <group title="Common Columns">
>>> diff --git a/ovn-sb.xml b/ovn-sb.xml
>>> index 17c45bb..02691bb 100644
>>> --- a/ovn-sb.xml
>>> +++ b/ovn-sb.xml
>>> @@ -1017,6 +1017,8 @@
>>>          <li><code>eth.mcast</code> expands to <code>eth.dst[40]</code></li>
>>>          <li><code>vlan.present</code> expands to <code>vlan.tci[12]</code></li>
>>>          <li><code>ip4</code> expands to <code>eth.type == 0x800</code></li>
>>> +        <li><code>ip4.src_mcast</code> expands to
>>> +        <code>ip4.src[28..31] == 0xe</code></li>
>>>          <li><code>ip4.mcast</code> expands to <code>ip4.dst[28..31] == 0xe</code></li>
>>>          <li><code>ip6</code> expands to <code>eth.type == 0x86dd</code></li>
>>>          <li><code>ip</code> expands to <code>ip4 || ip6</code></li>
>>> diff --git a/tests/ovn.at b/tests/ovn.at
>>> index 71eb390..52c044c 100644
>>> --- a/tests/ovn.at
>>> +++ b/tests/ovn.at
>>> @@ -14721,12 +14721,12 @@ AT_CHECK([ovn-sbctl get controller_event $uuid seq_num], [0], [dnl
>>>  OVN_CLEANUP([hv1], [hv2])
>>>  AT_CLEANUP
>>>
>>> -AT_SETUP([ovn -- IGMP snoop/querier])
>>> +AT_SETUP([ovn -- IGMP snoop/querier/relay])
>>>  AT_SKIP_IF([test $HAVE_PYTHON = no])
>>>  ovn_start
>>>
>>>  # Logical network:
>>> -# Two independent logical switches (sw1 and sw2).
>>> +# Three logical switches (sw1-sw3) connected to a logical router (rtr).
>>>  # sw1:
>>>  #   - subnet 10.0.0.0/8
>>>  #   - 2 ports bound on hv1 (sw1-p11, sw1-p12)
>>> @@ -14736,6 +14736,10 @@ ovn_start
>>>  #   - 1 port bound on hv1 (sw2-p1)
>>>  #   - 1 port bound on hv2 (sw2-p2)
>>>  #   - IGMP Querier from 20.0.0.254
>>> +# sw3:
>>> +#   - subnet 30.0.0.0/8
>>> +#   - 1 port bound on hv1 (sw3-p1)
>>> +#   - 1 port bound on hv2 (sw3-p2)
>>>
>>>  reset_pcap_file() {
>>>      local iface=$1
>>> @@ -14812,29 +14816,47 @@ store_igmp_v3_query() {
>>>  }
>>>
>>>  #
>>> -# send_ip_multicast_pkt INPORT HV ETH_SRC ETH_DST IP_SRC IP_DST IP_LEN
>>> -#    IP_PROTO DATA OUTFILE
>>> +# send_ip_multicast_pkt INPORT HV ETH_SRC ETH_DST IP_SRC IP_DST IP_LEN TTL
>>> +#    IP_CHKSUM IP_PROTO DATA
>>>  #
>>>  # This shell function causes an IP multicast packet to be received on INPORT
>>>  # of HV.
>>>  # The hexdump of the packet is stored in OUTFILE.
>>>  #
>>>  send_ip_multicast_pkt() {
>>> -    local inport=$1 hv=$2 eth_src=$3 eth_dst=$4 ip_src=$5 ip_dst=$6
>>> -    local ip_len=$7 ip_chksum=$8 proto=$9 data=${10} outfile=${11}
>>> -
>>> -    local ip_ttl=20
>>> +    local inport=$1 hv=$2 eth_src=$3 eth_dst=$4
>>> +    local ip_src=$5 ip_dst=$6 ip_len=$7 ip_ttl=$8 ip_chksum=$9 proto=${10}
>>> +    local data=${11}
>>>
>>>      local eth=${eth_dst}${eth_src}0800
>>>      local ip=450000${ip_len}95f14000${ip_ttl}${proto}${ip_chksum}${ip_src}${ip_dst}
>>>      local packet=${eth}${ip}${data}
>>>
>>>      as $hv ovs-appctl netdev-dummy/receive ${inport} ${packet}
>>> +}
>>> +
>>> +#
>>> +# store_ip_multicast_pkt ETH_SRC ETH_DST IP_SRC IP_DST IP_LEN TTL
>>> +#    IP_CHKSUM IP_PROTO DATA OUTFILE
>>> +#
>>> +# This shell function builds an IP multicast packet and stores the hexdump of
>>> +# the packet in OUTFILE.
>>> +#
>>> +store_ip_multicast_pkt() {
>>> +    local eth_src=$1 eth_dst=$2
>>> +    local ip_src=$3 ip_dst=$4 ip_len=$5 ip_ttl=$6 ip_chksum=$7 proto=$8
>>> +    local data=$9 outfile=${10}
>>> +
>>> +    local eth=${eth_dst}${eth_src}0800
>>> +    local ip=450000${ip_len}95f14000${ip_ttl}${proto}${ip_chksum}${ip_src}${ip_dst}
>>> +    local packet=${eth}${ip}${data}
>>> +
>>>      echo ${packet} >> ${outfile}
>>>  }
>>>
>>>  ovn-nbctl ls-add sw1
>>>  ovn-nbctl ls-add sw2
>>> +ovn-nbctl ls-add sw3
>>>
>>>  ovn-nbctl lsp-add sw1 sw1-p11
>>>  ovn-nbctl lsp-add sw1 sw1-p12
>>> @@ -14842,6 +14864,26 @@ ovn-nbctl lsp-add sw1 sw1-p21
>>>  ovn-nbctl lsp-add sw1 sw1-p22
>>>  ovn-nbctl lsp-add sw2 sw2-p1
>>>  ovn-nbctl lsp-add sw2 sw2-p2
>>> +ovn-nbctl lsp-add sw3 sw3-p1
>>> +ovn-nbctl lsp-add sw3 sw3-p2
>>> +
>>> +ovn-nbctl lr-add rtr
>>> +ovn-nbctl lrp-add rtr rtr-sw1 00:00:00:00:01:00 10.0.0.254/24
>>> +ovn-nbctl lrp-add rtr rtr-sw2 00:00:00:00:02:00 20.0.0.254/24
>>> +ovn-nbctl lrp-add rtr rtr-sw3 00:00:00:00:03:00 30.0.0.254/24
>>> +
>>> +ovn-nbctl lsp-add sw1 sw1-rtr                      \
>>> +    -- lsp-set-type sw1-rtr router                 \
>>> +    -- lsp-set-addresses sw1-rtr 00:00:00:00:01:00 \
>>> +    -- lsp-set-options sw1-rtr router-port=rtr-sw1
>>> +ovn-nbctl lsp-add sw2 sw2-rtr                      \
>>> +    -- lsp-set-type sw2-rtr router                 \
>>> +    -- lsp-set-addresses sw2-rtr 00:00:00:00:02:00 \
>>> +    -- lsp-set-options sw2-rtr router-port=rtr-sw2
>>> +ovn-nbctl lsp-add sw3 sw3-rtr                      \
>>> +    -- lsp-set-type sw3-rtr router                 \
>>> +    -- lsp-set-addresses sw3-rtr 00:00:00:00:03:00 \
>>> +    -- lsp-set-options sw3-rtr router-port=rtr-sw3
>>>
>>>  net_add n1
>>>  sim_add hv1
>>> @@ -14863,6 +14905,11 @@ ovs-vsctl -- add-port br-int hv1-vif3 -- \
>>>      options:tx_pcap=hv1/vif3-tx.pcap \
>>>      options:rxq_pcap=hv1/vif3-rx.pcap \
>>>      ofport-request=1
>>> +ovs-vsctl -- add-port br-int hv1-vif4 -- \
>>> +    set interface hv1-vif4 external-ids:iface-id=sw3-p1 \
>>> +    options:tx_pcap=hv1/vif4-tx.pcap \
>>> +    options:rxq_pcap=hv1/vif4-rx.pcap \
>>> +    ofport-request=1
>>>
>>>  sim_add hv2
>>>  as hv2
>>> @@ -14883,12 +14930,18 @@ ovs-vsctl -- add-port br-int hv2-vif3 -- \
>>>      options:tx_pcap=hv2/vif3-tx.pcap \
>>>      options:rxq_pcap=hv2/vif3-rx.pcap \
>>>      ofport-request=1
>>> +ovs-vsctl -- add-port br-int hv2-vif4 -- \
>>> +    set interface hv2-vif4 external-ids:iface-id=sw3-p2 \
>>> +    options:tx_pcap=hv2/vif4-tx.pcap \
>>> +    options:rxq_pcap=hv2/vif4-rx.pcap \
>>> +    ofport-request=1
>>>
>>>  OVN_POPULATE_ARP
>>>
>>>  # Enable IGMP snooping on sw1.
>>> -ovn-nbctl set Logical_Switch sw1 other_config:mcast_querier="false"
>>> -ovn-nbctl set Logical_Switch sw1 other_config:mcast_snoop="true"
>>> +ovn-nbctl set Logical_Switch sw1       \
>>> +    other_config:mcast_querier="false" \
>>> +    other_config:mcast_snoop="true"
>>>
>>>  # No IGMP query should be generated by sw1 (mcast_querier="false").
>>>  truncate -s 0 expected
>>> @@ -14921,9 +14974,12 @@ truncate -s 0 expected
>>>  truncate -s 0 expected_empty
>>>  send_ip_multicast_pkt hv1-vif2 hv1 \
>>>      000000000001 01005e000144 \
