[B.A.T.M.A.N.] [PATCHv5 0/3] batman-adv: Unicasting multicast reports to querier-node only

[B.A.T.M.A.N.] [PATCHv5 0/3] batman-adv: Unicasting multicast reports to querier-node only

[Linus Lüssing]

The last round of multicast patches send to the batman-adv mailinglist
to add support for the multicast optimizations in bridged scenarios, too,
unfortunately had one major conceptual flaw: It could lead to packet loss.
It's not sufficient to have the unicasting of reports implemented on
bridge-nodes only. Nodes without bridges need to treat reports the same
way.

The issue is described in detail here:

https://www.open-mesh.org/projects/batman-adv/wiki/Multicast-optimizations-listener-reports


Cheers, Linus

-----

Changelog:
v5:
 * Removed RFC tag: Needed exports got merged to net-next and are going to
   be available with Linux 4.2
 * Redid compat solution - now fully backwards compatible down to 2.6.33
v4:
 * excluded bridge part from this patchset, they should
   hopefully be added to net-next soon
 * Added a compat solution (PATCH 3/3)
 * Removed Kconfig-depends as by David's suggestion the needed parsing
   functions for MLD are going to be forced built-ins even if IPv6 is
   going to be built as a module
 * Removed unused variable 'int ret' in batadv_mcast_is_report_ipv6()
 * Adjusted to new folder structure
v3:
 * Adding Kconfig-depends and #if's
   (so basically adding similar dependancy constraints as the bridge code
    has, except that there are no depends if batman-adv gets compiled without
    multicast optimizations)
   -> the case of IPv6=M and batman-adv=y is still impossible if multicast
      optimizations are enabled; but I don't see the practical demand for that
      either - people who use IPv6 as a module will probably also want to
      use batman-adv as a module
v2:
 * various bugfixes (now runtime tested, too - should(tm) work)
 * added netdev+bridge mailinglists

[B.A.T.M.A.N.] [PATCHv5 1/3] batman-adv: Forward IGMP/MLD reports to selected querier (only)

[Linus Lüssing]

With this patch IGMP or MLD reports are only forwarded to the selected
IGMP/MLD querier as RFC4541 suggests. This is necessary to avoid
multicast packet loss in bridged scenarios later:

An IGMPv2/MLDv1 querier does not actively join the multicast group the
reports are sent to. Because of this, this leads to snooping
bridges/switches not being able to learn of multicast listeners in the
mesh and wrongly shutting down ports for multicast traffic to the mesh.

Signed-off-by: Linus Lüssing <linus.luessing@c0d3.blue>
---
 net/batman-adv/main.h           |    4 +-
 net/batman-adv/multicast.c      |  257 ++++++++++++++++++++++++++++++++++++++-
 net/batman-adv/multicast.h      |    8 ++
 net/batman-adv/soft-interface.c |   11 ++
 net/batman-adv/types.h          |   14 +++
 5 files changed, 287 insertions(+), 7 deletions(-)

diff --git a/net/batman-adv/main.h b/net/batman-adv/main.h
index 41d27c7..2832cf9 100644
--- a/net/batman-adv/main.h
+++ b/net/batman-adv/main.h
@@ -219,6 +219,7 @@ __be32 batadv_skb_crc32(struct sk_buff *skb, u8 *payload_ptr);
  * @BATADV_DBG_BLA: bridge loop avoidance messages
  * @BATADV_DBG_DAT: ARP snooping and DAT related messages
  * @BATADV_DBG_NC: network coding related messages
+ * @BATADV_DBG_MCAST: multicast related messages
  * @BATADV_DBG_ALL: the union of all the above log levels
  */
 enum batadv_dbg_level {
@@ -228,7 +229,8 @@ enum batadv_dbg_level {
 	BATADV_DBG_BLA    = BIT(3),
 	BATADV_DBG_DAT    = BIT(4),
 	BATADV_DBG_NC	  = BIT(5),
-	BATADV_DBG_ALL    = 63,
+	BATADV_DBG_MCAST  = BIT(6),
+	BATADV_DBG_ALL    = 127,
 };
 
 #ifdef CONFIG_BATMAN_ADV_DEBUG
diff --git a/net/batman-adv/multicast.c b/net/batman-adv/multicast.c
index 7aa480b..72d561c 100644
--- a/net/batman-adv/multicast.c
+++ b/net/batman-adv/multicast.c
@@ -24,6 +24,7 @@
 #include <linux/etherdevice.h>
 #include <linux/fs.h>
 #include <linux/if_ether.h>
+#include <linux/igmp.h>
 #include <linux/in6.h>
 #include <linux/in.h>
 #include <linux/ip.h>
@@ -40,6 +41,7 @@
 #include <linux/types.h>
 #include <net/addrconf.h>
 #include <net/ipv6.h>
+#include <net/mld.h>
 
 #include "packet.h"
 #include "translation-table.h"
@@ -270,10 +272,76 @@ out:
 }
 
 /**
+ * batadv_mcast_is_report_ipv4 - check for IGMP reports (and queries)
+ * @bat_priv: the bat priv with all the soft interface information
+ * @skb: the ethernet frame destined for the mesh
+ * @orig: if IGMP report: to be set to the querier to forward the skb to
+ *
+ * Checks whether the given frame is an IGMP report and if so sets the
+ * orig pointer to the originator of the selected IGMP querier if one exists
+ * and returns true. Otherwise returns false.
+ *
+ * If the packet is a general IGMP query instead then we delete the memorized,
+ * foreign IGMP querier (if one exists): We became the selected querier and
+ * therefore do not need to forward reports into the mesh.
+ *
+ * This call might reallocate skb data.
+ */
+static bool batadv_mcast_is_report_ipv4(struct batadv_priv *bat_priv,
+					struct sk_buff *skb,
+					struct batadv_orig_node **orig)
+{
+	struct batadv_mcast_querier_state *querier;
+	struct batadv_orig_node *orig_node;
+
+	if (ip_mc_check_igmp(skb, NULL) < 0)
+		return false;
+
+	querier = &bat_priv->mcast.querier_ipv4;
+
+	switch (igmp_hdr(skb)->type) {
+	case IGMP_HOST_MEMBERSHIP_REPORT:
+	case IGMPV2_HOST_MEMBERSHIP_REPORT:
+	case IGMPV3_HOST_MEMBERSHIP_REPORT:
+		rcu_read_lock();
+		orig_node = rcu_dereference(querier->orig);
+		if (orig_node && atomic_inc_not_zero(&orig_node->refcount)) {
+			/* TODO: include multicast routers via MRD (RFC4286) */
+			batadv_dbg(BATADV_DBG_MCAST, bat_priv,
+				   "Redirecting IGMP Report to %pM\n",
+				   orig_node->orig);
+			*orig = orig_node;
+		}
+		rcu_read_unlock();
+
+		return true;
+	case IGMP_HOST_MEMBERSHIP_QUERY:
+		/* RFC4541, section 2.1.1.1.b) says:
+		 * ignore general queries from 0.0.0.0
+		 */
+		if (!ip_hdr(skb)->saddr || igmp_hdr(skb)->group)
+			break;
+
+		spin_lock_bh(&querier->orig_lock);
+		orig_node = rcu_dereference(querier->orig);
+		if (orig_node)
+			rcu_assign_pointer(querier->orig, NULL);
+		spin_unlock_bh(&querier->orig_lock);
+
+		batadv_dbg(BATADV_DBG_MCAST, bat_priv,
+			   "Snooped own IGMP Query\n");
+		break;
+	}
+
+	return false;
+}
+
+/**
  * batadv_mcast_forw_mode_check_ipv4 - check for optimized forwarding potential
  * @bat_priv: the bat priv with all the soft interface information
  * @skb: the IPv4 packet to check
  * @is_unsnoopable: stores whether the destination is snoopable
+ * @orig: for IGMP reports: to be set to the querier to forward the skb to
  *
  * Checks whether the given IPv4 packet has the potential to be forwarded with a
  * mode more optimal than classic flooding.
@@ -283,7 +351,8 @@ out:
  */
 static int batadv_mcast_forw_mode_check_ipv4(struct batadv_priv *bat_priv,
 					     struct sk_buff *skb,
-					     bool *is_unsnoopable)
+					     bool *is_unsnoopable,
+					     struct batadv_orig_node **orig)
 {
 	struct iphdr *iphdr;
 
