ims_rdn: new experimental module for IMS RDNs by luca-nardin-kontron · Pull Request #4786 · kamailio/kamailio

luca-nardin-kontron · 2026-06-17T16:29:11Z

Pre-Submission Checklist

Commit message has the format required by CONTRIBUTING guide
Commits are split per component (core, individual modules, libs, utils, ...)
Each component has a single commit (if not, squash them into one commit)
Code is formatted with clang-format using the config file .clang-format
from source code folder
No commits to README files for modules (changes must be done to docbook files
in doc/ subfolder, the README file is autogenerated)

Type Of Change

Small bug fix (non-breaking change which fixes an issue)
New feature (non-breaking change which adds new functionality)
Breaking change (fix or feature that would change existing functionality)

Checklist:

PR should be backported to stable branches
Tested changes locally
Related to issue #XXXX (replace XXXX with an open issue number)

Description

We (Kontron Transportation) are issuing this pull request
with a new experimental module, which is called ims_rdn,
hopeful that this module might be useful for at least
some users of Kamailio IMS (I-CSCF and S-CSCF).

This module is most useful for terminating unregistered services
that are triggered by the I/S-CSCF and that are identified
by hierarchical address ranges of E.164 addresses and of
EIRENE addresses (in the Request-URI), not unlike wildcarded PSIs.
By configuring these address ranges at the I/S-CSCF,
one can save one SAR/SAA exchange with the HSS per call.
EIRENE addresses are railway specific telephone numbers,
which have been defined by the UIC.
This explains the name of the module RDN = Railway Dedicated Networks.

Additionally, this module defines kamailio transformations,
which are tailored for the use in RDNs,
e.g. for the transformation of functional numbers and
of MCData SDS into instant messages and vice versa.
This module is even more useful, when implementing
an IWF according to ETSI TS 103 792,
in which case some specific routing scripts will be needed.
We will be glad to help with those.

christoph-v-kontron · 2026-06-21T15:43:44Z

Hi Luca,

Two comments from my side.

I think the README must be updated. As far as I remember, we have removed the PVs from the module and now access the AVPs directly
Maybe, if Daniel and the others agree, we should bring our transformations to the transformations cookbook

Thank you for doing this important PR

henningw · 2026-06-23T07:49:40Z

Thanks for the PR. I only had a brief look to the code so far. Several things should be addressed:

please format the code with clang-format, the specific check is failing, then do a force-push to this PR
a cmake file needs to be added as well, as we switched to this build system some time ago
you are using system malloc in several places, this should be replaced by the Kamailio pgk_malloc/free variants unless its especially needed (e.g. for interaction with external libraries)
several global structs (e.g. hash_entry, config_type, template_type) should be prefixed, e.g. rdn_hash_struct or similar
Small utility functions in ims_rdn.c (e.g. key_exists) should be made static
Review also the utility functions in rdn_trans.{h,c}, many of them should be probably also made static
There are some German log/debug and also comments messages that should be translated to English

Regarding the note from @christoph-v-kontron, there seems to be pseudo-variables exported, which is fine and actually the preferred way since some time. But please also review of course.

henningw · 2026-07-14T08:48:39Z

@luca-nardin-kontron Any comments from your side to the earlier remarks?

luca-nardin-kontron · 2026-07-14T08:58:30Z

@luca-nardin-kontron Any comments from your side to the earlier remarks?

Hi Henning,

regarding clang, seems we used a more recent version of the tool that generates a format that is not compatible with the version you used. Regarding the other topics, we did already some correction and plan to update our pull request soon.

Kind Regards,
Luca

henningw · 2026-08-04T07:33:29Z

@luca-nardin-kontron Thanks for the previous update, are there any news?

luca-nardin-kontron · 2026-08-04T16:46:00Z

@luca-nardin-kontron Thanks for the previous update, are there any news?

Hi Henning,

all the requested changes are done and recompiled successfully against the community master, I just need to re-create the README, so I guess tomorrow or on friday I can update the pull request (if our machines survive the record heatwave of Vienna!)

Kind regards,
Luca

Added update_contacts(..) and parse_user_data(..) in the module's API.

We (Kontron Transportation) are issuing this pull request with a new experimental module, which is called ims_rdn, hopeful that this module might be useful for at least some users of Kamailio IMS (I-CSCF and S-CSCF). This module is most useful for terminating unregistered services that are triggered by the I/S-CSCF and that are identified by hierarchical address ranges of E.164 addresses and of EIRENE addresses (in the Request-URI), not unlike wildcarded PSIs. By configuring these address ranges at the I/S-CSCF, one can save one SAR/SAA exchange with the HSS per call. EIRENE addresses are railway specific telephone numbers, which have been defined by the UIC. This explains the name of the module RDN = Railway Dedicated Networks. Additionally, this module defines kamailio transformations, which are tailored for the use in RDNs, e.g. for the transformation of functional numbers and of MCData SDS into instant messages and vice versa. This module is even more useful, when implementing an IWF according to ETSI TS 103 792, in which case some specific routing scripts will be needed. We will be glad to help with those.

luca-nardin-kontron · 2026-08-07T21:41:53Z

We had to include in this PR small additions to ims_registrar_scscf as well, otherwise ims_rdn would not compile against community master. Please let us know if you prefer a separate PR for that change.

luca-nardin-kontron · 2026-08-27T08:42:51Z

@henningw it seems the failed check is due to a bug in github? Should we do something on our side to solve the issue or just wait that it gets fixed?

henningw · 2026-08-27T10:39:04Z

@henningw it seems the failed check is due to a bug in github? Should we do something on our side to solve the issue or just wait that it gets fixed?