>>> -    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e ca70 11 \
>>> -    e518e518000a3b3a0000 \
>>> -    expected
>>> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
>>> +    e518e518000a3b3a0000
>>> +store_ip_multicast_pkt \
>>> +    000000000001 01005e000144 \
>>> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
>>> +    e518e518000a3b3a0000 expected
>>>
>>>  OVN_CHECK_PACKETS([hv1/vif1-tx.pcap], [expected])
>>>  OVN_CHECK_PACKETS([hv2/vif1-tx.pcap], [expected])
>>> @@ -14944,17 +15000,19 @@ OVS_WAIT_UNTIL([
>>>      test "${total_entries}" = "1"
>>>  ])
>>>
>>> -# Send traffic traffic and make sure it gets forwarded only on the port that
>>> -# joined.
>>> +# Send traffic and make sure it gets forwarded only on the port that joined.
>>>  as hv1 reset_pcap_file hv1-vif1 hv1/vif1
>>>  as hv2 reset_pcap_file hv2-vif1 hv2/vif1
>>>  truncate -s 0 expected
>>>  truncate -s 0 expected_empty
>>>  send_ip_multicast_pkt hv1-vif2 hv1 \
>>>      000000000001 01005e000144 \
>>> -    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e ca70 11 \
>>> -    e518e518000a3b3a0000 \
>>> -    expected
>>> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
>>> +    e518e518000a3b3a0000
>>> +store_ip_multicast_pkt \
>>> +    000000000001 01005e000144 \
>>> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
>>> +    e518e518000a3b3a0000 expected
>>>
>>>  OVN_CHECK_PACKETS([hv1/vif1-tx.pcap], [expected_empty])
>>>  OVN_CHECK_PACKETS([hv2/vif1-tx.pcap], [expected])
>>> @@ -14988,6 +15046,111 @@ sleep 1
>>>  OVN_CHECK_PACKETS([hv1/vif3-tx.pcap], [expected])
>>>  OVN_CHECK_PACKETS([hv2/vif3-tx.pcap], [expected])
>>>
>>> +# Dissable IGMP querier on sw2.
>>> +ovn-nbctl set Logical_Switch sw2 \
>>> +    other_config:mcast_querier="false"
>>> +
>>> +# Enable IGMP snooping on sw3.
>>> +ovn-nbctl set Logical_Switch sw3       \
>>> +    other_config:mcast_querier="false" \
>>> +    other_config:mcast_snoop="true"
>>> +
>>> +# Send traffic from sw3 and make sure rtr doesn't relay it.
>>> +truncate -s 0 expected_empty
>>> +
>>> +as hv1 reset_pcap_file hv1-vif1 hv1/vif1
>>> +as hv1 reset_pcap_file hv1-vif2 hv1/vif2
>>> +as hv1 reset_pcap_file hv1-vif3 hv1/vif3
>>> +as hv1 reset_pcap_file hv1-vif4 hv1/vif4
>>> +as hv2 reset_pcap_file hv2-vif1 hv2/vif1
>>> +as hv2 reset_pcap_file hv2-vif2 hv2/vif2
>>> +as hv2 reset_pcap_file hv2-vif3 hv2/vif3
>>> +as hv2 reset_pcap_file hv2-vif4 hv2/vif4
>>> +
>>> +send_ip_multicast_pkt hv2-vif4 hv2 \
>>> +    000000000001 01005e000144 \
>>> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
>>> +    e518e518000a3b3a0000
>>> +
>>> +# Sleep a bit to make sure no traffic is received and then check.
>>> +sleep 1
>>> +OVN_CHECK_PACKETS([hv1/vif1-tx.pcap], [expected_empty])
>>> +OVN_CHECK_PACKETS([hv2/vif3-tx.pcap], [expected_empty])
>>> +OVN_CHECK_PACKETS([hv1/vif4-tx.pcap], [expected_empty])
>>> +OVN_CHECK_PACKETS([hv1/vif2-tx.pcap], [expected_empty])
>>> +OVN_CHECK_PACKETS([hv1/vif3-tx.pcap], [expected_empty])
>>> +OVN_CHECK_PACKETS([hv2/vif1-tx.pcap], [expected_empty])
>>> +OVN_CHECK_PACKETS([hv2/vif2-tx.pcap], [expected_empty])
>>> +OVN_CHECK_PACKETS([hv2/vif4-tx.pcap], [expected_empty])
>>> +
>>> +# Enable IGMP relay on rtr
>>> +ovn-nbctl set logical_router rtr \
>>> +    options:mcast_relay="true"
>>> +
>>> +# Inject IGMP Join for 239.0.1.68 on sw1-p11.
>>> +send_igmp_v3_report hv1-vif1 hv1 \
>>> +    000000000001 $(ip_to_hex 10 0 0 1) f9f8 \
>>> +    $(ip_to_hex 239 0 1 68) 04 e9b9 \
>>> +    /dev/null
>>> +# Inject IGMP Join for 239.0.1.68 on sw2-p2.
>>> +send_igmp_v3_report hv2-vif3 hv2 \
>>> +    000000000001 $(ip_to_hex 10 0 0 1) f9f8 \
>>> +    $(ip_to_hex 239 0 1 68) 04 e9b9 \
>>> +    /dev/null
>>> +# Inject IGMP Join for 239.0.1.68 on sw3-p1.
>>> +send_igmp_v3_report hv1-vif4 hv1 \
>>> +    000000000001 $(ip_to_hex 10 0 0 1) f9f8 \
>>> +    $(ip_to_hex 239 0 1 68) 04 e9b9 \
>>> +    /dev/null
>>> +
>>> +# Check that the IGMP Group is learned by all switches.
>>> +OVS_WAIT_UNTIL([
>>> +    total_entries=`ovn-sbctl find IGMP_Group | grep "239.0.1.68" | wc -l`
>>> +    test "${total_entries}" = "3"
>>> +])
>>> +
>>> +# Send traffic from sw3 and make sure it is relayed by rtr.
>>> +# and ports that joined.
>>> +truncate -s 0 expected_routed_sw1
>>> +truncate -s 0 expected_routed_sw2
>>> +truncate -s 0 expected_switched
>>> +truncate -s 0 expected_empty
>>> +
>>> +as hv1 reset_pcap_file hv1-vif1 hv1/vif1
>>> +as hv1 reset_pcap_file hv1-vif2 hv1/vif2
>>> +as hv1 reset_pcap_file hv1-vif3 hv1/vif3
>>> +as hv1 reset_pcap_file hv1-vif4 hv1/vif4
>>> +as hv2 reset_pcap_file hv2-vif1 hv2/vif1
>>> +as hv2 reset_pcap_file hv2-vif2 hv2/vif2
>>> +as hv2 reset_pcap_file hv2-vif3 hv2/vif3
>>> +as hv2 reset_pcap_file hv2-vif4 hv2/vif4
>>> +
>>> +send_ip_multicast_pkt hv2-vif4 hv2 \
>>> +    000000000001 01005e000144 \
>>> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
>>> +    e518e518000a3b3a0000
>>> +store_ip_multicast_pkt \
>>> +    000000000100 01005e000144 \
>>> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 1f cb70 11 \
>>> +    e518e518000a3b3a0000 expected_routed_sw1
>>> +store_ip_multicast_pkt \
>>> +    000000000200 01005e000144 \
>>> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 1f cb70 11 \
>>> +    e518e518000a3b3a0000 expected_routed_sw2
>>> +store_ip_multicast_pkt \
>>> +    000000000001 01005e000144 \
>>> +    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
>>> +    e518e518000a3b3a0000 expected_switched
>>> +
>>> +OVN_CHECK_PACKETS([hv1/vif1-tx.pcap], [expected_routed_sw1])
>>> +OVN_CHECK_PACKETS([hv2/vif3-tx.pcap], [expected_routed_sw2])
>>> +OVN_CHECK_PACKETS([hv1/vif4-tx.pcap], [expected_switched])
>>> +OVN_CHECK_PACKETS([hv1/vif2-tx.pcap], [expected_empty])
>>> +OVN_CHECK_PACKETS([hv1/vif3-tx.pcap], [expected_empty])
>>> +OVN_CHECK_PACKETS([hv2/vif1-tx.pcap], [expected_empty])
>>> +OVN_CHECK_PACKETS([hv2/vif2-tx.pcap], [expected_empty])
>>> +OVN_CHECK_PACKETS([hv2/vif4-tx.pcap], [expected_empty])
>>> +
>>>  OVN_CLEANUP([hv1], [hv2])
>>>  AT_CLEANUP
>>>
>>> --
>>> 1.8.3.1
>>>
>>> _______________________________________________
>>> dev mailing list
>>> dev@openvswitch.org
>>> https://mail.openvswitch.org/mailman/listinfo/ovs-dev
0-day Robot Aug. 19, 2019, 2:58 p.m. UTC | #4
Bleep bloop.  Greetings Dumitru Ceara, I am a robot and I have tried out your patch.
Thanks for your contribution.