@@ -291,6 +360,9 @@ static int batadv_mcast_forw_mode_check_ipv4(struct batadv_priv *bat_priv,
 	if (!pskb_may_pull(skb, sizeof(struct ethhdr) + sizeof(*iphdr)))
 		return -ENOMEM;
 
+	if (batadv_mcast_is_report_ipv4(bat_priv, skb, orig))
+		return 0;
+
 	iphdr = ip_hdr(skb);
 
 	/* TODO: Implement Multicast Router Discovery (RFC4286),
@@ -307,11 +379,73 @@ static int batadv_mcast_forw_mode_check_ipv4(struct batadv_priv *bat_priv,
 	return 0;
 }
 
+#if IS_ENABLED(CONFIG_IPV6)
+/**
+ * batadv_mcast_is_report_ipv6 - check for MLD reports (and queries)
+ * @bat_priv: the bat priv with all the soft interface information
+ * @skb: the ethernet frame destined for the mesh
+ * @orig: if MLD report: to be set to the querier to forward the skb to
+ *
+ * Checks whether the given frame is an MLD report and if so sets the
+ * orig pointer to the originator of the selected MLD querier if one exists
+ * and returns true. Otherwise returns false.
+ *
+ * If the packet is a general MLD query instead then we delete the memorized,
+ * foreign MLD querier (if one exists): We became the selected querier and
+ * therefore do not need to forward reports into the mesh.
+ *
+ * This call might reallocate skb data.
+ */
+static bool batadv_mcast_is_report_ipv6(struct batadv_priv *bat_priv,
+					struct sk_buff *skb,
+					struct batadv_orig_node **orig)
+{
+	struct mld_msg *mld;
+	struct batadv_mcast_querier_state *querier;
+	struct batadv_orig_node *orig_node;
+
+	if (ipv6_mc_check_mld(skb, NULL) < 0)
+		return false;
+
+	querier = &bat_priv->mcast.querier_ipv6;
+	mld = (struct mld_msg *)icmp6_hdr(skb);
+
+	switch (mld->mld_type) {
+	case ICMPV6_MGM_REPORT:
+	case ICMPV6_MLD2_REPORT:
+		rcu_read_lock();
+		orig_node = rcu_dereference(querier->orig);
+		if (orig_node && atomic_inc_not_zero(&orig_node->refcount)) {
+			/* TODO: include multicast routers via MRD (RFC4286) */
+			batadv_dbg(BATADV_DBG_MCAST, bat_priv,
+				   "Redirecting MLD Report to %pM\n",
+				   orig_node->orig);
+			*orig = orig_node;
+		}
+		rcu_read_unlock();
+
+		return true;
+	case ICMPV6_MGM_QUERY:
+		spin_lock_bh(&querier->orig_lock);
+		orig_node = rcu_dereference(querier->orig);
+		if (orig_node)
+			rcu_assign_pointer(querier->orig, NULL);
+		spin_unlock_bh(&querier->orig_lock);
+
+		batadv_dbg(BATADV_DBG_MCAST, bat_priv,
+			   "Snooped own MLD Query\n");
+		break;
+	}
+
+	return false;
+}
+
 /**
  * batadv_mcast_forw_mode_check_ipv6 - check for optimized forwarding potential
  * @bat_priv: the bat priv with all the soft interface information
  * @skb: the IPv6 packet to check
  * @is_unsnoopable: stores whether the destination is snoopable
+ * @orig: for MLD reports: to be set to the querier to forward the skb to
  *
  * Checks whether the given IPv6 packet has the potential to be forwarded with a
  * mode more optimal than classic flooding.
@@ -321,7 +455,8 @@ static int batadv_mcast_forw_mode_check_ipv4(struct batadv_priv *bat_priv,
  */
 static int batadv_mcast_forw_mode_check_ipv6(struct batadv_priv *bat_priv,
 					     struct sk_buff *skb,
-					     bool *is_unsnoopable)
+					     bool *is_unsnoopable,
+					     struct batadv_orig_node **orig)
 {
 	struct ipv6hdr *ip6hdr;
 
@@ -329,6 +464,9 @@ static int batadv_mcast_forw_mode_check_ipv6(struct batadv_priv *bat_priv,
 	if (!pskb_may_pull(skb, sizeof(struct ethhdr) + sizeof(*ip6hdr)))
 		return -ENOMEM;
 
+	if (batadv_mcast_is_report_ipv6(bat_priv, skb, orig))
+		return 0;
+
 	ip6hdr = ipv6_hdr(skb);
 
 	/* TODO: Implement Multicast Router Discovery (RFC4286),
@@ -345,6 +483,7 @@ static int batadv_mcast_forw_mode_check_ipv6(struct batadv_priv *bat_priv,
 
 	return 0;
 }
+#endif
 
 /**
  * batadv_mcast_forw_mode_check - check for optimized forwarding potential
@@ -360,7 +499,8 @@ static int batadv_mcast_forw_mode_check_ipv6(struct batadv_priv *bat_priv,
  */
 static int batadv_mcast_forw_mode_check(struct batadv_priv *bat_priv,
 					struct sk_buff *skb,
-					bool *is_unsnoopable)
+					bool *is_unsnoopable,
+					struct batadv_orig_node **orig)
 {
 	struct ethhdr *ethhdr = eth_hdr(skb);
 
@@ -373,10 +513,12 @@ static int batadv_mcast_forw_mode_check(struct batadv_priv *bat_priv,
 	switch (ntohs(ethhdr->h_proto)) {
 	case ETH_P_IP:
 		return batadv_mcast_forw_mode_check_ipv4(bat_priv, skb,
-							 is_unsnoopable);
+							 is_unsnoopable, orig);
+#if IS_ENABLED(CONFIG_IPV6)
 	case ETH_P_IPV6:
 		return batadv_mcast_forw_mode_check_ipv6(bat_priv, skb,
-							 is_unsnoopable);
+							 is_unsnoopable, orig);
+#endif
 	default:
 		return -EINVAL;
 	}
@@ -544,12 +686,16 @@ batadv_mcast_forw_mode(struct batadv_priv *bat_priv, struct sk_buff *skb,
 	bool is_unsnoopable = false;
 	struct ethhdr *ethhdr;
 
-	ret = batadv_mcast_forw_mode_check(bat_priv, skb, &is_unsnoopable);
+	ret = batadv_mcast_forw_mode_check(bat_priv, skb, &is_unsnoopable,
+					   orig);
 	if (ret == -ENOMEM)
 		return BATADV_FORW_NONE;
 	else if (ret < 0)
 		return BATADV_FORW_ALL;
 
+	if (*orig)
+		return BATADV_FORW_SINGLE;
+
 	ethhdr = eth_hdr(skb);
 
 	tt_count = batadv_tt_global_hash_count(bat_priv, ethhdr->h_dest,
@@ -581,6 +727,105 @@ batadv_mcast_forw_mode(struct batadv_priv *bat_priv, struct sk_buff *skb,
 }
 
