Project

General

Profile

Feature #3563

osmo-gsm-tester: add test case: test a5/3

Added by pespin 3 months ago. Updated 2 months ago.

Status:
Resolved
Priority:
Normal
Assignee:
Target version:
-
Start date:
09/18/2018
Due date:
% Done:

100%

Spec Reference:

Description

Like we do in register_a5_1_authreq.py with A5/1.


Related issues

Related to Cellular Network Infrastructure - Bug #3043: A5/3 encryption failsResolved2018-03-08

History

#1 Updated by pespin 3 months ago

  • Status changed from New to Feedback

Submitted here:
https://gerrit.osmocom.org/#/c/osmo-gsm-tester/+/11009 encryption: Reuse code using test lib features
https://gerrit.osmocom.org/#/c/osmo-gsm-tester/+/11010 suites: encryption: Add test register_a5_3_authreq

Authentication is fine but currently the test fails due to known issue with afaik malformed Classmark Request missing some extra fields snet from MSC->BSS to determine whether MS supports a5/3.

20180918131037033 DMM <0002> gsm_04_08.c:1640 IMSI:901700000009031: to determine whether A5/3 is supported, first ask for a Classmark Update to obtain Classmark 2

#2 Updated by pespin 3 months ago

  • Related to Bug #3043: A5/3 encryption fails added

#4 Updated by laforge 3 months ago

the point of this test is to run a configuration that permits only A5/3 and which
should succeed, given that the modems we use are all new enough to support A5/3.

#5 Updated by pespin 3 months ago

Yes, the tests under "encryption" suite test that only 1 specific a5/X algo is used. In the case of the test I added, cfg file in osmo-msc and osmo-bsc ends up having following line:

encryption a53

And in the case of osmo-msc, "authentication required" is set.

#6 Updated by neels 3 months ago

On Wed, Sep 19, 2018 at 10:45:32AM +0000, pespin [REDMINE] wrote:

> encryption a53
> 

you mean "encryption a5 3" right :)

And in the case of osmo-msc, "authentication required" is set.

BTW, that's not needed; as soon as you request encryption, authentication is
implied. That's why the other setting is called "authentication optional"
instead of "disabled" -- it is still going to happen for encryption.

#7 Updated by pespin 3 months ago

neels wrote:

On Wed, Sep 19, 2018 at 10:45:32AM +0000, pespin [REDMINE] wrote:

[...]

you mean "encryption a5 3" right :)

Yes, sorry.

And in the case of osmo-msc, "authentication required" is set.

BTW, that's not needed; as soon as you request encryption, authentication is
implied. That's why the other setting is called "authentication optional"
instead of "disabled" -- it is still going to happen for encryption.

In any case it doesn't hurt and it's clear this way anyway. Thanks for sharing the info.

#8 Updated by pespin 2 months ago

  • Status changed from Feedback to Resolved
  • % Done changed from 0 to 100

Merged, closing

Also available in: Atom PDF

Add picture from clipboard (Maximum size: 48.8 MB)