I encountered some error that I wasn't expecting.  See the details below.


git-am:
Failed to merge in the changes.
Patch failed at 0001 ovn-northd: Add IGMP Relay support
The copy of the patch that failed is found in:
   /var/lib/jenkins/jobs/upstream_build_from_pw/workspace/OVN/.git/rebase-apply/patch
When you have resolved this problem, run "git am --resolved".
If you prefer to skip this patch, run "git am --skip" instead.
To restore the original branch and stop patching, run "git am --abort".


Please check this out.  If you feel there has been an error, please email aconole@redhat.com

Thanks,
0-day Robot

Patch
diff mbox series

diff --git a/NEWS b/NEWS
index f476984..73045d6 100644
--- a/NEWS
+++ b/NEWS
@@ -39,6 +39,7 @@  Post-v2.11.0
        logical groups which results in tunnels only been formed between
        members of the same transport zone(s).
      * Support for new logical switch port type - 'virtual'.
+     * Support for IGMP Snooping/Querier and Relay.
    - New QoS type "linux-netem" on Linux.
    - Added support for TLS Server Name Indication (SNI).
 
diff --git a/lib/logical-fields.c b/lib/logical-fields.c
index 4ad5bf4..8fb591c 100644
--- a/lib/logical-fields.c
+++ b/lib/logical-fields.c
@@ -156,6 +156,8 @@  ovn_init_symtab(struct shash *symtab)
 
     expr_symtab_add_field(symtab, "ip4.src", MFF_IPV4_SRC, "ip4", false);
     expr_symtab_add_field(symtab, "ip4.dst", MFF_IPV4_DST, "ip4", false);
+    expr_symtab_add_predicate(symtab, "ip4.src_mcast",
+                              "ip4.src[28..31] == 0xe");
     expr_symtab_add_predicate(symtab, "ip4.mcast", "ip4.dst[28..31] == 0xe");
 
     expr_symtab_add_predicate(symtab, "icmp4", "ip4 && ip.proto == 1");
diff --git a/lib/mcast-group-index.h b/lib/mcast-group-index.h
index 15a1592..cb49ad7 100644
--- a/lib/mcast-group-index.h
+++ b/lib/mcast-group-index.h
@@ -20,8 +20,17 @@  struct ovsdb_idl;
 
 struct sbrec_datapath_binding;
 
-#define OVN_MCAST_FLOOD_TUNNEL_KEY   65535
-#define OVN_MCAST_UNKNOWN_TUNNEL_KEY (OVN_MCAST_FLOOD_TUNNEL_KEY - 1)
+#define OVN_MIN_MULTICAST 32768
+#define OVN_MAX_MULTICAST 65535
+
+enum ovn_mcast_tunnel_keys {
+
+    OVN_MCAST_FLOOD_TUNNEL_KEY = OVN_MIN_MULTICAST,
+    OVN_MCAST_UNKNOWN_TUNNEL_KEY,
+    OVN_MCAST_MROUTER_FLOOD_TUNNEL_KEY,
+    OVN_MIN_IP_MULTICAST,
+    OVN_MAX_IP_MULTICAST = OVN_MAX_MULTICAST,
+};
 
 struct ovsdb_idl_index *mcast_group_index_create(struct ovsdb_idl *);
 const struct sbrec_multicast_group *
diff --git a/northd/ovn-northd.8.xml b/northd/ovn-northd.8.xml
index 6d2fbe3..d45bb15 100644
--- a/northd/ovn-northd.8.xml
+++ b/northd/ovn-northd.8.xml
@@ -947,10 +947,40 @@  output;
 
     <ul>
       <li>
-        A priority-100 flow that outputs all packets with an Ethernet broadcast
+        A priority-100 flow that punts all IGMP packets to
+        <code>ovn-controller</code> if IGMP snooping is enabled on the
+        logical switch.
+      </li>
+
+      <li>
+        Priority-90 flows that forward registered IP multicast traffic to
+        their corresponding multicast group, which <code>ovn-northd</code>
+        creates based on learnt <ref table="IGMP_Group" db="OVN_Southbound"/>
+        entries.  The flows also forward packets to the
+        <code>MC_MROUTER_FLOOD</code> multicast group, which
+        <code>ovn-nortdh</code> populates with all the logical ports that
+        are connected to logical routers with
+        <ref column="options" table="Logical_Router"/>:mcast_relay='true'.
+      </li>
+
+      <li>
+        A priority-85 flow that forwards all IP multicast traffic destined to
+        224.0.0.X to the <code>MC_FLOOD</code> multicast group, which
+        <code>ovn-northd</code> populates with all enabled logical ports.
+      </li>
+
+      <li>
+        A priority-80 flow that forwards all unregistered IP multicast traffic
+        to the <code>MC_MROUTER_FLOOD</code> multicast group, if any.
+        Otherwise the flow drops all unregistered IP multicast packets.  This
+        flow is added only if <ref column="other_config"
+        table="Logical_Switch"/>:mcast_flood_unregistered='false'.
+      </li>
+
+      <li>
+        A priority-70 flow that outputs all packets with an Ethernet broadcast
         or multicast <code>eth.dst</code> to the <code>MC_FLOOD</code>
-        multicast group, which <code>ovn-northd</code> populates with all
-        enabled logical ports.
+        multicast group.
       </li>
 
       <li>
@@ -1228,6 +1258,14 @@  output;
 
       <li>
         <p>
+          A priority-95 flow allows IP multicast traffic if
+          <ref column="options" table="Logical_Router"/>:mcast_relay='true',
+          otherwise drops it.
+        </p>
+      </li>
+
+      <li>
+        <p>
           ICMP echo reply.  These flows reply to ICMP echo requests received
           for the router's IP address.  Let <var>A</var> be an IP address
           owned by a router port.  Then, for each <var>A</var> that is
@@ -1941,6 +1979,16 @@  output;
     <ul>
       <li>
         <p>
+          Priority-500 flows that match IP multicast traffic destined to
+          groups registered on any of the attached switches and sets
+          <code>outport</code> to the associated multicast group that will
+          eventually flood the traffic to all interested attached logical
+          switches. The flows also decrement TTL.
+        </p>
+      </li>
+
+      <li>
+        <p>
           For distributed logical routers where one of the logical router
           ports specifies a <code>redirect-chassis</code>, a priority-400
           logical flow for each ip source/destination couple that matches the
@@ -2074,6 +2122,15 @@  next;
     <ul>
       <li>
         <p>
+          A priority-500 flow that matches IP multicast traffic that was
+          allowed in the routing pipeline. For this kind of traffic the
+          <code>outport</code> was already set so the flow just advances to
+          the next table.
+        </p>
+      </li>
+
+      <li>
+        <p>
           For distributed logical routers where one of the logical router
           ports specifies a <code>redirect-chassis</code>, a priority-400
           logical flow with match <code>REGBIT_DISTRIBUTED_NAT == 1</code>
@@ -2641,9 +2698,19 @@  clone {
     <h3>Egress Table 3: Delivery</h3>
 
     <p>
-      Packets that reach this table are ready for delivery.  It contains
-      priority-100 logical flows that match packets on each enabled logical
-      router port, with action <code>output;</code>.
+      Packets that reach this table are ready for delivery.  It contains:
+      <ul>
+        <li>
+          Priority-110 logical flows that match IP multicast packets on each
+          enabled logical router port and modify the Ethernet source address
+          of the packets to the Ethernet address of the port and then execute
+          action <code>output;</code>.
+        </li>
+        <li>
+          Priority-100 logical flows that match packets on each enabled
+          logical router port, with action <code>output;</code>.
+        </li>
+      </ul>
     </p>
 
 </manpage>
diff --git a/northd/ovn-northd.c b/northd/ovn-northd.c
index e861344..592d3b9 100644
--- a/northd/ovn-northd.c
+++ b/northd/ovn-northd.c
@@ -433,32 +433,52 @@  struct ipam_info {
     bool mac_only;
 };
 
-#define OVN_MIN_MULTICAST 32768
-#define OVN_MAX_MULTICAST OVN_MCAST_FLOOD_TUNNEL_KEY
-BUILD_ASSERT_DECL(OVN_MIN_MULTICAST < OVN_MAX_MULTICAST);
-
-#define OVN_MIN_IP_MULTICAST OVN_MIN_MULTICAST
-#define OVN_MAX_IP_MULTICAST (OVN_MCAST_UNKNOWN_TUNNEL_KEY - 1)
-BUILD_ASSERT_DECL(OVN_MAX_IP_MULTICAST >= OVN_MIN_MULTICAST);
-
 /*
  * Multicast snooping and querier per datapath configuration.
  */
+struct mcast_switch_info {
+
+    bool enabled;               /* True if snooping enabled. */
+    bool querier;               /* True if querier enabled. */
+    bool flood_unregistered;    /* True if unregistered multicast should be
+                                 * flooded.
+                                 */
+    bool flood_relay;           /* True if the switch is connected to a
+                                 * multicast router and unregistered multicast
+                                 * should be flooded to the mrouter. Only
+                                 * applicable if flood_unregistered == false.
+                                 */
+
+    int64_t table_size;         /* Max number of IP multicast groups. */
+    int64_t idle_timeout;       /* Timeout after which an idle group is
+                                 * flushed.
+                                 */
+    int64_t query_interval;     /* Interval between multicast queries. */
+    char *eth_src;              /* ETH src address of the multicast queries. */
+    char *ipv4_src;             /* IP src address of the multicast queries. */
+    int64_t query_max_response; /* Expected time after which reports should
+                                 * be received for queries that were sent out.
+                                 */
+
+    uint32_t active_flows;      /* Current number of active IP multicast
+                                 * flows.
+                                 */
+};
+
+struct mcast_router_info {
+    bool relay; /* True if the router should relay IP multicast. */
+};
+
 struct mcast_info {
-    bool enabled;
-    bool querier;
-    bool flood_unregistered;
-
-    int64_t table_size;
-    int64_t idle_timeout;
-    int64_t query_interval;
-    char *eth_src;
-    char *ipv4_src;
-    int64_t  query_max_response;
-
-    struct hmap group_tnlids;
-    uint32_t group_tnlid_hint;
-    uint32_t active_flows;
+
+    struct hmap group_tnlids;  /* Group tunnel IDs in use on this DP. */
+    uint32_t group_tnlid_hint; /* Hint for allocating next group tunnel ID. */
+    struct ovs_list groups;    /* List of groups learnt on this DP. */
+
+    union {
+        struct mcast_switch_info sw;  /* Switch specific multicast info. */
+        struct mcast_router_info rtr; /* Router specific multicast info. */
+    };
 };
 