You can ignore this check, its a known issue. Regarding a second review, I planned it for the next days. Some delay during the holiday period.

henningw

Thanks for the update, I did another review. Mostly standard topics I added as comments:

missing logging and error handling for internal and external function calls
inconsistent logging, also inconsistent log messages, redundant prefixes etc.
dead and commented out code that should be removed.

Regarding the naming - the functions have somehow generic names (e.g. analyze_..._ruri). But this module is quite special so its probably not an issue for eventual overlapping to other modules.

henningw · 2026-09-02T13:53:55Z

+
+	e = (struct rdn_hash_entry *)shm_malloc(sizeof(struct rdn_hash_entry));
+	if(!e) {
+		ERR("Not enough memory left\n");


Use LM_ERR as elsewhere, or SHM_MEM_ERROR

henningw · 2026-09-02T13:54:06Z

+			sizeof(struct rdn_hash_entry *) * RDN_HASH_SIZE);
+
+	if(!dpsi_table) {
+		ERR("No memory left\n");


Use LM_ERR as elsewhere, or SHM_MEM_ERROR

henningw · 2026-09-02T13:55:10Z

+	int err = 0;
+	rdn_special_handling_t *new_sh = 0;
+	if(temp) {
+		new_sh = (rdn_special_handling_t *)shm_malloc(


Error handling and logging

henningw · 2026-09-02T13:56:00Z

+					strncpy(new_sh->iwf_domain.s, temp->iwf_domain.s,
+							temp->iwf_domain.len);
+				} else
+					err = 1;


henningw · 2026-09-02T13:56:08Z

+					err = 1;
+			}
+			if(temp->hardcoded_ruri.s && temp->hardcoded_ruri.len) {
+				new_sh->hardcoded_ruri.s = (char *)shm_malloc(


Error handling and logging

henningw · 2026-09-02T14:40:43Z

+	}
+	// Print timevalue as HEX number.
+	// The string is ASCII encoded (not binary) because "append_message_body_hex()" transforms it into binary bytes anyway.
+	return sprintf(dst, "%010llx", timevalue);


Error handling and logging

henningw · 2026-09-02T14:41:50Z

+
+	switch(subtype) {
+		case TR_RDN_UUI2FN:
+			length = base16_decode_update(_tr_buffer, &val->rs);


Error handling and logging for this and following function calls

henningw · 2026-09-02T14:43:47Z

+		case TR_RDN_BIN2INT:
+			// read any character from binary message body
+			// and cast it as integer.
+			val->ri = (unsigned char)val->rs.s[0];


Error handling and logging..

henningw · 2026-09-02T14:43:53Z

+			val->ri = get_sds_disposition_type(val->rs.len, val->rs.s);
+			// Check for Error
+			if(!val->ri)
+				return -1;


Logging here and the following function calls

henningw · 2026-09-02T14:44:31Z

+			break;
+		case TR_RDN_TXT_2_MSG_BODY:
+			// build SDS header for SDS DATA-PAYLOAD
+			length = build_sds_header_for_txt_body(


Error handling and logging..

henningw · 2026-09-22T17:42:54Z

@luca-nardin-kontron Any update from your side? Were you able to have a look to the comments?

luca-nardin-kontron added 2 commits August 7, 2026 22:45

ims_registrar_scscf: expose update_contacts and parse_user_data

9a8819f

Added update_contacts(..) and parse_user_data(..) in the module's API.

luca-nardin-kontron force-pushed the ims_rdn_pull_request branch from f5ed968 to 78d84b1 Compare August 7, 2026 21:30

henningw reviewed Sep 2, 2026

View reviewed changes

Conversation

luca-nardin-kontron commented Jun 17, 2026

Pre-Submission Checklist

Type Of Change

Checklist:

Description

Uh oh!

christoph-v-kontron commented Jun 21, 2026

Uh oh!

henningw commented Jun 23, 2026

Uh oh!

henningw commented Jul 14, 2026

Uh oh!

luca-nardin-kontron commented Jul 14, 2026

Uh oh!

henningw commented Aug 4, 2026

Uh oh!

luca-nardin-kontron commented Aug 4, 2026

Uh oh!

luca-nardin-kontron commented Aug 7, 2026

Uh oh!

luca-nardin-kontron commented Aug 27, 2026

Uh oh!

henningw commented Aug 27, 2026

Uh oh!

henningw left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

henningw commented Sep 22, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants