Currently this option just disables the default neighbor offset
synchronization method. It is useful for testing.

Signed-off-by: Marco Porsch <[email protected]>
---
 mesh.c |   24 +++++++++++++++++++++++-
 1 files changed, 23 insertions(+), 1 deletions(-)

diff --git a/mesh.c b/mesh.c
index 12147a2..4a3c29e 100644
--- a/mesh.c
+++ b/mesh.c
@@ -352,6 +352,7 @@ COMMAND(get, mesh_param, "[<param>]",
 static int join_mesh(struct nl80211_state *state, struct nl_cb *cb,
                     struct nl_msg *msg, int argc, char **argv)
 {
+       struct nlattr *container;
        float rate;
        char *end;
 
@@ -375,13 +376,34 @@ static int join_mesh(struct nl80211_state *state, struct 
nl_cb *cb,
                argc--;
        }
 
+       container = nla_nest_start(msg, NL80211_ATTR_MESH_SETUP);
+       if (!container)
+               return -ENOBUFS;
+
+       if (argc > 1 && strcmp(argv[0], "vendor_sync") == 0) {
+               argv++;
+               argc--;
+               if (strcmp(argv[0], "on") == 0)
+                       NLA_PUT_U8(msg,
+                                  NL80211_MESH_SETUP_ENABLE_VENDOR_SYNC, 1);
+               else
+                       NLA_PUT_U8(msg,
+                                  NL80211_MESH_SETUP_ENABLE_VENDOR_SYNC, 0);
+               argv++;
+               argc--;
+       }
+       /* parse and put other NL80211_ATTR_MESH_SETUP elements here */
+
+       nla_nest_end(msg, container);
+
        if (!argc)
                return 0;
        return set_interface_meshparam(state, cb, msg, argc, argv);
  nla_put_failure:
        return -ENOBUFS;
 }
-COMMAND(mesh, join, "<mesh ID> [mcast-rate <rate in Mbps>] [<param>=<value>]*",
+COMMAND(mesh, join, "<mesh ID> [mcast-rate <rate in Mbps>] [vendor_sync 
on|off]"
+       " [<param>=<value>]*",
        NL80211_CMD_JOIN_MESH, 0, CIB_NETDEV, join_mesh,
        "Join a mesh with the given mesh ID with mcast-rate and mesh 
parameters.");
 
-- 
1.7.5.4

_______________________________________________
Devel mailing list
[email protected]
http://lists.open80211s.org/cgi-bin/mailman/listinfo/devel

Reply via email to