 /**
+ * batadv_mcast_snoop_query_ipv4 - snoop for the selected MLD querier
+ * @skb: the unencapsulated ethernet frame coming from the mesh
+ * @orig_node: the originator this frame came from
+ *
+ * Checks whether the given frame is a general IGMP query from the selected
+ * querier and if so memorizes the originator this frame came from.
+ */
+static void batadv_mcast_snoop_query_ipv4(struct sk_buff *skb,
+					  struct batadv_orig_node *orig_node)
+{
+	struct batadv_mcast_querier_state *querier;
+
+	if (ip_mc_check_igmp(skb, NULL) < 0)
+		return;
+
+	/* we are only interested in general queries (group == 0.0.0.0) */
+	if (igmp_hdr(skb)->type != IGMP_HOST_MEMBERSHIP_QUERY ||
+	    igmp_hdr(skb)->group)
+		return;
+
+	/* RFC4541, section 2.1.1.1.b) says: ignore queries from 0.0.0.0 */
+	if (!ip_hdr(skb)->saddr)
+		return;
+
+	querier = &orig_node->bat_priv->mcast.querier_ipv4;
+
+	spin_lock_bh(&querier->orig_lock);
+	rcu_assign_pointer(querier->orig, orig_node);
+	spin_unlock_bh(&querier->orig_lock);
+
+	batadv_dbg(BATADV_DBG_MCAST, orig_node->bat_priv,
+		   "Snooped IGMP Query from originator %pM\n", orig_node->orig);
+}
+
+#if IS_ENABLED(CONFIG_IPV6)
+/**
+ * batadv_mcast_snoop_query_ipv6 - snoop for the selected MLD querier
+ * @skb: the unencapsulated ethernet frame coming from the mesh
+ * @orig_node: the originator this frame came from
+ *
+ * Checks whether the given frame is a general MLD query from the selected
+ * querier and if so memorizes the originator this frame came from.
+ */
+static void batadv_mcast_snoop_query_ipv6(struct sk_buff *skb,
+					  struct batadv_orig_node *orig_node)
+{
+	struct mld_msg *mld;
+	struct batadv_mcast_querier_state *querier;
+
+	if (ipv6_mc_check_mld(skb, NULL) < 0)
+		return;
+
+	mld = (struct mld_msg *)icmp6_hdr(skb);
+
+	/* we are only interested in general queries (mca == ::) */
+	if (mld->mld_type != ICMPV6_MGM_QUERY ||
+	    !ipv6_addr_any(&mld->mld_mca))
+		return;
+
+	querier = &orig_node->bat_priv->mcast.querier_ipv6;
+
+	spin_lock_bh(&querier->orig_lock);
+	rcu_assign_pointer(querier->orig, orig_node);
+	spin_unlock_bh(&querier->orig_lock);
+
+	batadv_dbg(BATADV_DBG_MCAST, orig_node->bat_priv,
+		   "Snooped MLD Query from originator %pM\n", orig_node->orig);
+}
+#endif
+
+/**
+ * batadv_mcast_snoop_query - snoop the selected IGMP/MLD querier
+ * @skb: the unencapsulated ethernet frame coming from the mesh
+ * @orig_node: the originator this frame came from
+ *
+ * Checks whether the given frame is a general IGMP or MLD query
+ * from the selected querier and if so memorizes the originator
+ * this frame came from.
+ *
+ * This call might reallocate skb data.
+ */
+void batadv_mcast_snoop_query(struct sk_buff *skb,
+			      struct batadv_orig_node *orig_node)
+{
+	struct ethhdr *ethhdr = eth_hdr(skb);
+
+	switch (ntohs(ethhdr->h_proto)) {
+	case ETH_P_IP:
+		batadv_mcast_snoop_query_ipv4(skb, orig_node);
+		break;
+#if IS_ENABLED(CONFIG_IPV6)
+	case ETH_P_IPV6:
+		batadv_mcast_snoop_query_ipv6(skb, orig_node);
+		break;
+#endif
+	}
+}
+
+/**
  * batadv_mcast_want_unsnoop_update - update unsnoop counter and list
  * @bat_priv: the bat priv with all the soft interface information
  * @orig: the orig_node which multicast state might have changed of
diff --git a/net/batman-adv/multicast.h b/net/batman-adv/multicast.h
index beb6e56..fb61db9 100644
--- a/net/batman-adv/multicast.h
+++ b/net/batman-adv/multicast.h
@@ -46,6 +46,9 @@ enum batadv_forw_mode
 batadv_mcast_forw_mode(struct batadv_priv *bat_priv, struct sk_buff *skb,
 		       struct batadv_orig_node **mcast_single_orig);
 
+void batadv_mcast_snoop_query(struct sk_buff *skb,
+			      struct batadv_orig_node *orig_node);
+
 void batadv_mcast_init(struct batadv_priv *bat_priv);
 
 void batadv_mcast_free(struct batadv_priv *bat_priv);
@@ -65,6 +68,11 @@ batadv_mcast_forw_mode(struct batadv_priv *bat_priv, struct sk_buff *skb,
 	return BATADV_FORW_ALL;
 }
 
+static inline void batadv_mcast_snoop_query(struct sk_buff *skb,
+					    struct batadv_orig_node *orig_node)
+{
+}
+
 static inline int batadv_mcast_init(struct batadv_priv *bat_priv)
 {
 	return 0;
diff --git a/net/batman-adv/soft-interface.c b/net/batman-adv/soft-interface.c
index c002961..0a8c2ae 100644
--- a/net/batman-adv/soft-interface.c
+++ b/net/batman-adv/soft-interface.c
@@ -206,6 +206,8 @@ static int batadv_interface_tx(struct sk_buff *skb,
 	if (atomic_read(&bat_priv->mesh_state) != BATADV_MESH_ACTIVE)
 		goto dropped;
 
+	skb_set_network_header(skb, ETH_HLEN);
+
 	soft_iface->trans_start = jiffies;
 	vid = batadv_get_vid(skb, 0);
 	ethhdr = eth_hdr(skb);
@@ -396,6 +398,7 @@ void batadv_interface_rx(struct net_device *soft_iface,
 
 	skb_pull_rcsum(skb, hdr_size);
 	skb_reset_mac_header(skb);
+	skb_set_network_header(skb, ETH_HLEN);
 
 	/* clean the netfilter state now that the batman-adv header has been
 	 * removed
@@ -458,6 +461,9 @@ void batadv_interface_rx(struct net_device *soft_iface,
 			skb->mark &= ~bat_priv->isolation_mark_mask;
 			skb->mark |= bat_priv->isolation_mark;
 		}
+
+		if (orig_node)
+			batadv_mcast_snoop_query(skb, orig_node);
 	} else if (batadv_is_ap_isolated(bat_priv, ethhdr->h_source,
 					 ethhdr->h_dest, vid)) {
 		goto dropped;
@@ -768,6 +774,11 @@ static int batadv_softif_init_late(struct net_device *dev)
 	atomic_set(&bat_priv->distributed_arp_table, 1);
 #endif
 #ifdef CONFIG_BATMAN_ADV_MCAST
+	rcu_assign_pointer(bat_priv->mcast.querier_ipv4.orig, NULL);
+	spin_lock_init(&bat_priv->mcast.querier_ipv4.orig_lock);
+	rcu_assign_pointer(bat_priv->mcast.querier_ipv6.orig, NULL);
+	spin_lock_init(&bat_priv->mcast.querier_ipv6.orig_lock);
+
 	bat_priv->mcast.flags = BATADV_NO_FLAGS;
 	atomic_set(&bat_priv->multicast_mode, 1);
 	atomic_set(&bat_priv->mcast.num_disabled, 0);
diff --git a/net/batman-adv/types.h b/net/batman-adv/types.h
index 67d6348..3a5365f 100644
--- a/net/batman-adv/types.h
+++ b/net/batman-adv/types.h
@@ -640,12 +640,24 @@ struct batadv_priv_dat {
 
 #ifdef CONFIG_BATMAN_ADV_MCAST
 /**
+ * struct batadv_mcast_querier_state - IGMP/MLD querier state when bridged
+ * @orig: node on which the selected querier resides
+ * @orig_lock: protects updates of the selected querier in 'orig'
+ */
+struct batadv_mcast_querier_state {
+	struct batadv_orig_node __rcu *orig; /* rcu protected pointer */
+	spinlock_t orig_lock; /* protects updates of orig */
+};
+
+/**
  * struct batadv_priv_mcast - per mesh interface mcast data
  * @mla_list: list of multicast addresses we are currently announcing via TT
  * @want_all_unsnoopables_list: a list of orig_nodes wanting all unsnoopable
  *  multicast traffic
  * @want_all_ipv4_list: a list of orig_nodes wanting all IPv4 multicast traffic
  * @want_all_ipv6_list: a list of orig_nodes wanting all IPv6 multicast traffic
+ * @querier_ipv4: the current state of an IGMP querier in the mesh
+ * @querier_ipv6: the current state of an MLD querier in the mesh
  * @flags: the flags we have last sent in our mcast tvlv
  * @enabled: whether the multicast tvlv is currently enabled
  * @num_disabled: number of nodes that have no mcast tvlv
@@ -660,6 +672,8 @@ struct batadv_priv_mcast {
 	struct hlist_head want_all_unsnoopables_list;
 	struct hlist_head want_all_ipv4_list;
 	struct hlist_head want_all_ipv6_list;
+	struct batadv_mcast_querier_state querier_ipv4;
+	struct batadv_mcast_querier_state querier_ipv6;
 	uint8_t flags;
 	bool enabled;
 	atomic_t num_disabled;
-- 
1.7.10.4

[B.A.T.M.A.N.] [PATCHv5 2/3] batman-adv: Increase BATADV_TVLV_MCAST version number to 2

[Linus Lüssing]

The multicast optimizations bridge integration will require the just
implemented IGMP/MLD report handling later. Therefore bumping the
version number.

Signed-off-by: Linus Lüssing <linus.luessing@c0d3.blue>
---
 net/batman-adv/multicast.c |   10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/net/batman-adv/multicast.c b/net/batman-adv/multicast.c
index 72d561c..87a5942 100644
--- a/net/batman-adv/multicast.c
+++ b/net/batman-adv/multicast.c
@@ -225,7 +225,7 @@ static bool batadv_mcast_mla_tvlv_update(struct batadv_priv *bat_priv)
 	if (batadv_mcast_has_bridge(bat_priv)) {
 		if (bat_priv->mcast.enabled) {
 			batadv_tvlv_container_unregister(bat_priv,
-							 BATADV_TVLV_MCAST, 1);
+							 BATADV_TVLV_MCAST, 2);
 			bat_priv->mcast.enabled = false;
 		}
 
@@ -234,7 +234,7 @@ static bool batadv_mcast_mla_tvlv_update(struct batadv_priv *bat_priv)
 
 	if (!bat_priv->mcast.enabled ||
 	    mcast_data.flags != bat_priv->mcast.flags) {
-		batadv_tvlv_container_register(bat_priv, BATADV_TVLV_MCAST, 1,
+		batadv_tvlv_container_register(bat_priv, BATADV_TVLV_MCAST, 2,
 					       &mcast_data, sizeof(mcast_data));
 		bat_priv->mcast.flags = mcast_data.flags;
 		bat_priv->mcast.enabled = true;
@@ -984,7 +984,7 @@ static void batadv_mcast_tvlv_ogm_handler_v1(struct batadv_priv *bat_priv,
 void batadv_mcast_init(struct batadv_priv *bat_priv)
 {
 	batadv_tvlv_handler_register(bat_priv, batadv_mcast_tvlv_ogm_handler_v1,
-				     NULL, BATADV_TVLV_MCAST, 1,
+				     NULL, BATADV_TVLV_MCAST, 2,
 				     BATADV_TVLV_HANDLER_OGM_CIFNOTFND);
 }
 
@@ -994,8 +994,8 @@ void batadv_mcast_init(struct batadv_priv *bat_priv)
  */
 void batadv_mcast_free(struct batadv_priv *bat_priv)
 {
-	batadv_tvlv_container_unregister(bat_priv, BATADV_TVLV_MCAST, 1);
-	batadv_tvlv_handler_unregister(bat_priv, BATADV_TVLV_MCAST, 1);
+	batadv_tvlv_container_unregister(bat_priv, BATADV_TVLV_MCAST, 2);
+	batadv_tvlv_handler_unregister(bat_priv, BATADV_TVLV_MCAST, 2);
 
 	batadv_mcast_mla_tt_retract(bat_priv, NULL);
 }
-- 
1.7.10.4

[B.A.T.M.A.N.] [PATCHv5 3/3] compat: Compat code for IGMP/MLD report unicast forwarding

[Linus Lüssing]

Signed-off-by: Linus Lüssing <linus.luessing@c0d3.blue>
---
 Makefile                              |    4 +
 compat-include/linux/igmp.h           |   13 ++
 compat-include/linux/skbuff.h         |   14 +++
 compat-include/net/addrconf.h         |   13 ++
 compat-include/net/ip6_checksum.h     |   18 +++
 compat-include/net/ipv6.h             |   17 +++
 compat-include/net/mld.h              |   52 ++++++++
 compat-sources/Makefile               |    3 +
 compat-sources/net/core/skbuff.c      |  136 +++++++++++++++++++++
 compat-sources/net/ipv4/igmp.c        |  169 ++++++++++++++++++++++++++
 compat-sources/net/ipv6/mcast_snoop.c |  216 +++++++++++++++++++++++++++++++++
 11 files changed, 655 insertions(+)
 create mode 100644 compat-include/linux/igmp.h
 create mode 100644 compat-include/net/addrconf.h
 create mode 100644 compat-include/net/ip6_checksum.h
 create mode 100644 compat-include/net/ipv6.h
 create mode 100644 compat-include/net/mld.h
 create mode 100644 compat-sources/Makefile
 create mode 100644 compat-sources/net/core/skbuff.c
 create mode 100644 compat-sources/net/ipv4/igmp.c
 create mode 100644 compat-sources/net/ipv6/mcast_snoop.c

diff --git a/Makefile b/Makefile
index ee3be1d..69056d7 100644
--- a/Makefile
+++ b/Makefile
@@ -50,6 +50,10 @@ ifneq ($(REVISION),)
 NOSTDINC_FLAGS += -DBATADV_SOURCE_VERSION=\"$(REVISION)\"
 endif
 
+include $(PWD)/compat-sources/Makefile
+
+export batman-adv-y
+
 BUILD_FLAGS := \
 	M=$(PWD)/net/batman-adv \
 	CONFIG_BATMAN_ADV=m \
diff --git a/compat-include/linux/igmp.h b/compat-include/linux/igmp.h
new file mode 100644
index 0000000..f61ab79
--- /dev/null
+++ b/compat-include/linux/igmp.h
@@ -0,0 +1,13 @@
+#ifndef _NET_BATMAN_ADV_COMPAT_LINUX_IGMP_H_
+#define _NET_BATMAN_ADV_COMPAT_LINUX_IGMP_H_
+
+#include <linux/version.h>
+#include_next <linux/igmp.h>
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(4, 2, 0)
+
+int ip_mc_check_igmp(struct sk_buff *skb, struct sk_buff **skb_trimmed);
+
+#endif /* < KERNEL_VERSION(4, 2, 0) */
+
+#endif	/* _NET_BATMAN_ADV_COMPAT_LINUX_IGMP_H_ */
diff --git a/compat-include/linux/skbuff.h b/compat-include/linux/skbuff.h
index d363cc0..1d4f569 100644
--- a/compat-include/linux/skbuff.h
+++ b/compat-include/linux/skbuff.h
@@ -89,6 +89,20 @@ static inline void skb_reset_mac_len(struct sk_buff *skb)
 
 #define pskb_copy_for_clone pskb_copy
 
+__sum16 skb_checksum_simple_validate(struct sk_buff *skb);
+
+__sum16
+skb_checksum_validate(struct sk_buff *skb, int proto,
+		      __wsum (*compute_pseudo)(struct sk_buff *skb, int proto));
+
 #endif /* < KERNEL_VERSION(3, 16, 0) */
 