 static uint32_t
@@ -559,6 +579,7 @@  ovn_datapath_create(struct hmap *datapaths, const struct uuid *key,
 }
 
 static void ovn_ls_port_group_destroy(struct hmap *nb_pgs);
+static void destroy_mcast_info_for_datapath(struct ovn_datapath *od);
 
 static void
 ovn_datapath_destroy(struct hmap *datapaths, struct ovn_datapath *od)
@@ -572,12 +593,7 @@  ovn_datapath_destroy(struct hmap *datapaths, struct ovn_datapath *od)
         bitmap_free(od->ipam_info.allocated_ipv4s);
         free(od->router_ports);
         ovn_ls_port_group_destroy(&od->nb_pgs);
-
-        if (od->nbs) {
-            free(od->mcast_info.eth_src);
-            free(od->mcast_info.ipv4_src);
-            destroy_tnlids(&od->mcast_info.group_tnlids);
-        }
+        destroy_mcast_info_for_datapath(od);
 
         free(od);
     }
@@ -714,23 +730,28 @@  init_ipam_info_for_datapath(struct ovn_datapath *od)
 }
 
 static void
-init_mcast_info_for_datapath(struct ovn_datapath *od)
+init_mcast_info_for_router_datapath(struct ovn_datapath *od)
 {
-    if (!od->nbs) {
-        return;
-    }
+    struct mcast_router_info *mcast_rtr_info = &od->mcast_info.rtr;
 
-    struct mcast_info *mcast_info = &od->mcast_info;
+    mcast_rtr_info->relay = smap_get_bool(&od->nbr->options, "mcast_relay",
+                                          false);
+}
 
-    mcast_info->enabled =
+static void
+init_mcast_info_for_switch_datapath(struct ovn_datapath *od)
+{
+    struct mcast_switch_info *mcast_sw_info = &od->mcast_info.sw;
+
+    mcast_sw_info->enabled =
         smap_get_bool(&od->nbs->other_config, "mcast_snoop", false);
-    mcast_info->querier =
+    mcast_sw_info->querier =
         smap_get_bool(&od->nbs->other_config, "mcast_querier", true);
-    mcast_info->flood_unregistered =
+    mcast_sw_info->flood_unregistered =
         smap_get_bool(&od->nbs->other_config, "mcast_flood_unregistered",
                       false);
 
-    mcast_info->table_size =
+    mcast_sw_info->table_size =
         smap_get_ullong(&od->nbs->other_config, "mcast_table_size",
                         OVN_MCAST_DEFAULT_MAX_ENTRIES);
 
@@ -742,54 +763,94 @@  init_mcast_info_for_datapath(struct ovn_datapath *od)
     } else if (idle_timeout > OVN_MCAST_MAX_IDLE_TIMEOUT_S) {
         idle_timeout = OVN_MCAST_MAX_IDLE_TIMEOUT_S;
     }
-    mcast_info->idle_timeout = idle_timeout;
+    mcast_sw_info->idle_timeout = idle_timeout;
 
     uint32_t query_interval =
         smap_get_ullong(&od->nbs->other_config, "mcast_query_interval",
-                        mcast_info->idle_timeout / 2);
+                        mcast_sw_info->idle_timeout / 2);
     if (query_interval < OVN_MCAST_MIN_QUERY_INTERVAL_S) {
         query_interval = OVN_MCAST_MIN_QUERY_INTERVAL_S;
     } else if (query_interval > OVN_MCAST_MAX_QUERY_INTERVAL_S) {
         query_interval = OVN_MCAST_MAX_QUERY_INTERVAL_S;
     }
-    mcast_info->query_interval = query_interval;
+    mcast_sw_info->query_interval = query_interval;
 
-    mcast_info->eth_src =
+    mcast_sw_info->eth_src =
         nullable_xstrdup(smap_get(&od->nbs->other_config, "mcast_eth_src"));
-    mcast_info->ipv4_src =
+    mcast_sw_info->ipv4_src =
         nullable_xstrdup(smap_get(&od->nbs->other_config, "mcast_ip4_src"));
 
-    mcast_info->query_max_response =
+    mcast_sw_info->query_max_response =
         smap_get_ullong(&od->nbs->other_config, "mcast_query_max_response",
                         OVN_MCAST_DEFAULT_QUERY_MAX_RESPONSE_S);
 
-    hmap_init(&mcast_info->group_tnlids);
-    mcast_info->group_tnlid_hint = OVN_MIN_IP_MULTICAST;
-    mcast_info->active_flows = 0;
+    mcast_sw_info->active_flows = 0;
+}
+
+static void
+init_mcast_info_for_datapath(struct ovn_datapath *od)
+{
+    if (!od->nbr && !od->nbs) {
+        return;
+    }
+
+    hmap_init(&od->mcast_info.group_tnlids);
+    od->mcast_info.group_tnlid_hint = OVN_MIN_IP_MULTICAST;
+    ovs_list_init(&od->mcast_info.groups);
+
+    if (od->nbs) {
+        init_mcast_info_for_switch_datapath(od);
+    } else {
+        init_mcast_info_for_router_datapath(od);
+    }
+}
+
+static void
+destroy_mcast_info_for_switch_datapath(struct ovn_datapath *od)
+{
+    struct mcast_switch_info *mcast_sw_info = &od->mcast_info.sw;
+
+    free(mcast_sw_info->eth_src);
+    free(mcast_sw_info->ipv4_src);
+}
+
+static void
+destroy_mcast_info_for_datapath(struct ovn_datapath *od)
+{
+    if (!od->nbr && !od->nbs) {
+        return;
+    }
+
+    if (od->nbs) {
+        destroy_mcast_info_for_switch_datapath(od);
+    }
+
+    destroy_tnlids(&od->mcast_info.group_tnlids);
 }
 
 static void
-store_mcast_info_for_datapath(const struct sbrec_ip_multicast *sb,
-                              struct ovn_datapath *od)
+store_mcast_info_for_switch_datapath(const struct sbrec_ip_multicast *sb,
+                                     struct ovn_datapath *od)
 {
-    struct mcast_info *mcast_info = &od->mcast_info;
+    struct mcast_switch_info *mcast_sw_info = &od->mcast_info.sw;
 
     sbrec_ip_multicast_set_datapath(sb, od->sb);
-    sbrec_ip_multicast_set_enabled(sb, &mcast_info->enabled, 1);
-    sbrec_ip_multicast_set_querier(sb, &mcast_info->querier, 1);
-    sbrec_ip_multicast_set_table_size(sb, &mcast_info->table_size, 1);
-    sbrec_ip_multicast_set_idle_timeout(sb, &mcast_info->idle_timeout, 1);
+    sbrec_ip_multicast_set_enabled(sb, &mcast_sw_info->enabled, 1);
+    sbrec_ip_multicast_set_querier(sb, &mcast_sw_info->querier, 1);
+    sbrec_ip_multicast_set_table_size(sb, &mcast_sw_info->table_size, 1);
+    sbrec_ip_multicast_set_idle_timeout(sb, &mcast_sw_info->idle_timeout, 1);
     sbrec_ip_multicast_set_query_interval(sb,
-                                          &mcast_info->query_interval, 1);
+                                          &mcast_sw_info->query_interval, 1);
     sbrec_ip_multicast_set_query_max_resp(sb,
-                                          &mcast_info->query_max_response, 1);
+                                          &mcast_sw_info->query_max_response,
+                                          1);
 
-    if (mcast_info->eth_src) {
-        sbrec_ip_multicast_set_eth_src(sb, mcast_info->eth_src);
+    if (mcast_sw_info->eth_src) {
+        sbrec_ip_multicast_set_eth_src(sb, mcast_sw_info->eth_src);
     }
 
-    if (mcast_info->ipv4_src) {
-        sbrec_ip_multicast_set_ip4_src(sb, mcast_info->ipv4_src);
+    if (mcast_sw_info->ipv4_src) {
+        sbrec_ip_multicast_set_ip4_src(sb, mcast_sw_info->ipv4_src);
     }
 }
 
@@ -906,6 +967,7 @@  join_datapaths(struct northd_context *ctx, struct hmap *datapaths,
                                      NULL, nbr, NULL);
             ovs_list_push_back(nb_only, &od->list);
         }
+        init_mcast_info_for_datapath(od);
         ovs_list_push_back(lr_list, &od->lr_list);
     }
 }
@@ -1999,6 +2061,13 @@  join_logical_ports(struct northd_context *ctx,
                     break;
                 }
             }
+
+            /* If the router is multicast enabled then set relay on the switch
+             * datapath.
+             */
+            if (peer->od && peer->od->mcast_info.rtr.relay) {
+                op->od->mcast_info.sw.flood_relay = true;
+            }
         } else if (op->nbrp && op->nbrp->peer && !op->derived) {
             struct ovn_port *peer = ovn_port_find(ports, op->nbrp->peer);
             if (peer) {
@@ -2846,6 +2915,10 @@  struct multicast_group {
 static const struct multicast_group mc_flood =
     { MC_FLOOD, OVN_MCAST_FLOOD_TUNNEL_KEY };
 
