Message-ID: <CAJceWd-o4ubk=-rC_3DQfj55QqRMPr4T5BF1odxdiz9Gk1J1Bg@mail.gmail.com>
Date:   Tue, 15 Feb 2022 02:53:49 +0530
From:   Mobashshera Rasool <mobash.rasool.linux@...il.com>
To:     netdev@...r.kernel.org, davem@...emloft.net, kuznet@....inr.ac.ru,
        yoshfuji@...ux-ipv6.org, kuba@...nel.org
Cc:     mrasool@...are.com, equinox@...nsourcerouting.org,
        mobash.rasool@...il.com
Subject: [PATCH net-next] net: ip6mr: add support for passing full packet on
 wrong mif

This patch adds support for MRT6MSG_WRMIFWHOLE which is used to pass
full packet and real vif id when the incoming interface is wrong.
While the RP and FHR are setting up state we need to be sending the
registers encapsulated with all the data inside otherwise we lose it.
The RP then decapsulates it and forwards it to the interested parties.
Currently with WRONGMIF we can only be sending empty register packets
and will lose that data.
This behaviour can be enabled by using MRT_PIM with
val == MRT6MSG_WRMIFWHOLE. This doesn't prevent MRT6MSG_WRONGMIF from
happening, it happens in addition to it, also it is controlled by the same
throttling parameters as WRONGMIF (i.e. 1 packet per 3 seconds currently).
Both messages are generated to keep backwards compatibily and avoid
breaking someone who was enabling MRT_PIM with val == 4, since any
positive val is accepted and treated the same.

Signed-off-by: Mobashshera Rasool <mobash.rasool@...il.com>
---
 include/uapi/linux/mroute6.h |  1 +
 net/ipv6/ip6mr.c             | 18 ++++++++++++++----
 2 files changed, 15 insertions(+), 4 deletions(-)

diff --git a/include/uapi/linux/mroute6.h b/include/uapi/linux/mroute6.h
index a1fd617..1d90c21 100644
--- a/include/uapi/linux/mroute6.h
+++ b/include/uapi/linux/mroute6.h
@@ -134,6 +134,7 @@ struct mrt6msg {
 #define MRT6MSG_NOCACHE         1
 #define MRT6MSG_WRONGMIF        2
 #define MRT6MSG_WHOLEPKT        3               /* used for use level encap */
+#define MRT6MSG_WRMIFWHOLE     4               /* For PIM Register
and assert processing */
         __u8            im6_mbz;                /* must be zero            */
         __u8            im6_msgtype;            /* what type of message    */
         __u16           im6_mif;                /* mif rec'd on            */
diff --git a/net/ipv6/ip6mr.c b/net/ipv6/ip6mr.c
index 7cf73e6..1eed315 100644
--- a/net/ipv6/ip6mr.c
+++ b/net/ipv6/ip6mr.c
@@ -1040,7 +1040,7 @@ static int ip6mr_cache_report(struct mr_table
*mrt, struct sk_buff *pkt,
         int ret;

 #ifdef CONFIG_IPV6_PIMSM_V2
-       if (assert == MRT6MSG_WHOLEPKT)
+       if (assert == MRT6MSG_WHOLEPKT || assert == MRT6MSG_WRMIFWHOLE)
                 skb = skb_realloc_headroom(pkt, -skb_network_offset(pkt)
                                                 +sizeof(*msg));
         else
@@ -1056,7 +1056,7 @@ static int ip6mr_cache_report(struct mr_table
*mrt, struct sk_buff *pkt,
         skb->ip_summed = CHECKSUM_UNNECESSARY;

 #ifdef CONFIG_IPV6_PIMSM_V2
-       if (assert == MRT6MSG_WHOLEPKT) {
+       if (assert == MRT6MSG_WHOLEPKT || assert == MRT6MSG_WRMIFWHOLE) {
                 /* Ugly, but we have no choice with this interface.
                    Duplicate old header, fix length etc.
                    And all this only to mangle msg->im6_msgtype and
@@ -1068,8 +1068,11 @@ static int ip6mr_cache_report(struct mr_table
*mrt, struct sk_buff *pkt,
                 skb_reset_transport_header(skb);
                 msg = (struct mrt6msg *)skb_transport_header(skb);
                 msg->im6_mbz = 0;
-               msg->im6_msgtype = MRT6MSG_WHOLEPKT;
-               msg->im6_mif = mrt->mroute_reg_vif_num;
+               msg->im6_msgtype = assert;
+               if (assert == MRT6MSG_WRMIFWHOLE)
+                       msg->im6_mif = mifi;
+               else
+                       msg->im6_mif = mrt->mroute_reg_vif_num;
                 msg->im6_pad = 0;
                 msg->im6_src = ipv6_hdr(pkt)->saddr;
                 msg->im6_dst = ipv6_hdr(pkt)->daddr;
@@ -1633,6 +1636,7 @@ int ip6_mroute_setsockopt(struct sock *sk, int
optname, sockptr_t optval,
         mifi_t mifi;
         struct net *net = sock_net(sk);
         struct mr_table *mrt;
+       bool do_wrmifwhole;

         if (sk->sk_type != SOCK_RAW ||
             inet_sk(sk)->inet_num != IPPROTO_ICMPV6)
@@ -1746,12 +1750,15 @@ int ip6_mroute_setsockopt(struct sock *sk, int
optname, sockptr_t optval,
                         return -EINVAL;
                 if (copy_from_sockptr(&v, optval, sizeof(v)))
                         return -EFAULT;
+
+               do_wrmifwhole = (v == MRT6MSG_WRMIFWHOLE);
                 v = !!v;
                 rtnl_lock();
                 ret = 0;
                 if (v != mrt->mroute_do_pim) {
                         mrt->mroute_do_pim = v;
                         mrt->mroute_do_assert = v;
+                       mrt->mroute_do_wrvifwhole = do_wrmifwhole;
                 }
                 rtnl_unlock();
                 return ret;
@@ -2127,6 +2134,9 @@ static void ip6_mr_forward(struct net *net,
struct mr_table *mrt,
                                MFC_ASSERT_THRESH)) {
                         c->_c.mfc_un.res.last_assert = jiffies;
                         ip6mr_cache_report(mrt, skb, true_vifi,
MRT6MSG_WRONGMIF);
+                       if (mrt->mroute_do_wrvifwhole)
+                               ip6mr_cache_report(mrt, skb, true_vifi,
+                                                  MRT6MSG_WRMIFWHOLE);
                 }
                 goto dont_forward;
         }
--
2.7.4

Powered by blists - more mailing lists