+#if LINUX_VERSION_CODE < KERNEL_VERSION(4, 2, 0)
+
+struct sk_buff *skb_checksum_trimmed(struct sk_buff *skb,
+				     unsigned int transport_len,
+				     __sum16(*skb_chkf)(struct sk_buff *skb));
+
+#endif /* < KERNEL_VERSION(4, 2, 0) */
+
 #endif	/* _NET_BATMAN_ADV_COMPAT_LINUX_SKBUFF_H_ */
diff --git a/compat-include/net/addrconf.h b/compat-include/net/addrconf.h
new file mode 100644
index 0000000..69c45d0
--- /dev/null
+++ b/compat-include/net/addrconf.h
@@ -0,0 +1,13 @@
+#ifndef _NET_BATMAN_ADV_COMPAT_NET_ADDRCONF_H_
+#define _NET_BATMAN_ADV_COMPAT_NET_ADDRCONF_H_
+
+#include <linux/version.h>
+#include_next <net/addrconf.h>
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(4, 2, 0)
+
+int ipv6_mc_check_mld(struct sk_buff *skb, struct sk_buff **skb_trimmed);
+
+#endif /* < KERNEL_VERSION(4, 2, 0) */
+
+#endif	/* _NET_BATMAN_ADV_COMPAT_NET_ADDRCONF_H_ */
diff --git a/compat-include/net/ip6_checksum.h b/compat-include/net/ip6_checksum.h
new file mode 100644
index 0000000..fda0c07
--- /dev/null
+++ b/compat-include/net/ip6_checksum.h
@@ -0,0 +1,18 @@
+#ifndef _NET_BATMAN_ADV_COMPAT_NET_IP6_CHECKSUM_H_
+#define _NET_BATMAN_ADV_COMPAT_NET_IP6_CHECKSUM_H_
+
+#include <linux/version.h>
+#include_next <net/ip6_checksum.h>
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(3, 16, 0)
+
+static inline __wsum ip6_compute_pseudo(struct sk_buff *skb, int proto)
+{
+	return ~csum_unfold(csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
+					    &ipv6_hdr(skb)->daddr,
+					    skb->len, proto, 0));
+}
+
+#endif /* < KERNEL_VERSION(3, 16, 0) */
+
+#endif	/* _NET_BATMAN_ADV_COMPAT_NET_IP6_CHECKSUM_H_ */
diff --git a/compat-include/net/ipv6.h b/compat-include/net/ipv6.h
new file mode 100644
index 0000000..1e190d8
--- /dev/null
+++ b/compat-include/net/ipv6.h
@@ -0,0 +1,17 @@
+#ifndef _NET_BATMAN_ADV_COMPAT_NET_IPV6_H_
+#define _NET_BATMAN_ADV_COMPAT_NET_IPV6_H_
+
+#include <linux/version.h>
+#include_next <net/ipv6.h>
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(3, 3, 0)
+
+#define ipv6_skip_exthdr(skb, start, nexthdrp, frag_offp) \
+	({ \
+		(void)frag_offp; \
+		ipv6_skip_exthdr(skb, start, nexthdrp); \
+	})
+
+#endif /* < KERNEL_VERSION(3, 3, 0) */
+
+#endif /* _NET_BATMAN_ADV_COMPAT_NET_IPV6_H_ */
diff --git a/compat-include/net/mld.h b/compat-include/net/mld.h
new file mode 100644
index 0000000..e041eb6
--- /dev/null
+++ b/compat-include/net/mld.h
@@ -0,0 +1,52 @@
+#ifndef _NET_BATMAN_ADV_COMPAT_NET_MLD_H_
+#define _NET_BATMAN_ADV_COMPAT_NET_MLD_H_
+
+#include <linux/version.h>
+
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2, 6, 35)
+#include_next <net/mld.h>
+#endif /* >= KERNEL_VERSION(2, 6, 35) */
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 35)
+struct mld_msg {
+	struct icmp6hdr		mld_hdr;
+	struct in6_addr		mld_mca;
+};
+
+#define mld_type		mld_hdr.icmp6_type
+
+struct mld2_grec {
+	__u8		grec_type;
+	__u8		grec_auxwords;
+	__be16		grec_nsrcs;
+	struct in6_addr	grec_mca;
+	struct in6_addr	grec_src[0];
+};
+
+struct mld2_report {
+	struct icmp6hdr		mld2r_hdr;
+	struct mld2_grec	mld2r_grec[0];
+};
+
+struct mld2_query {
+	struct icmp6hdr		mld2q_hdr;
+	struct in6_addr		mld2q_mca;
+#if defined(__LITTLE_ENDIAN_BITFIELD)
+	__u8			mld2q_qrv:3,
+				mld2q_suppress:1,
+				mld2q_resv2:4;
+#elif defined(__BIG_ENDIAN_BITFIELD)
+	__u8			mld2q_resv2:4,
+				mld2q_suppress:1,
+				mld2q_qrv:3;
+#else
+#error "Please fix <asm/byteorder.h>"
+#endif
+	__u8			mld2q_qqic;
+	__be16			mld2q_nsrcs;
+	struct in6_addr		mld2q_srcs[0];
+};
+
+#endif /* < KERNEL_VERSION(2, 6, 35) */
+
+#endif	/* _NET_BATMAN_ADV_COMPAT_NET_MLD_H_ */
diff --git a/compat-sources/Makefile b/compat-sources/Makefile
new file mode 100644
index 0000000..c364ded
--- /dev/null
+++ b/compat-sources/Makefile
@@ -0,0 +1,3 @@
+batman-adv-y += ../../compat-sources/net/core/skbuff.o
+batman-adv-y += ../../compat-sources/net/ipv4/igmp.o
+batman-adv-y += ../../compat-sources/net/ipv6/mcast_snoop.o
diff --git a/compat-sources/net/core/skbuff.c b/compat-sources/net/core/skbuff.c
new file mode 100644
index 0000000..eb15adf
--- /dev/null
+++ b/compat-sources/net/core/skbuff.c
@@ -0,0 +1,136 @@
+#include <linux/ipv6.h>
+#include <linux/skbuff.h>
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(3, 16, 0)
+
+__sum16 skb_checksum_simple_validate(struct sk_buff *skb)
+{
+	switch (skb->ip_summed) {
+	case CHECKSUM_COMPLETE:
+		if (!csum_fold(skb->csum))
+			break;
+		/* fall through */
+	case CHECKSUM_NONE:
+		skb->csum = 0;
+		if (skb_checksum_complete(skb))
+			return -EINVAL;
+	}
+
+	return 0;
+}
+
+__sum16
+skb_checksum_validate(struct sk_buff *skb, int proto,
+		      __wsum (*compute_pseudo)(struct sk_buff *skb, int proto))
+{
+	const struct ipv6hdr *ip6h = ipv6_hdr(skb);
+
+	switch (skb->ip_summed) {
+	case CHECKSUM_COMPLETE:
+		if (!csum_ipv6_magic(&ip6h->saddr, &ip6h->daddr, skb->len,
+				     IPPROTO_ICMPV6, skb->csum))
+			break;
+		/*FALLTHROUGH*/
+	case CHECKSUM_NONE:
+		skb->csum = ~csum_unfold(csum_ipv6_magic(&ip6h->saddr,
+							 &ip6h->daddr,
+							 skb->len,
+							 IPPROTO_ICMPV6, 0));
+		if (__skb_checksum_complete(skb))
+			return -EINVAL;
+	}
+
+	return 0;
+}
+
+#endif /* < KERNEL_VERSION(3, 16, 0) */
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(4, 2, 0)
+
+/**
+ * skb_checksum_maybe_trim - maybe trims the given skb
+ * @skb: the skb to check
+ * @transport_len: the data length beyond the network header
+ *
+ * Checks whether the given skb has data beyond the given transport length.
+ * If so, returns a cloned skb trimmed to this transport length.
+ * Otherwise returns the provided skb. Returns NULL in error cases
+ * (e.g. transport_len exceeds skb length or out-of-memory).
+ *
+ * Caller needs to set the skb transport header and release the returned skb.
+ * Provided skb is consumed.
+ */
+static struct sk_buff *skb_checksum_maybe_trim(struct sk_buff *skb,
+					       unsigned int transport_len)
+{
+	struct sk_buff *skb_chk;