+#define MC_MROUTER_FLOOD "_MC_mrouter_flood"
+static const struct multicast_group mc_mrouter_flood =
+    { MC_MROUTER_FLOOD, OVN_MCAST_MROUTER_FLOOD_TUNNEL_KEY };
+
 #define MC_UNKNOWN "_MC_unknown"
 static const struct multicast_group mc_unknown =
     { MC_UNKNOWN, OVN_MCAST_UNKNOWN_TUNNEL_KEY };
@@ -2955,7 +3028,8 @@  ovn_multicast_update_sbrec(const struct ovn_multicast *mc,
  */
 struct ovn_igmp_group_entry {
     struct ovs_list list_node; /* Linkage in the list of entries. */
-    const struct sbrec_igmp_group *sb;
+    size_t n_ports;
+    struct ovn_port **ports;
 };
 
 /*
@@ -2964,12 +3038,13 @@  struct ovn_igmp_group_entry {
  */
 struct ovn_igmp_group {
     struct hmap_node hmap_node; /* Index on 'datapath' and 'address'. */
+    struct ovs_list list_node;  /* Linkage in the per-dp igmp group list. */
 
     struct ovn_datapath *datapath;
     struct in6_addr address; /* Multicast IPv6-mapped-IPv4 or IPv4 address. */
     struct multicast_group mcgroup;
 
-    struct ovs_list sb_entries; /* List of SB entries for this group. */
+    struct ovs_list entries; /* List of SB entries for this group. */
 };
 
 static uint32_t
@@ -2997,77 +3072,120 @@  ovn_igmp_group_find(struct hmap *igmp_groups,
     return NULL;
 }
 
-static void
+static struct ovn_igmp_group *
 ovn_igmp_group_add(struct northd_context *ctx, struct hmap *igmp_groups,
                    struct ovn_datapath *datapath,
-                   const struct sbrec_igmp_group *sb_igmp_group)
+                   const struct in6_addr *address,
+                   const char *address_s)
 {
-    struct in6_addr group_address;
-    ovs_be32 ipv4;
-
-    if (ip_parse(sb_igmp_group->address, &ipv4)) {
-        group_address = in6_addr_mapped_ipv4(ipv4);
-    } else if (!ipv6_parse(sb_igmp_group->address, &group_address)) {
-        static struct vlog_rate_limit rl = VLOG_RATE_LIMIT_INIT(1, 1);
-        VLOG_WARN_RL(&rl, "invalid IGMP group address: %s",
-                     sb_igmp_group->address);
-        return;
-    }
-
     struct ovn_igmp_group *igmp_group =
-        ovn_igmp_group_find(igmp_groups, datapath, &group_address);
+        ovn_igmp_group_find(igmp_groups, datapath, address);
 
     if (!igmp_group) {
         igmp_group = xmalloc(sizeof *igmp_group);
 
         const struct sbrec_multicast_group *mcgroup =
-            mcast_group_lookup(ctx->sbrec_mcast_group_by_name_dp,
-                               sb_igmp_group->address, datapath->sb);
+            mcast_group_lookup(ctx->sbrec_mcast_group_by_name_dp, address_s,
+                               datapath->sb);
 
         igmp_group->datapath = datapath;
-        igmp_group->address = group_address;
+        igmp_group->address = *address;
         if (mcgroup) {
             igmp_group->mcgroup.key = mcgroup->tunnel_key;
             add_tnlid(&datapath->mcast_info.group_tnlids, mcgroup->tunnel_key);
         } else {
             igmp_group->mcgroup.key = 0;
         }
-        igmp_group->mcgroup.name = sb_igmp_group->address;
-        ovs_list_init(&igmp_group->sb_entries);
+        igmp_group->mcgroup.name = address_s;
+        ovs_list_init(&igmp_group->entries);
 
         hmap_insert(igmp_groups, &igmp_group->hmap_node,
-                    ovn_igmp_group_hash(datapath, &group_address));
+                    ovn_igmp_group_hash(datapath, address));
+        ovs_list_push_back(&datapath->mcast_info.groups,
+                           &igmp_group->list_node);
+    }
+
+    return igmp_group;
+}
+
+static bool
+ovn_igmp_group_get_address(const struct sbrec_igmp_group *sb_igmp_group,
+                           struct in6_addr *address)
+{
+    ovs_be32 ipv4;
+
+    if (ip_parse(sb_igmp_group->address, &ipv4)) {
+        *address = in6_addr_mapped_ipv4(ipv4);
+        return true;
+    }
+    if (!ipv6_parse(sb_igmp_group->address, address)) {
+        return false;
     }
+    return true;
+}
 
+static struct ovn_port **
+ovn_igmp_group_get_ports(const struct sbrec_igmp_group *sb_igmp_group,
+                         size_t *n_ports, struct hmap *ovn_ports)
+{
+    struct ovn_port **ports = xmalloc(sb_igmp_group->n_ports * sizeof *ports);
+
+     *n_ports = 0;
+     for (size_t i = 0; i < sb_igmp_group->n_ports; i++) {
+        ports[(*n_ports)] =
+            ovn_port_find(ovn_ports, sb_igmp_group->ports[i]->logical_port);
+        if (ports[(*n_ports)]) {
+            (*n_ports)++;
+        }
+    }
+
+    return ports;
+}
+
+static void
+ovn_igmp_group_add_entry(struct ovn_igmp_group *igmp_group,
+                         struct ovn_port **ports, size_t n_ports)
+{
     struct ovn_igmp_group_entry *entry = xmalloc(sizeof *entry);
 
-    entry->sb = sb_igmp_group;
-    ovs_list_push_back(&igmp_group->sb_entries , &entry->list_node);
+    entry->ports = ports;
+    entry->n_ports = n_ports;
+    ovs_list_push_back(&igmp_group->entries, &entry->list_node);
+}
+
+static void
+ovn_igmp_group_destroy_entry(struct ovn_igmp_group_entry *entry)
+{
+    free(entry->ports);
+}
+
+static bool
+ovn_igmp_group_allocate_id(struct ovn_igmp_group *igmp_group)
+{
+    if (igmp_group->mcgroup.key == 0) {
+        struct mcast_info *mcast_info = &igmp_group->datapath->mcast_info;
+        igmp_group->mcgroup.key = ovn_mcast_group_allocate_key(mcast_info);
+    }
+
+    if (igmp_group->mcgroup.key == 0) {
+        return false;
+    }
+
+    return true;
 }
 
 static void
 ovn_igmp_group_aggregate_ports(struct ovn_igmp_group *igmp_group,
-                               struct hmap *ovn_ports,
                                struct hmap *mcast_groups)
 {
     struct ovn_igmp_group_entry *entry;
 
-    LIST_FOR_EACH_POP (entry, list_node, &igmp_group->sb_entries) {
-        size_t n_oports = 0;
-        struct ovn_port **oports =
-            xmalloc(entry->sb->n_ports * sizeof *oports);
-
-        for (size_t i = 0; i < entry->sb->n_ports; i++) {
-            oports[n_oports] =
-                ovn_port_find(ovn_ports, entry->sb->ports[i]->logical_port);
-            if (oports[n_oports]) {
-                n_oports++;
-            }
-        }
-
+    LIST_FOR_EACH_POP (entry, list_node, &igmp_group->entries) {
         ovn_multicast_add_ports(mcast_groups, igmp_group->datapath,
-                                &igmp_group->mcgroup, oports, n_oports);
-        free(oports);
+                                &igmp_group->mcgroup, entry->ports,
+                                entry->n_ports);
+
+        ovn_igmp_group_destroy_entry(entry);
         free(entry);
     }
 }
@@ -3079,10 +3197,12 @@  ovn_igmp_group_destroy(struct hmap *igmp_groups,
     if (igmp_group) {
         struct ovn_igmp_group_entry *entry;
 
-        LIST_FOR_EACH_POP (entry, list_node, &igmp_group->sb_entries) {
+        LIST_FOR_EACH_POP (entry, list_node, &igmp_group->entries) {
+            ovn_igmp_group_destroy_entry(entry);
             free(entry);
         }
         hmap_remove(igmp_groups, &igmp_group->hmap_node);
+        ovs_list_remove(&igmp_group->list_node);
         free(igmp_group);
     }
 }
@@ -5286,7 +5406,9 @@  build_lswitch_flows(struct hmap *datapaths, struct hmap *ports,
             continue;
         }
 
-        if (od->mcast_info.enabled) {
+        struct mcast_switch_info *mcast_sw_info = &od->mcast_info.sw;
+
+        if (mcast_sw_info->enabled) {
             /* Punt IGMP traffic to controller. */
             ovn_lflow_add(lflows, od, S_SWITCH_IN_L2_LKUP, 100,
                           "ip4 && ip.proto == 2", "igmp;");
@@ -5299,9 +5421,16 @@  build_lswitch_flows(struct hmap *datapaths, struct hmap *ports,
                           "outport = \""MC_FLOOD"\"; output;");
 
