Project

General

Profile

Bug #3914

PAP PCO not handled correctly

Added by laforge about 1 month ago. Updated about 1 month ago.

Status:
In Progress
Priority:
Normal
Assignee:
Category:
openggsn
Target version:
-
Start date:
04/10/2019
Due date:
% Done:

90%

Spec Reference:

Description

I don't have a GTP trace, but the attached Gb interface trace shows the following oddities:

  • The PCO as sent by the MS/UE in the PDP CTX ACT REQ contains a PAP Auth Req
    • This Auth Req is malformed according to the wireshark dissector (can we trust it?), but that's not the main point here
  • The PCO as returned by the network to te MS doesn't contain any PAP Auth Resp
    • not sure if support for PAP is required, but TS 24.008 at least states:
      • At least the following protocol identifiers (as defined in RFC 3232 [103]) shall be
        supported in this version of the protocol:C021H (LCP); C023H (PAP); C223H (CHAP);and 8021H (IPCP).
    • Instead, our PCO response contains twice the PCO for IPCP (DNS servers), which clearly is wrong
act_pdp_req_with_pap.pcapng act_pdp_req_with_pap.pcapng 792 Bytes laforge, 04/10/2019 05:53 AM

History

#1 Updated by laforge about 1 month ago

Note that the PAP inside the capture is clarly invalid as per specification. It sends a Peer-Id-Length of 04, but there's 6 (excluding NUL byte) or 7 (including NUL byte) characters of Peer Identifier.

Nevertheless, OsmoGGSN should handle this somewhat intelligently, e.g. by sending an ACK in return.

Under no circumstances should OsmoGGSN send duplicate DNS PCOs

#2 Updated by laforge about 1 month ago

  • Status changed from New to In Progress
  • % Done changed from 0 to 30

Note that the PAP inside the capture is clarly invalid as per specification. It sends a Peer-Id-Length of 04, but there's 6 (excluding NUL byte) or 7 (including NUL byte) characters of Peer Identifier.

Nevertheless, OsmoGGSN should handle this somewhat intelligently, e.g. by sending an ACK in return.

Under no circumstances should OsmoGGSN send duplicate DNS PCOs

See https://gerrit.osmocom.org/#/c/osmo-ttcn3-hacks/+/13563 for improving our TTCN3 tests to avoid duplicate PCO protocolIDs and https://gerrit.osmocom.org/#/c/osmo-ttcn3-hacks/+/13564 for a test case reproducing exactly the PCOs as observed by that phone.

#3 Updated by laforge about 1 month ago

  • % Done changed from 30 to 60

I just pushed a series of patches, and https://gerrit.osmocom.org/#/c/osmo-ggsn/+/13570 is removing the duplicate PCO response from osmo-ggsn.

we still don't have any PAP handling.

#4 Updated by laforge about 1 month ago

See https://gerrit.osmocom.org/#/c/osmo-ggsn/+/13608 for the patch adding minimalistic PAP support. Hopefully this will make some modems happy.

#5 Updated by laforge about 1 month ago

  • % Done changed from 60 to 90

Also available in: Atom PDF

Add picture from clipboard (Maximum size: 48.8 MB)