+	unsigned int len = skb_transport_offset(skb) + transport_len;
+	int ret;
+
+	if (skb->len < len) {
+		kfree_skb(skb);
+		return NULL;
+	} else if (skb->len == len) {
+		return skb;
+	}
+
+	skb_chk = skb_clone(skb, GFP_ATOMIC);
+	kfree_skb(skb);
+
+	if (!skb_chk)
+		return NULL;
+
+	ret = pskb_trim_rcsum(skb_chk, len);
+	if (ret) {
+		kfree_skb(skb_chk);
+		return NULL;
+	}
+
+	return skb_chk;
+}
+
+/**
+ * skb_checksum_trimmed - validate checksum of an skb
+ * @skb: the skb to check
+ * @transport_len: the data length beyond the network header
+ * @skb_chkf: checksum function to use
+ *
+ * Applies the given checksum function skb_chkf to the provided skb.
+ * Returns a checked and maybe trimmed skb. Returns NULL on error.
+ *
+ * If the skb has data beyond the given transport length, then a
+ * trimmed & cloned skb is checked and returned.
+ *
+ * Caller needs to set the skb transport header and release the returned skb.
+ * Provided skb is consumed.
+ */
+struct sk_buff *skb_checksum_trimmed(struct sk_buff *skb,
+				     unsigned int transport_len,
+				     __sum16(*skb_chkf)(struct sk_buff *skb))
+{
+	struct sk_buff *skb_chk;
+	unsigned int offset = skb_transport_offset(skb);
+	__sum16 ret;
+
+	skb_chk = skb_checksum_maybe_trim(skb, transport_len);
+	if (!skb_chk)
+		return NULL;
+
+	if (!pskb_may_pull(skb_chk, offset)) {
+		kfree_skb(skb_chk);
+		return NULL;
+	}
+
+	__skb_pull(skb_chk, offset);
+	ret = skb_chkf(skb_chk);
+	__skb_push(skb_chk, offset);
+
+	if (ret) {
+		kfree_skb(skb_chk);
+		return NULL;
+	}
+
+	return skb_chk;
+}
+
+#endif /* < KERNEL_VERSION(4, 2, 0) */
diff --git a/compat-sources/net/ipv4/igmp.c b/compat-sources/net/ipv4/igmp.c
new file mode 100644
index 0000000..7dd69d7
--- /dev/null
+++ b/compat-sources/net/ipv4/igmp.c
@@ -0,0 +1,169 @@
+#if LINUX_VERSION_CODE < KERNEL_VERSION(4, 2, 0)
+
+#include <linux/igmp.h>
+#include <linux/ip.h>
+#include <linux/skbuff.h>
+#include <net/ip.h>
+
+static int ip_mc_check_iphdr(struct sk_buff *skb)
+{
+	const struct iphdr *iph;
+	unsigned int len;
+	unsigned int offset = skb_network_offset(skb) + sizeof(*iph);
+
+	if (!pskb_may_pull(skb, offset))
+		return -EINVAL;
+
+	iph = ip_hdr(skb);
+
+	if (iph->version != 4 || ip_hdrlen(skb) < sizeof(*iph))
+		return -EINVAL;
+
+	offset += ip_hdrlen(skb) - sizeof(*iph);
+
+	if (!pskb_may_pull(skb, offset))
+		return -EINVAL;
+
+	iph = ip_hdr(skb);
+
+	if (unlikely(ip_fast_csum((u8 *)iph, iph->ihl)))
+		return -EINVAL;
+
+	len = skb_network_offset(skb) + ntohs(iph->tot_len);
+	if (skb->len < len || len < offset)
+		return -EINVAL;
+
+	skb_set_transport_header(skb, offset);
+
+	return 0;
+}
+
+static int ip_mc_check_igmp_reportv3(struct sk_buff *skb)
+{
+	unsigned int len = skb_transport_offset(skb);
+
+	len += sizeof(struct igmpv3_report);
+
+	return pskb_may_pull(skb, len) ? 0 : -EINVAL;
+}
+
+static int ip_mc_check_igmp_query(struct sk_buff *skb)
+{
+	unsigned int len = skb_transport_offset(skb);
+
+	len += sizeof(struct igmphdr);
+	if (skb->len < len)
+		return -EINVAL;
+
+	/* IGMPv{1,2}? */
+	if (skb->len != len) {
+		/* or IGMPv3? */
+		len += sizeof(struct igmpv3_query) - sizeof(struct igmphdr);
+		if (skb->len < len || !pskb_may_pull(skb, len))
+			return -EINVAL;
+	}
+
+	/* RFC2236+RFC3376 (IGMPv2+IGMPv3) require the multicast link layer
+	 * all-systems destination addresses (224.0.0.1) for general queries
+	 */
+	if (!igmp_hdr(skb)->group &&
+	    ip_hdr(skb)->daddr != htonl(INADDR_ALLHOSTS_GROUP))
+		return -EINVAL;
+
+	return 0;
+}
+
+static int ip_mc_check_igmp_msg(struct sk_buff *skb)
+{
+	switch (igmp_hdr(skb)->type) {
+	case IGMP_HOST_LEAVE_MESSAGE:
+	case IGMP_HOST_MEMBERSHIP_REPORT:
+	case IGMPV2_HOST_MEMBERSHIP_REPORT:
+		/* fall through */
+		return 0;
+	case IGMPV3_HOST_MEMBERSHIP_REPORT:
+		return ip_mc_check_igmp_reportv3(skb);
+	case IGMP_HOST_MEMBERSHIP_QUERY:
+		return ip_mc_check_igmp_query(skb);
+	default:
+		return -ENOMSG;
+	}
+}
+
+static inline __sum16 ip_mc_validate_checksum(struct sk_buff *skb)
+{
+	return skb_checksum_simple_validate(skb);
+}
+
+static int __ip_mc_check_igmp(struct sk_buff *skb, struct sk_buff **skb_trimmed)
+
+{
+	struct sk_buff *skb_chk;
+	unsigned int transport_len;
+	unsigned int len = skb_transport_offset(skb) + sizeof(struct igmphdr);
+	int ret;
+
+	transport_len = ntohs(ip_hdr(skb)->tot_len) - ip_hdrlen(skb);
+
+	skb_get(skb);
+	skb_chk = skb_checksum_trimmed(skb, transport_len,
+				       ip_mc_validate_checksum);
+	if (!skb_chk)
+		return -EINVAL;
+
+	if (!pskb_may_pull(skb_chk, len)) {
+		kfree_skb(skb_chk);
+		return -EINVAL;
+	}
+
+	ret = ip_mc_check_igmp_msg(skb_chk);
+	if (ret) {
+		kfree_skb(skb_chk);
+		return ret;
+	}
+
+	if (skb_trimmed)
+		*skb_trimmed = skb_chk;
+	else
+		kfree_skb(skb_chk);
+
+	return 0;
+}
+
+/**
+ * ip_mc_check_igmp - checks whether this is a sane IGMP packet
+ * @skb: the skb to validate
+ * @skb_trimmed: to store an skb pointer trimmed to IPv4 packet tail (optional)
+ *
+ * Checks whether an IPv4 packet is a valid IGMP packet. If so sets
+ * skb network and transport headers accordingly and returns zero.
+ *
+ * -EINVAL: A broken packet was detected, i.e. it violates some internet
+ *  standard
+ * -ENOMSG: IP header validation succeeded but it is not an IGMP packet.
+ * -ENOMEM: A memory allocation failure happened.
+ *
+ * Optionally, an skb pointer might be provided via skb_trimmed (or set it
+ * to NULL): After parsing an IGMP packet successfully it will point to
+ * an skb which has its tail aligned to the IP packet end. This might
+ * either be the originally provided skb or a trimmed, cloned version if
+ * the skb frame had data beyond the IP packet. A cloned skb allows us
+ * to leave the original skb and its full frame unchanged (which might be
+ * desirable for layer 2 frame jugglers).
+ *
+ * The caller needs to release a reference count from any returned skb_trimmed.
+ */
+int ip_mc_check_igmp(struct sk_buff *skb, struct sk_buff **skb_trimmed)
