sysmocom
/
open5gs
forked from acouzens/open5gs
11
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
open5gs/src/smf/n4-handler.c

1279 lines
46 KiB
C
Raw Normal View History

Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
/*
* Copyright (C) 2019 by Sukchan Lee <acetcom@gmail.com>
*
* This file is part of Open5GS.
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU Affero General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <https://www.gnu.org/licenses/>.
*/
#include "context.h"
#include "timer.h"
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
#include "s5c-build.h"
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
#include "pfcp-path.h"
#include "gtp-path.h"
#include "n4-handler.h"
feat: Add dedicated QoS flow
2021-01-01 02:07:08 +00:00
#include "binding.h"
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
#include "sbi-path.h"
feat: Add dedicated QoS flow
2021-01-01 02:07:08 +00:00
#include "ngap-path.h"
[SMF] Introduce optional Gy interface support (#1479) The use of the Gy interface (SMF acting as CTF towards an OCS node) is mandated through configuration file. Default value "enable: auto" will only make use of it in case a Diameter peer announcing support for the Credit-Control Application is found. Upon subscriber session creation, and after auth check over Gx, the SMF will create a Gy session with the OCS and only after that step the SMF will accept the session back to the subscriber. The OCS may then grant some traffic volumes/time and ask to be notified back with updated measurements. In order to get the measurements, the SMF relies on PFCP URR configured to the UPF through Session Repoort Request messages. When closing the subscriber session, the SMF will also terminate the Gy session at the OCS. So far only some specifics parts of the Gy interface as well as the PFCP side are implemented. Those should be enough to at least have volume/time thresholds granted by the OCS, which then will be able to track subsriber resource use. This patch doesn't implement the OCS side of the Gy interface, that's left as a future exercise. The interface was tested using an OCS emulator implemented in TTCN-3 [1] [1] https://cgit.osmocom.org/osmo-ttcn3-hacks/
2022-04-08 23:26:28 +00:00
#include "fd-path.h"
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
Pespin/pr (#1395) * [SMF] Gn: Avoid assert crash if no PDP resources available * [SMF] Gn: Rearrange IE handling order in CreatePDPContextRequest Let's handle the GTPC remote addr + TEID first, since those should be used in the CreatePDPContextResponse ideally if available. Let's then handle parsing of all IEs not related to bearers/UserPlane, then those missing, and finally do all the IP resource allocation.
2022-02-25 13:39:44 +00:00
uint8_t gtp_cause_from_pfcp(uint8_t pfcp_cause, uint8_t gtp_version)
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
{
Introduce Gn interface (GTPv1C) Support to PGW (#1351) * [CORE] tlv: Store mode in ogs_tlv_t This allows specifying the format of the IE for each individual IE, hence allowing messages containing IEs formatted in different ways. This is needed in order to support parsing GTPv1-C, since messages contain IEs with different structure (TLV vs TV). Hence, this is a preparation patch to add support for parsing TVs in ogs-tlv.c/.h. * [CORE] tlv: Support parsing msg with both TLV and TV in it IEs of type TV are sometimes used in GTPv1-C. Current tlv parser/builder doesn't provide with ways to parse messages which contain TV formatted IEs. This patch adds the relevant types and ways to encode/decode them. Furthermore, the current parser/builder allows parsing/building messages containing the exact same format in all its IEs. A new parser function is added which allows parsing messages of different types (TV, TLV) mixed in the same message. In order to be able to do so, it uses the general msg_mode passed to it in order to know the general TLV format (in essence, the length of the Tag field, and also the length of the Length field if applicable each IE). Looking up the instance in the TLV description is left undone and hadcoded to 0, since the only user so far requiring this API is GTPv1-C, which has no instances. * [CORE] tlv: Support repeated tag+instance parsing TLV message In GTPv2C, repeated IEs (same tag) are easily differentiated by the Instance byte, which provides info to match different decoded structures. In GTPv1C though, there's no Instance byte, and we still encounter repeated IEs (like GSN Address in Create PDP Context Request). Hence, the TLV decoder needs to be updated to track count of IEs found (identified by tag+instance, where instance is always 0 in GTPv1C) and get the proper description index + offset into the decoded structure. * [GTP]: Move GTPv2-C specifics to its own libgtp subdir This will allow adding GTPv1-C code by the side. Most GTPv2 code is left in this patch as "gtp" instead of renaming it to "gtp2" in order to avoid massive changes. It can be done at a later stage if wanted. * [GTP] Support generating GTPv1-C messages * [SMF] Add Gn interface support This patch introduces GTPv1C support to open5gs-smfd. With it, open5gs-becomes a GGSN too, where SGSN can connect to, hence supporting GERAN and UTRAN networks.
