udp: Fix the SNMP counter of UDP_MIB_INDATAGRAMS
authorWei Yongjun <yjwei@cn.fujitsu.com>
Sun, 2 Nov 2008 16:11:01 +0000 (16:11 +0000)
committerDavid S. Miller <davem@davemloft.net>
Mon, 3 Nov 2008 07:52:45 +0000 (23:52 -0800)
If UDP echo is sent to xinetd/echo-dgram, the UDP reply will be received
at the sender. But the SNMP counter of UDP_MIB_INDATAGRAMS will be not
increased, UDP6_MIB_INDATAGRAMS will be increased instead.

  Endpoint A                      Endpoint B
  UDP Echo request ----------->
  (IPv4, Dst port=7)
                   <----------    UDP Echo Reply
                                  (IPv4, Src port=7)

This bug is come from this patch cb75994ec311b2cd50e5205efdcc0696abd6675d.

It do counter UDP[6]_MIB_INDATAGRAMS until udp[v6]_recvmsg. Because
xinetd used IPv6 socket to receive UDP messages, thus, when received
UDP packet, the UDP6_MIB_INDATAGRAMS will be increased in function
udpv6_recvmsg() even if the packet is a IPv4 UDP packet.

This patch fixed the problem.

Signed-off-by: Wei Yongjun <yjwei@cn.fujitsu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/ipv6/udp.c

index 71e259e866a16c93cad70fb65a9bab6313da4b93..18696af106d66c2876ac4c718ab38937a1172ef8 100644 (file)
@@ -138,6 +138,7 @@ int udpv6_recvmsg(struct kiocb *iocb, struct sock *sk,
        int peeked;
        int err;
        int is_udplite = IS_UDPLITE(sk);
+       int is_udp4;
 
        if (addr_len)
                *addr_len=sizeof(struct sockaddr_in6);
@@ -158,6 +159,8 @@ try_again:
        else if (copied < ulen)
                msg->msg_flags |= MSG_TRUNC;
 
+       is_udp4 = (skb->protocol == htons(ETH_P_IP));
+
        /*
         * If checksum is needed at all, try to do it while copying the
         * data.  If the data is truncated, or if we only want a partial
@@ -180,9 +183,14 @@ try_again:
        if (err)
                goto out_free;
 
-       if (!peeked)
-               UDP6_INC_STATS_USER(sock_net(sk),
-                               UDP_MIB_INDATAGRAMS, is_udplite);
+       if (!peeked) {
+               if (is_udp4)
+                       UDP_INC_STATS_USER(sock_net(sk),
+                                       UDP_MIB_INDATAGRAMS, is_udplite);
+               else
+                       UDP6_INC_STATS_USER(sock_net(sk),
+                                       UDP_MIB_INDATAGRAMS, is_udplite);
+       }
 
        sock_recv_timestamp(msg, sk, skb);
 
@@ -196,7 +204,7 @@ try_again:
                sin6->sin6_flowinfo = 0;
                sin6->sin6_scope_id = 0;
 
-               if (skb->protocol == htons(ETH_P_IP))
+               if (is_udp4)
                        ipv6_addr_set(&sin6->sin6_addr, 0, 0,
                                      htonl(0xffff), ip_hdr(skb)->saddr);
                else {
@@ -207,7 +215,7 @@ try_again:
                }
 
        }
-       if (skb->protocol == htons(ETH_P_IP)) {
+       if (is_udp4) {
                if (inet->cmsg_flags)
                        ip_cmsg_recv(msg, skb);
        } else {