+{
+	int ret = ip_mc_check_iphdr(skb);
+
+	if (ret < 0)
+		return ret;
+
+	if (ip_hdr(skb)->protocol != IPPROTO_IGMP)
+		return -ENOMSG;
+
+	return __ip_mc_check_igmp(skb, skb_trimmed);
+}
+
+#endif /* < KERNEL_VERSION(4, 2, 0) */
diff --git a/compat-sources/net/ipv6/mcast_snoop.c b/compat-sources/net/ipv6/mcast_snoop.c
new file mode 100644
index 0000000..32a57bc
--- /dev/null
+++ b/compat-sources/net/ipv6/mcast_snoop.c
@@ -0,0 +1,216 @@
+#if LINUX_VERSION_CODE < KERNEL_VERSION(4, 2, 0)
+
+/* Copyright (C) 2010: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
+ * Copyright (C) 2015: Linus Lüssing <linus.luessing@c0d3.blue>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of version 2 of the GNU General Public
+ * License as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, see <http://www.gnu.org/licenses/>.
+ *
+ *
+ * Based on the MLD support added to br_multicast.c by YOSHIFUJI Hideaki.
+ */
+
+#include <linux/skbuff.h>
+#include <net/ipv6.h>
+#include <net/mld.h>
+#include <net/addrconf.h>
+#include <net/ip6_checksum.h>
+
+static int ipv6_mc_check_ip6hdr(struct sk_buff *skb)
+{
+	const struct ipv6hdr *ip6h;
+	unsigned int len;
+	unsigned int offset = skb_network_offset(skb) + sizeof(*ip6h);
+
+	if (!pskb_may_pull(skb, offset))
+		return -EINVAL;
+
+	ip6h = ipv6_hdr(skb);
+
+	if (ip6h->version != 6)
+		return -EINVAL;
+
+	len = offset + ntohs(ip6h->payload_len);
+	if (skb->len < len || len <= offset)
+		return -EINVAL;
+
+	return 0;
+}
+
+static int ipv6_mc_check_exthdrs(struct sk_buff *skb)
+{
+	const struct ipv6hdr *ip6h;
+	int offset;
+	u8 nexthdr;
+	__be16 frag_off;
+
+	ip6h = ipv6_hdr(skb);
+
+	if (ip6h->nexthdr != IPPROTO_HOPOPTS)
+		return -ENOMSG;
+
+	nexthdr = ip6h->nexthdr;
+	offset = skb_network_offset(skb) + sizeof(*ip6h);
+	offset = ipv6_skip_exthdr(skb, offset, &nexthdr, &frag_off);
+
+	if (offset < 0)
+		return -EINVAL;
+
+	if (nexthdr != IPPROTO_ICMPV6)
+		return -ENOMSG;
+
+	skb_set_transport_header(skb, offset);
+
+	return 0;
+}
+
+static int ipv6_mc_check_mld_reportv2(struct sk_buff *skb)
+{
+	unsigned int len = skb_transport_offset(skb);
+
+	len += sizeof(struct mld2_report);
+
+	return pskb_may_pull(skb, len) ? 0 : -EINVAL;
+}
+
+static int ipv6_mc_check_mld_query(struct sk_buff *skb)
+{
+	struct mld_msg *mld;
+	unsigned int len = skb_transport_offset(skb);
+
+	/* RFC2710+RFC3810 (MLDv1+MLDv2) require link-local source addresses */
+	if (!(ipv6_addr_type(&ipv6_hdr(skb)->saddr) & IPV6_ADDR_LINKLOCAL))
+		return -EINVAL;
+
+	len += sizeof(struct mld_msg);
+	if (skb->len < len)
+		return -EINVAL;
+
+	/* MLDv1? */
+	if (skb->len != len) {
+		/* or MLDv2? */
+		len += sizeof(struct mld2_query) - sizeof(struct mld_msg);
+		if (skb->len < len || !pskb_may_pull(skb, len))
+			return -EINVAL;
+	}
+
+	mld = (struct mld_msg *)skb_transport_header(skb);
+
+	/* RFC2710+RFC3810 (MLDv1+MLDv2) require the multicast link layer
+	 * all-nodes destination address (ff02::1) for general queries
+	 */
+	if (ipv6_addr_any(&mld->mld_mca) &&
+	    !ipv6_addr_is_ll_all_nodes(&ipv6_hdr(skb)->daddr))
+		return -EINVAL;
+
+	return 0;
+}
+
+static int ipv6_mc_check_mld_msg(struct sk_buff *skb)
+{
+	struct mld_msg *mld = (struct mld_msg *)skb_transport_header(skb);
+
+	switch (mld->mld_type) {
+	case ICMPV6_MGM_REDUCTION:
+	case ICMPV6_MGM_REPORT:
+		/* fall through */
+		return 0;
+	case ICMPV6_MLD2_REPORT:
+		return ipv6_mc_check_mld_reportv2(skb);
+	case ICMPV6_MGM_QUERY:
+		return ipv6_mc_check_mld_query(skb);
+	default:
+		return -ENOMSG;
+	}
+}
+
+static inline __sum16 ipv6_mc_validate_checksum(struct sk_buff *skb)
+{
+	return skb_checksum_validate(skb, IPPROTO_ICMPV6, ip6_compute_pseudo);
+}
+
+static int __ipv6_mc_check_mld(struct sk_buff *skb,
+			       struct sk_buff **skb_trimmed)
+
+{
+	struct sk_buff *skb_chk = NULL;
+	unsigned int transport_len;
+	unsigned int len = skb_transport_offset(skb) + sizeof(struct mld_msg);
+	int ret;
+
+	transport_len = ntohs(ipv6_hdr(skb)->payload_len);
+	transport_len -= skb_transport_offset(skb) - sizeof(struct ipv6hdr);
+
+	skb_get(skb);
+	skb_chk = skb_checksum_trimmed(skb, transport_len,
+				       ipv6_mc_validate_checksum);
+	if (!skb_chk)
+		return -EINVAL;
+
+	if (!pskb_may_pull(skb_chk, len)) {
+		kfree_skb(skb_chk);
+		return -EINVAL;
+	}
+
+	ret = ipv6_mc_check_mld_msg(skb_chk);
+	if (ret) {
+		kfree_skb(skb_chk);
+		return ret;
+	}
+
+	if (skb_trimmed)
+		*skb_trimmed = skb_chk;
+	else
+		kfree_skb(skb_chk);
+
+	return 0;
+}
+
+/**
+ * ipv6_mc_check_mld - checks whether this is a sane MLD packet
+ * @skb: the skb to validate
+ * @skb_trimmed: to store an skb pointer trimmed to IPv6 packet tail (optional)
+ *
+ * Checks whether an IPv6 packet is a valid MLD packet. If so sets
+ * skb network and transport headers accordingly and returns zero.
+ *
+ * -EINVAL: A broken packet was detected, i.e. it violates some internet
+ *  standard
+ * -ENOMSG: IP header validation succeeded but it is not an MLD packet.
+ * -ENOMEM: A memory allocation failure happened.
+ *
+ * Optionally, an skb pointer might be provided via skb_trimmed (or set it
+ * to NULL): After parsing an MLD packet successfully it will point to
+ * an skb which has its tail aligned to the IP packet end. This might
+ * either be the originally provided skb or a trimmed, cloned version if
+ * the skb frame had data beyond the IP packet. A cloned skb allows us
+ * to leave the original skb and its full frame unchanged (which might be
+ * desirable for layer 2 frame jugglers).
+ *
+ * The caller needs to release a reference count from any returned skb_trimmed.
+ */
+int ipv6_mc_check_mld(struct sk_buff *skb, struct sk_buff **skb_trimmed)
+{
+	int ret;
+
+	ret = ipv6_mc_check_ip6hdr(skb);
+	if (ret < 0)
+		return ret;
+
+	ret = ipv6_mc_check_exthdrs(skb);
+	if (ret < 0)
+		return ret;
+
+	return __ipv6_mc_check_mld(skb, skb_trimmed);
+}
+
+#endif /* < KERNEL_VERSION(4, 2, 0) */
-- 
1.7.10.4

