)]}'
{
  "commit": "dafee490858f79e144c5e6cdd84ceb9efa20a3f1",
  "tree": "b14c78f41845026fd2ab161a7348d898a2fa6d3e",
  "parents": [
    "32c524d1c48b62be49fa1b1dd93fed10792debc0"
  ],
  "author": {
    "name": "Wei Dong",
    "email": "weid@nanjing-fnst.com",
    "time": "Wed Aug 02 13:41:21 2006 -0700"
  },
  "committer": {
    "name": "David S. Miller",
    "email": "davem@davemloft.net",
    "time": "Wed Aug 02 13:41:21 2006 -0700"
  },
  "message": "[IPV6]: SNMPv2 \"ipv6IfStatsOutFragCreates\" counter error\n\n  When I tested linux kernel 2.6.71.7 about statistics\n\"ipv6IfStatsOutFragCreates\", and found that it couldn\u0027t increase\ncorrectly. The criteria is RFC 2465:\n\n  ipv6IfStatsOutFragCreates OBJECT-TYPE\n      SYNTAX      Counter32\n      MAX-ACCESS  read-only\n      STATUS      current\n      DESCRIPTION\n         \"The number of output datagram fragments that have\n         been generated as a result of fragmentation at\n         this output interface.\"\n      ::\u003d { ipv6IfStatsEntry 15 }\n\nI think there are two issues in Linux kernel. \n1st:\nRFC2465 specifies the counter is \"The number of output datagram\nfragments...\". I think increasing this counter after output a fragment\nsuccessfully is better. And it should not be increased even though a\nfragment is created but failed to output.\n\n2nd:\nIf we send a big ICMP/ICMPv6 echo request to a host, and receive\nICMP/ICMPv6 echo reply consisted of some fragments. As we know that in\nLinux kernel first fragmentation occurs in ICMP layer(maybe saying\ntransport layer is better), but this is not the \"real\"\nfragmentation,just do some \"pre-fragment\" -- allocate space for date,\nand form a frag_list, etc. The \"real\" fragmentation happens in IP layer\n-- set offset and MF flag and so on. So I think in \"fast path\" for\nip_fragment/ip6_fragment, if we send a fragment which \"pre-fragment\" by\nupper layer we should also increase \"ipv6IfStatsOutFragCreates\".\n\nSigned-off-by: Wei Dong \u003cweid@nanjing-fnst.com\u003e\nAcked-by: YOSHIFUJI Hideaki \u003cyoshfuji@linux-ipv6.org\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "7c9f9a6421b8f3acdf7ccff1a106402913c75dd9",
      "old_mode": 33188,
      "old_path": "net/ipv4/ip_output.c",
      "new_id": "9bf307a2978355bd68fb66d663630b79720db85c",
      "new_mode": 33188,
      "new_path": "net/ipv4/ip_output.c"
    },
    {
      "type": "modify",
      "old_id": "70c9234b70e7da6b36dd65099def4addec3224b9",
      "old_mode": 33188,
      "old_path": "net/ipv6/ip6_output.c",
      "new_id": "69451af6abe7dd2aaff636498eb1894e45f43fc7",
      "new_mode": 33188,
      "new_path": "net/ipv6/ip6_output.c"
    }
  ]
}