It appears that the annoying Linux behavior wherein the "source routed"

flag is never set on packets sent by the machine on which the capture is
being done is not a 2.0[.x]-only problem; it's in 2.4.18, and may well
be in all 2.x kernels with at least some drivers.

git-svn-id: http://anonsvn.wireshark.org/wireshark/trunk@5812 f5534014-38df-0310-8fa8-9805f1628bb7

diff --git a/packet-tr.c b/packet-tr.c
index 8e2f39fdc2702e800f57e8026ec92ec82fd2242e..b5f2d40f09ebc910510b4212fb767061994f655a 100644 (file)
--- a/packet-tr.c
+++ b/packet-tr.c
@@ -2,7 +2,7 @@
  * Routines for Token-Ring packet disassembly
  * Gilbert Ramirez <gram@alumni.rice.edu>
  *
- * $Id: packet-tr.c,v 1.70 2002/04/14 23:04:04 guy Exp $
+ * $Id: packet-tr.c,v 1.71 2002/07/02 08:23:53 guy Exp $
  *
  * Ethereal - Network traffic analyzer
  * By Gerald Combs <gerald@ethereal.com>
@@ -226,6 +226,11 @@ capture_tr(const u_char *pd, int offset, int len, packet_counts *ld) {
         * which packets should have been SR here. I'll check to
         * see if there's a SNAP or IPX field right after
         * my RIF fields.
+        *
+        * The Linux 2.4.18 code, at least appears to do the
+        * same thing, from a capture I got from somebody running
+        * 2.4.18 (RH 7.1, so perhaps this is a Red Hat
+        * "improvement").
         */
        if (!source_routed && trn_rif_bytes > 0) {
                if (pd[offset + 0x0e] != pd[offset + 0x0f]) {
@@ -365,6 +370,11 @@ dissect_tr(tvbuff_t *tvb, packet_info *pinfo, proto_tree *tree)
         * which packets should have been SR here. I'll check to
         * see if there's a SNAP or IPX field right after
         * my RIF fields.
+        *
+        * The Linux 2.4.18 code, at least appears to do the
+        * same thing, from a capture I got from somebody running
+        * 2.4.18 (RH 7.1, so perhaps this is a Red Hat
+        * "improvement").
         */
        if (frame_type == 1 && !source_routed && trn_rif_bytes > 0) {
                TRY {