[bugzilla-daemon@xxxxxxxxxxxxx]

Comment # 5 on bug 13110 from Uli Heilmeier

IMHO issue here is that frame 724 has two DICOM segments. The second one is
fragmented.
The first pass calls dissect_dcm_pdv_body() for the first segment (which is
correct). The second pass however jumps into the reassemble_data part falsely
for the first DICOM segment.

Therefore I guess the bug is here:

--- epan/dissectors/packet-dcm.c ---
6684         head = fragment_add_seq_next(&dcm_pdv_reassembly_table,¬
6685                                 tvb, offset, pinfo, reassembly_id, NULL,¬
6686                                 pdv_body_len,¬
6687                                 !(pdv->is_last_fragment));¬
6688 ¬
6689         if (head && (head->next == NULL)) {¬
6690             /* Was not really fragmented, therefore use 'conventional'
decoding¬
6691                fragment_add_seq_next() won't add any items to the list,
when last fragment only¬
6692             */¬
6693 ¬
6694             offset = dissect_dcm_pdv_body(tvb, pinfo, tree, pdv, offset,
pdv_body_len, pdv_description);¬
6695         }¬
6696         else {¬
6697             next_tvb = process_reassembled_data(tvb, offset, pinfo,¬
6698                                         "Reassembled PDV", head,¬
---

---

      You are receiving this mail because:

• You are watching all bug changes.