2022-02-18 13:23:45 +00:00
switch (gtp_version) {
case 1:
switch (pfcp_cause) {
case OGS_PFCP_CAUSE_REQUEST_ACCEPTED:
return OGS_GTP1_CAUSE_REQUEST_ACCEPTED;
case OGS_PFCP_CAUSE_REQUEST_REJECTED:
return OGS_GTP1_CAUSE_REJ_MS_NOT_GPRS_RESPONDING;
case OGS_PFCP_CAUSE_SESSION_CONTEXT_NOT_FOUND:
return OGS_GTP1_CAUSE_CONTEXT_NOT_FOUND;
case OGS_PFCP_CAUSE_MANDATORY_IE_MISSING:
case OGS_PFCP_CAUSE_CONDITIONAL_IE_MISSING:
return OGS_GTP1_CAUSE_MANDATORY_IE_MISSING;
case OGS_PFCP_CAUSE_INVALID_LENGTH:
return OGS_GTP1_CAUSE_INVALID_MESSAGE_FORMAT;
case OGS_PFCP_CAUSE_MANDATORY_IE_INCORRECT:
return OGS_GTP1_CAUSE_MANDATORY_IE_INCORRECT;
case OGS_PFCP_CAUSE_INVALID_FORWARDING_POLICY:
case OGS_PFCP_CAUSE_INVALID_F_TEID_ALLOCATION_OPTION:
return OGS_GTP1_CAUSE_INVALID_MESSAGE_FORMAT;
case OGS_PFCP_CAUSE_NO_ESTABLISHED_PFCP_ASSOCIATION:
return OGS_GTP1_CAUSE_NO_RESOURCES_AVAILABLE;
case OGS_PFCP_CAUSE_RULE_CREATION_MODIFICATION_FAILURE:
return OGS_GTP1_CAUSE_SEMANTIC_ERR_TFT_OPERATION;
case OGS_PFCP_CAUSE_PFCP_ENTITY_IN_CONGESTION:
return OGS_GTP1_CAUSE_APN_CONGESTION;
case OGS_PFCP_CAUSE_NO_RESOURCES_AVAILABLE:
return OGS_GTP1_CAUSE_NO_RESOURCES_AVAILABLE;
case OGS_PFCP_CAUSE_SERVICE_NOT_SUPPORTED:
return OGS_GTP1_CAUSE_SERVICE_NOT_SUPPORTED;
case OGS_PFCP_CAUSE_SYSTEM_FAILURE:
return OGS_GTP1_CAUSE_SYSTEM_FAILURE;
case OGS_PFCP_CAUSE_ALL_DYNAMIC_ADDRESS_ARE_OCCUPIED:
return OGS_GTP1_CAUSE_ALL_DYNAMIC_PDP_ADDRS_OCCUPIED;
default:
return OGS_GTP1_CAUSE_SYSTEM_FAILURE;
}
break;
case 2:
switch (pfcp_cause) {
case OGS_PFCP_CAUSE_REQUEST_ACCEPTED:
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
return OGS_GTP2_CAUSE_REQUEST_ACCEPTED;
Introduce Gn interface (GTPv1C) Support to PGW (#1351) * [CORE] tlv: Store mode in ogs_tlv_t This allows specifying the format of the IE for each individual IE, hence allowing messages containing IEs formatted in different ways. This is needed in order to support parsing GTPv1-C, since messages contain IEs with different structure (TLV vs TV). Hence, this is a preparation patch to add support for parsing TVs in ogs-tlv.c/.h. * [CORE] tlv: Support parsing msg with both TLV and TV in it IEs of type TV are sometimes used in GTPv1-C. Current tlv parser/builder doesn't provide with ways to parse messages which contain TV formatted IEs. This patch adds the relevant types and ways to encode/decode them. Furthermore, the current parser/builder allows parsing/building messages containing the exact same format in all its IEs. A new parser function is added which allows parsing messages of different types (TV, TLV) mixed in the same message. In order to be able to do so, it uses the general msg_mode passed to it in order to know the general TLV format (in essence, the length of the Tag field, and also the length of the Length field if applicable each IE). Looking up the instance in the TLV description is left undone and hadcoded to 0, since the only user so far requiring this API is GTPv1-C, which has no instances. * [CORE] tlv: Support repeated tag+instance parsing TLV message In GTPv2C, repeated IEs (same tag) are easily differentiated by the Instance byte, which provides info to match different decoded structures. In GTPv1C though, there's no Instance byte, and we still encounter repeated IEs (like GSN Address in Create