             /* Drop unregistered IP multicast if not allowed. */
-            if (!od->mcast_info.flood_unregistered) {
-                ovn_lflow_add(lflows, od, S_SWITCH_IN_L2_LKUP, 80,
-                              "ip4 && ip4.mcast", "drop;");
+            if (!mcast_sw_info->flood_unregistered) {
+                /* Forward unregistered IP multicast to mrouter (if any). */
+                if (mcast_sw_info->flood_relay) {
+                    ovn_lflow_add(lflows, od, S_SWITCH_IN_L2_LKUP, 80,
+                                  "ip4 && ip4.mcast",
+                                  "outport = \""MC_MROUTER_FLOOD"\"; output;");
+                } else {
+                    ovn_lflow_add(lflows, od, S_SWITCH_IN_L2_LKUP, 80,
+                                  "ip4 && ip4.mcast", "drop;");
+                }
             }
         }
 
@@ -5318,18 +5447,26 @@  build_lswitch_flows(struct hmap *datapaths, struct hmap *ports,
             continue;
         }
 
-        struct mcast_info *mcast_info = &igmp_group->datapath->mcast_info;
+        struct mcast_switch_info *mcast_sw_info =
+            &igmp_group->datapath->mcast_info.sw;
 
-        if (mcast_info->active_flows >= mcast_info->table_size) {
+        if (mcast_sw_info->active_flows >= mcast_sw_info->table_size) {
             continue;
         }
-        mcast_info->active_flows++;
+        mcast_sw_info->active_flows++;
 
         ds_clear(&match);
         ds_clear(&actions);
 
         ds_put_format(&match, "eth.mcast && ip4 && ip4.dst == %s ",
                       igmp_group->mcgroup.name);
+        /* Also flood traffic to all multicast routers with relay enabled. */
+        if (mcast_sw_info->flood_relay) {
+            ds_put_cstr(&actions,
+                        "clone { "
+                            "outport = \""MC_MROUTER_FLOOD "\"; output; "
+                        "};");
+        }
         ds_put_format(&actions, "outport = \"%s\"; output; ",
                       igmp_group->mcgroup.name);
 
@@ -6209,7 +6346,7 @@  build_lrouter_flows(struct hmap *datapaths, struct hmap *ports,
          * source or destination, and zero network source or destination
          * (priority 100). */
         ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_INPUT, 100,
-                      "ip4.mcast || "
+                      "ip4.src_mcast ||"
                       "ip4.src == 255.255.255.255 || "
                       "ip4.src == 127.0.0.0/8 || "
                       "ip4.dst == 127.0.0.0/8 || "
@@ -6217,6 +6354,16 @@  build_lrouter_flows(struct hmap *datapaths, struct hmap *ports,
                       "ip4.dst == 0.0.0.0/8",
                       "drop;");
 
+        /* Allow multicast if relay enabled (priority 95). */
+        ds_clear(&actions);
+        if (od->mcast_info.rtr.relay) {
+            ds_put_cstr(&actions, "next;");
+        } else {
+            ds_put_cstr(&actions, "drop;");
+        }
+        ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_INPUT, 95,
+                      "ip4.mcast", ds_cstr(&actions));
+
         /* ARP reply handling.  Use ARP replies to populate the logical
          * router's ARP table. */
         ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_INPUT, 90, "arp.op == 2",
@@ -7487,6 +7634,27 @@  build_lrouter_flows(struct hmap *datapaths, struct hmap *ports,
         }
     }
 
+    /* IP Multicast lookup. Here we set the output port, adjust TTL and
+     * advance to next table (priority 500).
+     */
+    HMAP_FOR_EACH (od, key_node, datapaths) {
+        if (!od->nbr || !od->mcast_info.rtr.relay) {
+            continue;
+        }
+        struct ovn_igmp_group *igmp_group;
+
+        LIST_FOR_EACH (igmp_group, list_node, &od->mcast_info.groups) {
+            ds_clear(&match);
+            ds_clear(&actions);
+            ds_put_format(&match, "ip4 && ip4.dst == %s ",
+                          igmp_group->mcgroup.name);
+            ds_put_format(&actions, "outport = \"%s\"; ip.ttl--; next;",
+                          igmp_group->mcgroup.name);
+            ovn_lflow_add(lflows, od, S_ROUTER_IN_IP_ROUTING, 500,
+                          ds_cstr(&match), ds_cstr(&actions));
+        }
+    }
+
     /* Logical router ingress table 8: Policy.
      *
      * A packet that arrives at this table is an IP packet that should be
@@ -7517,10 +7685,24 @@  build_lrouter_flows(struct hmap *datapaths, struct hmap *ports,
 
     /* Local router ingress table 9: ARP Resolution.
      *
-     * Any packet that reaches this table is an IP packet whose next-hop IP
-     * address is in reg0. (ip4.dst is the final destination.) This table
-     * resolves the IP address in reg0 into an output port in outport and an
-     * Ethernet address in eth.dst. */
+     * Multicast packets already have the outport set so just advance to next
+     * table (priority 500). */
+    HMAP_FOR_EACH (od, key_node, datapaths) {
+        if (!od->nbr) {
+            continue;
+        }
+
+        ovn_lflow_add(lflows, od, S_ROUTER_IN_ARP_RESOLVE, 500,
+                      "ip4.mcast", "next;");
+    }
+
+    /* Local router ingress table 9: ARP Resolution.
+     *
+     * Any unicast packet that reaches this table is an IP packet whose
+     * next-hop IP address is in reg0. (ip4.dst is the final destination.)
+     * This table resolves the IP address in reg0 into an output port in
+     * outport and an Ethernet address in eth.dst.
+     */
     HMAP_FOR_EACH (op, key_node, ports) {
         if (op->nbsp && !lsp_is_enabled(op->nbsp)) {
             continue;
@@ -8002,9 +8184,13 @@  build_lrouter_flows(struct hmap *datapaths, struct hmap *ports,
         ovn_lflow_add(lflows, od, S_ROUTER_IN_ARP_REQUEST, 0, "1", "output;");
     }
 
-    /* Logical router egress table 1: Delivery (priority 100).
+    /* Logical router egress table 1: Delivery (priority 100-110).
      *
-     * Priority 100 rules deliver packets to enabled logical ports. */
+     * Priority 100 rules deliver packets to enabled logical ports.
+     * Priority 110 rules match multicast packets and update the source
+     * mac before delivering to enabled logical ports. IP multicast traffic
+     * bypasses S_ROUTER_IN_IP_ROUTING route lookups.
+     */
     HMAP_FOR_EACH (op, key_node, ports) {
         if (!op->nbrp) {
             continue;
@@ -8024,6 +8210,20 @@  build_lrouter_flows(struct hmap *datapaths, struct hmap *ports,
             continue;
         }
 
+        /* If multicast relay is enabled then also adjust source mac for IP
+         * multicast traffic.
+         */
+        if (op->od->mcast_info.rtr.relay) {
+            ds_clear(&match);
+            ds_clear(&actions);
+            ds_put_format(&match, "ip4.mcast && outport == %s",
+                          op->json_key);
+            ds_put_format(&actions, "eth.src = %s; output;",
+                          op->lrp_networks.ea_s);
+            ovn_lflow_add(lflows, op->od, S_ROUTER_OUT_DELIVERY, 110,
+                        ds_cstr(&match), ds_cstr(&actions));
+        }
+
         ds_clear(&match);
         ds_put_format(&match, "outport == %s", op->json_key);
         ovn_lflow_add(lflows, op->od, S_ROUTER_OUT_DELIVERY, 100,
@@ -8574,7 +8774,7 @@  build_ip_mcast(struct northd_context *ctx, struct hmap *datapaths)
         if (!ip_mcast) {
             ip_mcast = sbrec_ip_multicast_insert(ctx->ovnsb_txn);
         }
-        store_mcast_info_for_datapath(ip_mcast, od);
+        store_mcast_info_for_switch_datapath(ip_mcast, od);
     }
 
     /* Delete southbound records without northbound matches. */
@@ -8606,6 +8806,14 @@  build_mcast_groups(struct northd_context *ctx,
 
         if (lsp_is_enabled(op->nbsp)) {
             ovn_multicast_add(mcast_groups, &mc_flood, op);
+
+            /* If this port is connected to a multicast router then add it
+             * to the MC_MROUTER_FLOOD group.
+             */
+            if (op->od->mcast_info.sw.flood_relay && op->peer &&
+                    op->peer->od && op->peer->od->mcast_info.rtr.relay) {
+                ovn_multicast_add(mcast_groups, &mc_mrouter_flood, op);
+            }
         }
     }
 
@@ -8628,10 +8836,61 @@  build_mcast_groups(struct northd_context *ctx,
             continue;
         }
 
+        struct in6_addr group_address;
+        if (!ovn_igmp_group_get_address(sb_igmp, &group_address)) {
+            static struct vlog_rate_limit rl = VLOG_RATE_LIMIT_INIT(1, 1);
+            VLOG_WARN_RL(&rl, "invalid IGMP group address: %s",
+                         sb_igmp->address);
+            continue;
+        }
+
         /* Add the IGMP group entry. Will also try to allocate an ID for it
          * if the multicast group already exists.
          */
-        ovn_igmp_group_add(ctx, igmp_groups, od, sb_igmp);
+        struct ovn_igmp_group *igmp_group =
+            ovn_igmp_group_add(ctx, igmp_groups, od, &group_address,
+                               sb_igmp->address);
+
+        /* Extract the IGMP group ports from the SB entry and store them
+         * in the IGMP group.
+         */
+        size_t n_igmp_ports;
+        struct ovn_port **igmp_ports =
+            ovn_igmp_group_get_ports(sb_igmp, &n_igmp_ports, ports);
+        ovn_igmp_group_add_entry(igmp_group, igmp_ports, n_igmp_ports);
+    }
+
+    /* Build IGMP groups for multicast routers with relay enabled. The router
+     * IGMP groups are based on the groups learnt by their multicast enabled
+     * peers.
+     */
+    struct ovn_datapath *od;
+    HMAP_FOR_EACH (od, key_node, datapaths) {
+
+        if (ovs_list_is_empty(&od->mcast_info.groups)) {
+            continue;
+        }
+
+        for (size_t i = 0; i < od->n_router_ports; i++) {
+            struct ovn_port *router_port = od->router_ports[i]->peer;
+
+            if (!router_port || !router_port->od ||
+                    !router_port->od->mcast_info.rtr.relay) {
+                continue;
+            }
+
+            struct ovn_igmp_group *igmp_group;
+            LIST_FOR_EACH (igmp_group, list_node, &od->mcast_info.groups) {
+                struct ovn_igmp_group *igmp_group_rtr =
+                    ovn_igmp_group_add(ctx, igmp_groups, router_port->od,
+                                       &igmp_group->address,
+                                       igmp_group->mcgroup.name);
+                struct ovn_port **router_igmp_ports =
+                    xmalloc(sizeof *router_igmp_ports);
+                router_igmp_ports[0] = router_port;
+                ovn_igmp_group_add_entry(igmp_group_rtr, router_igmp_ports, 1);
+            }
+        }
     }
 
