]> bbs.cooldavid.org Git - net-next-2.6.git/commitdiff
[IPV6]: SNMPv2 "ipv6IfStatsInHdrErrors" counter error
authorWei Dong <weid@nanjing-fnst.com>
Wed, 2 Aug 2006 20:39:57 +0000 (13:39 -0700)
committerDavid S. Miller <davem@davemloft.net>
Wed, 2 Aug 2006 20:39:57 +0000 (13:39 -0700)
  When I tested Linux kernel 2.6.17.7 about statistics
"ipv6IfStatsInHdrErrors", found that this counter couldn't increase
correctly. The criteria is RFC2465:
  ipv6IfStatsInHdrErrors OBJECT-TYPE
      SYNTAX     Counter3
      MAX-ACCESS read-only
      STATUS     current
      DESCRIPTION
         "The number of input datagrams discarded due to
         errors in their IPv6 headers, including version
         number mismatch, other format errors, hop count
         exceeded, errors discovered in processing their
         IPv6 options, etc."
      ::= { ipv6IfStatsEntry 2 }

When I send TTL=0 and TTL=1 a packet to a router which need to be
forwarded, router just sends an ICMPv6 message to tell the sender that
TIME_EXCEED and HOPLIMITS, but no increments for this counter(in the
function ip6_forward).

Signed-off-by: Wei Dong <weid@nanjing-fnst.com>
Acked-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/ipv6/ip6_output.c

index 5e74a37695f7eea53a95b4bbffd5ec16734bf836..70c9234b70e7da6b36dd65099def4addec3224b9 100644 (file)
@@ -356,6 +356,7 @@ int ip6_forward(struct sk_buff *skb)
                skb->dev = dst->dev;
                icmpv6_send(skb, ICMPV6_TIME_EXCEED, ICMPV6_EXC_HOPLIMIT,
                            0, skb->dev);
+               IP6_INC_STATS_BH(IPSTATS_MIB_INHDRERRORS);
 
                kfree_skb(skb);
                return -ETIMEDOUT;