PDP Context Request). Hence, the TLV decoder needs to be updated to track count of IEs found (identified by tag+instance, where instance is always 0 in GTPv1C) and get the proper description index + offset into the decoded structure. * [GTP]: Move GTPv2-C specifics to its own libgtp subdir This will allow adding GTPv1-C code by the side. Most GTPv2 code is left in this patch as "gtp" instead of renaming it to "gtp2" in order to avoid massive changes. It can be done at a later stage if wanted. * [GTP] Support generating GTPv1-C messages * [SMF] Add Gn interface support This patch introduces GTPv1C support to open5gs-smfd. With it, open5gs-becomes a GGSN too, where SGSN can connect to, hence supporting GERAN and UTRAN networks.
2022-02-18 13:23:45 +00:00
case OGS_PFCP_CAUSE_REQUEST_REJECTED:
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
return OGS_GTP2_CAUSE_REQUEST_REJECTED_REASON_NOT_SPECIFIED;
Introduce Gn interface (GTPv1C) Support to PGW (#1351) * [CORE] tlv: Store mode in ogs_tlv_t This allows specifying the format of the IE for each individual IE, hence allowing messages containing IEs formatted in different ways. This is needed in order to support parsing GTPv1-C, since messages contain IEs with different structure (TLV vs TV). Hence, this is a preparation patch to add support for parsing TVs in ogs-tlv.c/.h. * [CORE] tlv: Support parsing msg with both TLV and TV in it IEs of type TV are sometimes used in GTPv1-C. Current tlv parser/builder doesn't provide with ways to parse messages which contain TV formatted IEs. This patch adds the relevant types and ways to encode/decode them. Furthermore, the current parser/builder allows parsing/building messages containing the exact same format in all its IEs. A new parser function is added which allows parsing messages of different types (TV, TLV) mixed in the same message. In order to be able to do so, it uses the general msg_mode passed to it in order to know the general TLV format (in essence, the length of the Tag field, and also the length of the Length field if applicable each IE). Looking up the instance in the TLV description is left undone and hadcoded to 0, since the only user so far requiring this API is GTPv1-C, which has no instances. * [CORE] tlv: Support repeated tag+instance parsing TLV message In GTPv2C, repeated IEs (same tag) are easily differentiated by the Instance byte, which provides info to match different decoded structures. In GTPv1C though, there's no Instance byte, and we still encounter repeated IEs (like GSN Address in Create PDP Context Request). Hence, the TLV decoder needs to be updated to track count of IEs found (identified by tag+instance, where instance is always 0 in GTPv1C) and get the proper description index + offset into the decoded structure. * [GTP]: Move GTPv2-C specifics to its own libgtp subdir This will allow adding GTPv1-C code by the side. Most GTPv2 code is left in this patch as "gtp" instead of renaming it to "gtp2" in order to avoid massive changes. It can be done at a later stage if wanted. * [GTP] Support generating GTPv1-C messages * [SMF] Add Gn interface support This patch introduces GTPv1C support to open5gs-smfd. With it, open5gs-becomes a GGSN too, where SGSN can connect to, hence supporting GERAN and UTRAN networks.