     /* Walk the aggregated IGMP groups and allocate IDs for new entries.
@@ -8639,21 +8898,17 @@  build_mcast_groups(struct northd_context *ctx,
      */
     struct ovn_igmp_group *igmp_group, *igmp_group_next;
     HMAP_FOR_EACH_SAFE (igmp_group, igmp_group_next, hmap_node, igmp_groups) {
-        if (igmp_group->mcgroup.key == 0) {
-            struct mcast_info *mcast_info = &igmp_group->datapath->mcast_info;
-            igmp_group->mcgroup.key = ovn_mcast_group_allocate_key(mcast_info);
-        }
 
-        /* If we ran out of keys just destroy the entry. */
-        if (igmp_group->mcgroup.key == 0) {
+        if (!ovn_igmp_group_allocate_id(igmp_group)) {
+            /* If we ran out of keys just destroy the entry. */
             ovn_igmp_group_destroy(igmp_groups, igmp_group);
             continue;
         }
 
-        /* Aggregate the ports from all SB entries corresponding to this
+        /* Aggregate the ports from all entries corresponding to this
          * group.
          */
-        ovn_igmp_group_aggregate_ports(igmp_group, ports, mcast_groups);
+        ovn_igmp_group_aggregate_ports(igmp_group, mcast_groups);
     }
 }
 
diff --git a/ovn-nb.xml b/ovn-nb.xml
index e166190..1f8d751 100644
--- a/ovn-nb.xml
+++ b/ovn-nb.xml
@@ -1527,6 +1527,12 @@ 
           address.
         </p>
       </column>
+      <column name="options" key="mcast_relay" type'{"type": "boolean"}'>
+        <p>
+          Enables/disables IP multicast relay between logical switches
+          connected to the logical router. Default: False.
+        </p>
+      </column>
     </group>
 
     <group title="Common Columns">
diff --git a/ovn-sb.xml b/ovn-sb.xml
index 17c45bb..02691bb 100644
--- a/ovn-sb.xml
+++ b/ovn-sb.xml
@@ -1017,6 +1017,8 @@ 
         <li><code>eth.mcast</code> expands to <code>eth.dst[40]</code></li>
         <li><code>vlan.present</code> expands to <code>vlan.tci[12]</code></li>
         <li><code>ip4</code> expands to <code>eth.type == 0x800</code></li>
+        <li><code>ip4.src_mcast</code> expands to
+        <code>ip4.src[28..31] == 0xe</code></li>
         <li><code>ip4.mcast</code> expands to <code>ip4.dst[28..31] == 0xe</code></li>
         <li><code>ip6</code> expands to <code>eth.type == 0x86dd</code></li>
         <li><code>ip</code> expands to <code>ip4 || ip6</code></li>
diff --git a/tests/ovn.at b/tests/ovn.at
index 71eb390..52c044c 100644
--- a/tests/ovn.at
+++ b/tests/ovn.at
@@ -14721,12 +14721,12 @@  AT_CHECK([ovn-sbctl get controller_event $uuid seq_num], [0], [dnl
 OVN_CLEANUP([hv1], [hv2])
 AT_CLEANUP
 
-AT_SETUP([ovn -- IGMP snoop/querier])
+AT_SETUP([ovn -- IGMP snoop/querier/relay])
 AT_SKIP_IF([test $HAVE_PYTHON = no])
 ovn_start
 
 # Logical network:
-# Two independent logical switches (sw1 and sw2).
+# Three logical switches (sw1-sw3) connected to a logical router (rtr).
 # sw1:
 #   - subnet 10.0.0.0/8
 #   - 2 ports bound on hv1 (sw1-p11, sw1-p12)
@@ -14736,6 +14736,10 @@  ovn_start
 #   - 1 port bound on hv1 (sw2-p1)
 #   - 1 port bound on hv2 (sw2-p2)
 #   - IGMP Querier from 20.0.0.254
+# sw3:
+#   - subnet 30.0.0.0/8
+#   - 1 port bound on hv1 (sw3-p1)
+#   - 1 port bound on hv2 (sw3-p2)
 
 reset_pcap_file() {
     local iface=$1
@@ -14812,29 +14816,47 @@  store_igmp_v3_query() {
 }
 
 #
-# send_ip_multicast_pkt INPORT HV ETH_SRC ETH_DST IP_SRC IP_DST IP_LEN
-#    IP_PROTO DATA OUTFILE
+# send_ip_multicast_pkt INPORT HV ETH_SRC ETH_DST IP_SRC IP_DST IP_LEN TTL
+#    IP_CHKSUM IP_PROTO DATA
 #
 # This shell function causes an IP multicast packet to be received on INPORT
 # of HV.
 # The hexdump of the packet is stored in OUTFILE.
 #
 send_ip_multicast_pkt() {
-    local inport=$1 hv=$2 eth_src=$3 eth_dst=$4 ip_src=$5 ip_dst=$6
-    local ip_len=$7 ip_chksum=$8 proto=$9 data=${10} outfile=${11}
-
-    local ip_ttl=20
+    local inport=$1 hv=$2 eth_src=$3 eth_dst=$4
+    local ip_src=$5 ip_dst=$6 ip_len=$7 ip_ttl=$8 ip_chksum=$9 proto=${10}
+    local data=${11}
 
     local eth=${eth_dst}${eth_src}0800
     local ip=450000${ip_len}95f14000${ip_ttl}${proto}${ip_chksum}${ip_src}${ip_dst}
     local packet=${eth}${ip}${data}
 
     as $hv ovs-appctl netdev-dummy/receive ${inport} ${packet}
+}
+
+#
+# store_ip_multicast_pkt ETH_SRC ETH_DST IP_SRC IP_DST IP_LEN TTL
+#    IP_CHKSUM IP_PROTO DATA OUTFILE
+#
+# This shell function builds an IP multicast packet and stores the hexdump of
+# the packet in OUTFILE.
+#
+store_ip_multicast_pkt() {
+    local eth_src=$1 eth_dst=$2
+    local ip_src=$3 ip_dst=$4 ip_len=$5 ip_ttl=$6 ip_chksum=$7 proto=$8
+    local data=$9 outfile=${10}
+
+    local eth=${eth_dst}${eth_src}0800
+    local ip=450000${ip_len}95f14000${ip_ttl}${proto}${ip_chksum}${ip_src}${ip_dst}
+    local packet=${eth}${ip}${data}
+
     echo ${packet} >> ${outfile}
 }
 
 ovn-nbctl ls-add sw1
 ovn-nbctl ls-add sw2
+ovn-nbctl ls-add sw3
 
 ovn-nbctl lsp-add sw1 sw1-p11
 ovn-nbctl lsp-add sw1 sw1-p12
@@ -14842,6 +14864,26 @@  ovn-nbctl lsp-add sw1 sw1-p21
 ovn-nbctl lsp-add sw1 sw1-p22
 ovn-nbctl lsp-add sw2 sw2-p1
 ovn-nbctl lsp-add sw2 sw2-p2
+ovn-nbctl lsp-add sw3 sw3-p1
+ovn-nbctl lsp-add sw3 sw3-p2
+
+ovn-nbctl lr-add rtr
+ovn-nbctl lrp-add rtr rtr-sw1 00:00:00:00:01:00 10.0.0.254/24
+ovn-nbctl lrp-add rtr rtr-sw2 00:00:00:00:02:00 20.0.0.254/24
+ovn-nbctl lrp-add rtr rtr-sw3 00:00:00:00:03:00 30.0.0.254/24
+
+ovn-nbctl lsp-add sw1 sw1-rtr                      \
+    -- lsp-set-type sw1-rtr router                 \
+    -- lsp-set-addresses sw1-rtr 00:00:00:00:01:00 \
+    -- lsp-set-options sw1-rtr router-port=rtr-sw1
+ovn-nbctl lsp-add sw2 sw2-rtr                      \
+    -- lsp-set-type sw2-rtr router                 \
+    -- lsp-set-addresses sw2-rtr 00:00:00:00:02:00 \
+    -- lsp-set-options sw2-rtr router-port=rtr-sw2
+ovn-nbctl lsp-add sw3 sw3-rtr                      \
+    -- lsp-set-type sw3-rtr router                 \
+    -- lsp-set-addresses sw3-rtr 00:00:00:00:03:00 \
+    -- lsp-set-options sw3-rtr router-port=rtr-sw3
 
 net_add n1
 sim_add hv1
@@ -14863,6 +14905,11 @@  ovs-vsctl -- add-port br-int hv1-vif3 -- \
     options:tx_pcap=hv1/vif3-tx.pcap \
     options:rxq_pcap=hv1/vif3-rx.pcap \
     ofport-request=1
+ovs-vsctl -- add-port br-int hv1-vif4 -- \
+    set interface hv1-vif4 external-ids:iface-id=sw3-p1 \
+    options:tx_pcap=hv1/vif4-tx.pcap \
+    options:rxq_pcap=hv1/vif4-rx.pcap \
+    ofport-request=1
 
 sim_add hv2
 as hv2
@@ -14883,12 +14930,18 @@  ovs-vsctl -- add-port br-int hv2-vif3 -- \
     options:tx_pcap=hv2/vif3-tx.pcap \
     options:rxq_pcap=hv2/vif3-rx.pcap \
     ofport-request=1
+ovs-vsctl -- add-port br-int hv2-vif4 -- \
+    set interface hv2-vif4 external-ids:iface-id=sw3-p2 \
+    options:tx_pcap=hv2/vif4-tx.pcap \
+    options:rxq_pcap=hv2/vif4-rx.pcap \
+    ofport-request=1
 
