Don't let the existence of a protocol tree affect what's in the Info column.

A dissector must never assume that it will, or won't, be called with a
protocol tree; it's up to the Wireshark/TShark/etc. core to decide
whether to do it, and it can change its behavior over time or even
change it from release to release.

Have dissect_epath() take an argument that explicitly indicates whether
to add the CIP class to the Info column, rather than assuming that you
do so only if the tree pointer passed to it is null.

Bug: 12257
Change-Id: Ide8a6fc21252880f849a8d0aa4659a675bb3ae04
Reviewed-on: https://code.wireshark.org/review/14456
Reviewed-by: Guy Harris <guy@alum.mit.edu>

1 files changed, 2 insertions, 2 deletions

diff --git a/epan/dissectors/packet-cipsafety.c b/epan/dissectors/packet-cipsafety.c
index 11e67a4c3e..44a2466d4a 100644
--- a/epan/dissectors/packet-cipsafety.c
+++ b/epan/dissectors/packet-cipsafety.c
@@ -945,7 +945,7 @@ static int dissect_s_supervisor_output_connection_point_owners(packet_info *pinf
 
          epath_tree = proto_tree_add_subtree(entry_tree,
                          tvb, offset+attr_len, app_path_size, ett_path, &app_path_item, "Application Resource: ");
-         dissect_epath( tvb, pinfo, epath_tree, app_path_item, offset+attr_len, app_path_size, FALSE, TRUE, NULL, NULL, NO_DISPLAY, NULL);
+         dissect_epath(tvb, pinfo, epath_tree, app_path_item, offset+attr_len, app_path_size, FALSE, TRUE, NULL, NULL, NO_DISPLAY, NULL, TRUE);
          attr_len += app_path_size;
       }
    }
@@ -1097,7 +1097,7 @@ static int dissect_s_validator_app_data_path(packet_info *pinfo, proto_tree *tre
 {
    proto_item* pi;
    proto_tree* epath_tree = proto_tree_add_subtree(tree, NULL, 0, 0, ett_path, &pi, "Application Data Path: ");
-   dissect_epath(tvb, pinfo, epath_tree, pi, offset, total_len, FALSE, FALSE, NULL, NULL, NO_DISPLAY, NULL);
+   dissect_epath(tvb, pinfo, epath_tree, pi, offset, total_len, FALSE, FALSE, NULL, NULL, NO_DISPLAY, NULL, TRUE);
    return total_len;
 }