Re: [B.A.T.M.A.N.] [PATCHv5 0/3] batman-adv: Unicasting multicast reports to querier-node only

[Linus Lüssing]

PS: If you like the new compat solution better, then please let me
know what copyright header you'd like to have added.

Currently only compat-sources/net/ipv6/mcast_snoop.c has a header,
identical to the one in the upstream kernel. For the other files
I'm a little unsure about what to add. For instance
compat-sources/net/ipv4/igmp.c is a lot based on Herbert Xu's
work (similar to how mcast_snoop.c is based on Yoshifuji's work).


On Wed, May 06, 2015 at 09:29:19PM +0200, Linus Lüssing wrote:
> The last round of multicast patches send to the batman-adv mailinglist
> to add support for the multicast optimizations in bridged scenarios, too,
> unfortunately had one major conceptual flaw: It could lead to packet loss.
> It's not sufficient to have the unicasting of reports implemented on
> bridge-nodes only. Nodes without bridges need to treat reports the same
> way.
> 
> The issue is described in detail here:
> 
> https://www.open-mesh.org/projects/batman-adv/wiki/Multicast-optimizations-listener-reports
> 
> 
> Cheers, Linus
> 
> -----
> 
> Changelog:
> v5:
>  * Removed RFC tag: Needed exports got merged to net-next and are going to
>    be available with Linux 4.2
>  * Redid compat solution - now fully backwards compatible down to 2.6.33
> v4:
>  * excluded bridge part from this patchset, they should
>    hopefully be added to net-next soon
>  * Added a compat solution (PATCH 3/3)
>  * Removed Kconfig-depends as by David's suggestion the needed parsing
>    functions for MLD are going to be forced built-ins even if IPv6 is
>    going to be built as a module
>  * Removed unused variable 'int ret' in batadv_mcast_is_report_ipv6()
>  * Adjusted to new folder structure
> v3:
>  * Adding Kconfig-depends and #if's
>    (so basically adding similar dependancy constraints as the bridge code
>     has, except that there are no depends if batman-adv gets compiled without
>     multicast optimizations)
>    -> the case of IPv6=M and batman-adv=y is still impossible if multicast
>       optimizations are enabled; but I don't see the practical demand for that
>       either - people who use IPv6 as a module will probably also want to
>       use batman-adv as a module
> v2:
>  * various bugfixes (now runtime tested, too - should(tm) work)
>  * added netdev+bridge mailinglists

Re: [B.A.T.M.A.N.] [PATCHv5 3/3] compat: Compat code for IGMP/MLD report unicast forwarding

[Sven Eckelmann]

On Wednesday 06 May 2015 21:29:22 Linus Lüssing wrote:
> Signed-off-by: Linus Lüssing <linus.luessing@c0d3.blue>
> ---
>  Makefile                              |    4 +
>  compat-include/linux/igmp.h           |   13 ++
>  compat-include/linux/skbuff.h         |   14 +++
>  compat-include/net/addrconf.h         |   13 ++
>  compat-include/net/ip6_checksum.h     |   18 +++
>  compat-include/net/ipv6.h             |   17 +++
>  compat-include/net/mld.h              |   52 ++++++++
>  compat-sources/Makefile               |    3 +
>  compat-sources/net/core/skbuff.c      |  136 +++++++++++++++++++++
>  compat-sources/net/ipv4/igmp.c        |  169 ++++++++++++++++++++++++++
>  compat-sources/net/ipv6/mcast_snoop.c |  216
> +++++++++++++++++++++++++++++++++ 11 files changed, 655 insertions(+)
>  create mode 100644 compat-include/linux/igmp.h
>  create mode 100644 compat-include/net/addrconf.h
>  create mode 100644 compat-include/net/ip6_checksum.h
>  create mode 100644 compat-include/net/ipv6.h
>  create mode 100644 compat-include/net/mld.h
>  create mode 100644 compat-sources/Makefile
>  create mode 100644 compat-sources/net/core/skbuff.c
>  create mode 100644 compat-sources/net/ipv4/igmp.c
>  create mode 100644 compat-sources/net/ipv6/mcast_snoop.c

The split header files were only there to fix the problem of the header mess 
which couldn't be resolved with a single compat.h that forced a specific 
include order. It is not really necessary to have the same for the source 
files. Maybe your solution is good or maybe not. Only the future will tell. :)

The other solution would be more like backports [1]. They seem to use one 
backport-/compat-*.c file for each kernel they support and introduced features 
which they require. So all kernel < 3.19 will build the content of 
backport-3.19.c to get the functions introduced by this version. But there are 
also some larger ones for special stuff. Maybe you can check this out and 
decide what you like more.

To the copyright stuff... hm, if you didn't write it yourself then the 
copyright stuff from the original source should be there. backports [1] 
doesn't seem to do this very well. They have for example used larger portions 
from files marked as copyright Yu Zhao <yu.zhao@intel.com> but have only the 
copyright header with Copyright (c) 2013  Luis R. Rodriguez <mcgrof@do-not-
panic.com>. Not sure that I can give a good answer here. Maybe someone else 
has an opinion about this part

To the Makefile integration: I did it with compat.c similar before dropping 
it. The only difference was that I only had one file

    export batman-adv-y += ../../compat.o

So you solution seems to be fine.

Kind regards,
	Sven


[1] https://www.kernel.org/pub/linux/kernel/projects/backports/stable/

Re: [B.A.T.M.A.N.] [PATCHv5 3/3] compat: Compat code for IGMP/MLD report unicast forwarding

[Linus Lüssing]

On Thu, May 07, 2015 at 09:33:17AM +0200, Sven Eckelmann wrote:
> The split header files were only there to fix the problem of the header mess 
> which couldn't be resolved with a single compat.h that forced a specific 
> include order. It is not really necessary to have the same for the source 
> files. Maybe your solution is good or maybe not. Only the future will tell. :)
> 
> The other solution would be more like backports [1]. They seem to use one 
> backport-/compat-*.c file for each kernel they support and introduced features 
> which they require. So all kernel < 3.19 will build the content of 
> backport-3.19.c to get the functions introduced by this version. But there are 
> also some larger ones for special stuff. Maybe you can check this out and 
> decide what you like more.

For the backport compat code I'm always getting confused with the
kernel versions. But maybe that's more because their latest
downloads, for instance the one with 4.0 in its name, is actually
not just for 4.0 but older kernels, too (unless I'm again mixing
things up...). With our current implementation I like it quite a
lot that while reading it's always clear which code is being used
thanks to the explicit #if's and #endifs. And by keeping the
upstream folder structure, it's immediately clear where code came
from, too (though that could be figured out via git-grep as well,
of course).

> 
> To the copyright stuff... hm, if you didn't write it yourself then the 
> copyright stuff from the original source should be there. backports [1] 
> doesn't seem to do this very well. They have for example used larger portions 
> from files marked as copyright Yu Zhao <yu.zhao@intel.com> but have only the 
> copyright header with Copyright (c) 2013  Luis R. Rodriguez <mcgrof@do-not-
> panic.com>. Not sure that I can give a good answer here. Maybe someone else 
> has an opinion about this part

Had some small discussions with Simon today, too. I now simply copied the
copyright headers from the upstream files. Even though I know that
ip_mc_check_igmp() in igmp.c was not written by Alan Cox (more by
Herbert Xu with refactoring by me). Or the copied skbuff.c stuff
wasn't written by Alan Cox/Florian La Roche either (where some
stuff was written by me and some was moved from br_multicast.c
where I don't know who actually wrote that - would need to do
deeper git-blame analysis...).

But I guess if people didn't claim copyright for things in the
header of the C file, then I'd assume they are okay with what we
do. Otherwise they would need to complain in the upstream kernel
too or everyone else would need to do hours of git history digging
and analysis to find out where input might have come from.

As always, thanks for your valuable feedback!

Cheers, Linus