 OVN_POPULATE_ARP
 
 # Enable IGMP snooping on sw1.
-ovn-nbctl set Logical_Switch sw1 other_config:mcast_querier="false"
-ovn-nbctl set Logical_Switch sw1 other_config:mcast_snoop="true"
+ovn-nbctl set Logical_Switch sw1       \
+    other_config:mcast_querier="false" \
+    other_config:mcast_snoop="true"
 
 # No IGMP query should be generated by sw1 (mcast_querier="false").
 truncate -s 0 expected
@@ -14921,9 +14974,12 @@  truncate -s 0 expected
 truncate -s 0 expected_empty
 send_ip_multicast_pkt hv1-vif2 hv1 \
     000000000001 01005e000144 \
-    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e ca70 11 \
-    e518e518000a3b3a0000 \
-    expected
+    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
+    e518e518000a3b3a0000
+store_ip_multicast_pkt \
+    000000000001 01005e000144 \
+    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
+    e518e518000a3b3a0000 expected
 
 OVN_CHECK_PACKETS([hv1/vif1-tx.pcap], [expected])
 OVN_CHECK_PACKETS([hv2/vif1-tx.pcap], [expected])
@@ -14944,17 +15000,19 @@  OVS_WAIT_UNTIL([
     test "${total_entries}" = "1"
 ])
 
-# Send traffic traffic and make sure it gets forwarded only on the port that
-# joined.
+# Send traffic and make sure it gets forwarded only on the port that joined.
 as hv1 reset_pcap_file hv1-vif1 hv1/vif1
 as hv2 reset_pcap_file hv2-vif1 hv2/vif1
 truncate -s 0 expected
 truncate -s 0 expected_empty
 send_ip_multicast_pkt hv1-vif2 hv1 \
     000000000001 01005e000144 \
-    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e ca70 11 \
-    e518e518000a3b3a0000 \
-    expected
+    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
+    e518e518000a3b3a0000
+store_ip_multicast_pkt \
+    000000000001 01005e000144 \
+    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
+    e518e518000a3b3a0000 expected
 
 OVN_CHECK_PACKETS([hv1/vif1-tx.pcap], [expected_empty])
 OVN_CHECK_PACKETS([hv2/vif1-tx.pcap], [expected])
@@ -14988,6 +15046,111 @@  sleep 1
 OVN_CHECK_PACKETS([hv1/vif3-tx.pcap], [expected])
 OVN_CHECK_PACKETS([hv2/vif3-tx.pcap], [expected])
 
+# Dissable IGMP querier on sw2.
+ovn-nbctl set Logical_Switch sw2 \
+    other_config:mcast_querier="false"
+
+# Enable IGMP snooping on sw3.
+ovn-nbctl set Logical_Switch sw3       \
+    other_config:mcast_querier="false" \
+    other_config:mcast_snoop="true"
+
+# Send traffic from sw3 and make sure rtr doesn't relay it.
+truncate -s 0 expected_empty
+
+as hv1 reset_pcap_file hv1-vif1 hv1/vif1
+as hv1 reset_pcap_file hv1-vif2 hv1/vif2
+as hv1 reset_pcap_file hv1-vif3 hv1/vif3
+as hv1 reset_pcap_file hv1-vif4 hv1/vif4
+as hv2 reset_pcap_file hv2-vif1 hv2/vif1
+as hv2 reset_pcap_file hv2-vif2 hv2/vif2
+as hv2 reset_pcap_file hv2-vif3 hv2/vif3
+as hv2 reset_pcap_file hv2-vif4 hv2/vif4
+
+send_ip_multicast_pkt hv2-vif4 hv2 \
+    000000000001 01005e000144 \
+    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
+    e518e518000a3b3a0000
+
+# Sleep a bit to make sure no traffic is received and then check.
+sleep 1
+OVN_CHECK_PACKETS([hv1/vif1-tx.pcap], [expected_empty])
+OVN_CHECK_PACKETS([hv2/vif3-tx.pcap], [expected_empty])
+OVN_CHECK_PACKETS([hv1/vif4-tx.pcap], [expected_empty])
+OVN_CHECK_PACKETS([hv1/vif2-tx.pcap], [expected_empty])
+OVN_CHECK_PACKETS([hv1/vif3-tx.pcap], [expected_empty])
+OVN_CHECK_PACKETS([hv2/vif1-tx.pcap], [expected_empty])
+OVN_CHECK_PACKETS([hv2/vif2-tx.pcap], [expected_empty])
+OVN_CHECK_PACKETS([hv2/vif4-tx.pcap], [expected_empty])
+
+# Enable IGMP relay on rtr
+ovn-nbctl set logical_router rtr \
+    options:mcast_relay="true"
+
+# Inject IGMP Join for 239.0.1.68 on sw1-p11.
+send_igmp_v3_report hv1-vif1 hv1 \
+    000000000001 $(ip_to_hex 10 0 0 1) f9f8 \
+    $(ip_to_hex 239 0 1 68) 04 e9b9 \
+    /dev/null
+# Inject IGMP Join for 239.0.1.68 on sw2-p2.
+send_igmp_v3_report hv2-vif3 hv2 \
+    000000000001 $(ip_to_hex 10 0 0 1) f9f8 \
+    $(ip_to_hex 239 0 1 68) 04 e9b9 \
+    /dev/null
+# Inject IGMP Join for 239.0.1.68 on sw3-p1.
+send_igmp_v3_report hv1-vif4 hv1 \
+    000000000001 $(ip_to_hex 10 0 0 1) f9f8 \
+    $(ip_to_hex 239 0 1 68) 04 e9b9 \
+    /dev/null
+
+# Check that the IGMP Group is learned by all switches.
+OVS_WAIT_UNTIL([
+    total_entries=`ovn-sbctl find IGMP_Group | grep "239.0.1.68" | wc -l`
+    test "${total_entries}" = "3"
+])
+
+# Send traffic from sw3 and make sure it is relayed by rtr.
+# and ports that joined.
+truncate -s 0 expected_routed_sw1
+truncate -s 0 expected_routed_sw2
+truncate -s 0 expected_switched
+truncate -s 0 expected_empty
+
+as hv1 reset_pcap_file hv1-vif1 hv1/vif1
+as hv1 reset_pcap_file hv1-vif2 hv1/vif2
+as hv1 reset_pcap_file hv1-vif3 hv1/vif3
+as hv1 reset_pcap_file hv1-vif4 hv1/vif4
+as hv2 reset_pcap_file hv2-vif1 hv2/vif1
+as hv2 reset_pcap_file hv2-vif2 hv2/vif2
+as hv2 reset_pcap_file hv2-vif3 hv2/vif3
+as hv2 reset_pcap_file hv2-vif4 hv2/vif4
+
+send_ip_multicast_pkt hv2-vif4 hv2 \
+    000000000001 01005e000144 \
+    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
+    e518e518000a3b3a0000
+store_ip_multicast_pkt \
+    000000000100 01005e000144 \
+    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 1f cb70 11 \
+    e518e518000a3b3a0000 expected_routed_sw1
+store_ip_multicast_pkt \
+    000000000200 01005e000144 \
+    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 1f cb70 11 \
+    e518e518000a3b3a0000 expected_routed_sw2
+store_ip_multicast_pkt \
+    000000000001 01005e000144 \
+    $(ip_to_hex 10 0 0 42) $(ip_to_hex 239 0 1 68) 1e 20 ca70 11 \
+    e518e518000a3b3a0000 expected_switched
+
+OVN_CHECK_PACKETS([hv1/vif1-tx.pcap], [expected_routed_sw1])
+OVN_CHECK_PACKETS([hv2/vif3-tx.pcap], [expected_routed_sw2])
+OVN_CHECK_PACKETS([hv1/vif4-tx.pcap], [expected_switched])
+OVN_CHECK_PACKETS([hv1/vif2-tx.pcap], [expected_empty])
+OVN_CHECK_PACKETS([hv1/vif3-tx.pcap], [expected_empty])
+OVN_CHECK_PACKETS([hv2/vif1-tx.pcap], [expected_empty])
+OVN_CHECK_PACKETS([hv2/vif2-tx.pcap], [expected_empty])
+OVN_CHECK_PACKETS([hv2/vif4-tx.pcap], [expected_empty])
+
 OVN_CLEANUP([hv1], [hv2])
 AT_CLEANUP