Wireshark-dev: Re: [Wireshark-dev] Display IPv6 extension headers outside of the IPv6 subtree

From: João Valverde <joao.valverde@xxxxxxxxxxxxxxxxxx>
Date: Sat, 22 Aug 2015 14:58:12 +0100


On 08/22/2015 02:28 PM, Jo�o Valverde wrote:


On 08/22/2015 09:29 AM, Alexis La Goutte wrote:


On Sat, Aug 22, 2015 at 8:10 AM, Jo�o Valverde
<joao.valverde@xxxxxxxxxxxxxxxxxx
<mailto:joao.valverde@xxxxxxxxxxxxxxxxxx>> wrote:

    I'm not so sure now that I've done some tests with short and long
    header chains, I think I'll drop it.

Can you push your proto ? or add some screenshot to see the different ?


For now all I did was:

diff --git a/epan/dissectors/packet-ipv6.c b/epan/dissectors/packet-ipv6.c
index 1c07776..76e5285 100644
--- a/epan/dissectors/packet-ipv6.c
+++ b/epan/dissectors/packet-ipv6.c
@@ -2225,7 +2225,7 @@ dissect_ipv6(tvbuff_t *tvb, packet_info *pinfo,
proto_tree *tree)

      if (nxt == IP_PROTO_HOPOPTS) {
          options_tvb = tvb_new_subset_remaining(tvb, offset);
-        advance = dissect_hopopts(options_tvb, pinfo, ipv6_tree, &iph);
+        advance = dissect_hopopts(options_tvb, pinfo, tree, &iph);
          if (advance > 0) {
              nxt = tvb_get_guint8(tvb, offset);
              offset += advance;
@@ -2274,7 +2274,7 @@ again:
      nxt_handle =
dissector_get_uint_handle(ipv6_next_header_dissector_table, nxt);

      if ((nxt_handle) &&
-        ((advance = call_dissector_with_data(nxt_handle, options_tvb,
pinfo, ipv6_tree, &iph)) > 0)) {
+        ((advance = call_dissector_with_data(nxt_handle, options_tvb,
pinfo, tree, &iph)) > 0)) {
          nxt = tvb_get_guint8(tvb, offset);
          offset += advance;
          plen -= advance;


I had meant to change the ipv6.nxt dissectors to ip.proto but didn't get
that far.


For most practical cases I still feel it would be a net benefit visually and functionally, although you can get into some pretty pathological cases with ridiculously long extension header chains and IP over IP.

The one header that gets to me the most being hidden is AH, although it is consistent with the current policy (MobileIPv6 is not IMO).




    Regards,

    Jo�o V.


    On 08/21/2015 06:04 PM, Jo�o Valverde wrote:

        I forgot to include:

        f) it would fix this[1] bug too.

        [1]https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=9996

        On 08/21/2015 05:45 PM, Jo�o Valverde wrote:

            Hi,

            I would like to propose a patch to remove IPv6 extension
            headers from
            the IPv6 subtree and display them as normal IP protocols in
            the UI.

            Arguments for:

            a) The current behaviour really, really, (really) hurts
            legibility
            b) It would be consistent among all extension headers (see
            for example
            MobileIPv6 vs AH - why the difference?)[1]
            c) Extension headers aren't special in this regard, the IP
            protocol
            space is shared
            d) IPv6 extension headers over IPv4 would be dissected, as
            they should.
            An expert info error can then be displayed where appropriate.
            e) It would do away with the hack (?) used for the AH
            dissector (I
            think, I haven't looked very closely yet here).

            Arguments against:
            ?)

            Would anyone object to this proposal?

            Thanks!

            Jo�o V.

            [1]https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=10705


___________________________________________________________________________

    Sent via:    Wireshark-dev mailing list <wireshark-dev@xxxxxxxxxxxxx
    <mailto:wireshark-dev@xxxxxxxxxxxxx>>
    Archives: https://www.wireshark.org/lists/wireshark-dev
    Unsubscribe: https://wireshark.org/mailman/options/wireshark-dev
                 mailto:wireshark-dev-request@xxxxxxxxxxxxx
    <mailto:wireshark-dev-request@xxxxxxxxxxxxx>?subject=unsubscribe




___________________________________________________________________________

Sent via:    Wireshark-dev mailing list <wireshark-dev@xxxxxxxxxxxxx>
Archives:    https://www.wireshark.org/lists/wireshark-dev
Unsubscribe: https://wireshark.org/mailman/options/wireshark-dev

mailto:wireshark-dev-request@xxxxxxxxxxxxx?subject=unsubscribe