2022-02-18 13:23:45 +00:00
case OGS_PFCP_CAUSE_SESSION_CONTEXT_NOT_FOUND:
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
return OGS_GTP2_CAUSE_CONTEXT_NOT_FOUND;
Introduce Gn interface (GTPv1C) Support to PGW (#1351) * [CORE] tlv: Store mode in ogs_tlv_t This allows specifying the format of the IE for each individual IE, hence allowing messages containing IEs formatted in different ways. This is needed in order to support parsing GTPv1-C, since messages contain IEs with different structure (TLV vs TV). Hence, this is a preparation patch to add support for parsing TVs in ogs-tlv.c/.h. * [CORE] tlv: Support parsing msg with both TLV and TV in it IEs of type TV are sometimes used in GTPv1-C. Current tlv parser/builder doesn't provide with ways to parse messages which contain TV formatted IEs. This patch adds the relevant types and ways to encode/decode them. Furthermore, the current parser/builder allows parsing/building messages containing the exact same format in all its IEs. A new parser function is added which allows parsing messages of different types (TV, TLV) mixed in the same message. In order to be able to do so, it uses the general msg_mode passed to it in order to know the general TLV format (in essence, the length of the Tag field, and also the length of the Length field if applicable each IE). Looking up the instance in the TLV description is left undone and hadcoded to 0, since the only user so far requiring this API is GTPv1-C, which has no instances. * [CORE] tlv: Support repeated tag+instance parsing TLV message In GTPv2C, repeated IEs (same tag) are easily differentiated by the Instance byte, which provides info to match different decoded structures. In GTPv1C though, there's no Instance byte, and we still encounter repeated IEs (like GSN Address in Create PDP Context Request). Hence, the TLV decoder needs to be updated to track count of IEs found (identified by tag+instance, where instance is always 0 in GTPv1C) and get the proper description index + offset into the decoded structure. * [GTP]: Move GTPv2-C specifics to its own libgtp subdir This will allow adding GTPv1-C code by the side. Most GTPv2 code is left in this patch as "gtp" instead of renaming it to "gtp2" in order to avoid massive changes. It can be done at a later stage if wanted. * [GTP] Support generating GTPv1-C messages * [SMF] Add Gn interface support This patch introduces GTPv1C support to open5gs-smfd. With it, open5gs-becomes a GGSN too, where SGSN can connect to, hence supporting GERAN and UTRAN networks.
2022-02-18 13:23:45 +00:00
case OGS_PFCP_CAUSE_MANDATORY_IE_MISSING:
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
return OGS_GTP2_CAUSE_MANDATORY_IE_MISSING;
Introduce Gn interface (GTPv1C) Support to PGW (#1351) * [CORE] tlv: Store mode in ogs_tlv_t This allows specifying the format of the IE for each individual IE, hence allowing messages containing IEs formatted in different ways. This is needed in order to support parsing GTPv1-C, since messages contain IEs with different structure (TLV vs TV). Hence, this is a preparation patch to add support for parsing TVs in ogs-tlv.c/.h. * [CORE] tlv: Support parsing msg with both TLV and TV in it IEs of type TV are sometimes used in GTPv1-C. Current tlv parser/builder doesn't provide with ways to parse messages which contain TV formatted IEs. This patch adds the relevant types and ways to encode/decode them. Furthermore, the current parser/builder allows parsing/building messages containing the exact same format in all its IEs. A new parser function is added which allows parsing messages of different types (TV, TLV) mixed in the same message. In order to be able to do so, it uses the general msg_mode passed to it in order to know the general TLV format (in essence, the length of the Tag field, and also the length of the Length field if applicable each IE). Looking up the instance in the TLV description is left undone and hadcoded to 0, since the only user so far requiring this API is GTPv1-C, which has no instances. * [CORE] tlv: Support repeated tag+instance parsing TLV message In GTPv2C, repeated IEs (same tag) are easily differentiated by the Instance byte, which provides info to match different decoded structures. In GTPv1C though, there's no Instance byte, and we still encounter repeated IEs (like GSN Address in Create PDP Context Request). Hence, the TLV decoder needs to be updated to track count of IEs found (identified by tag+instance, where instance is always 0 in GTPv1C) and get the proper description index + offset into the decoded structure. * [GTP]: Move GTPv2-C specifics to its own libgtp subdir This will allow adding GTPv1-C code by the side. Most GTPv2 code is left in this patch as "gtp" instead of renaming it to "gtp2" in order to avoid massive changes. It can be done at a later stage if wanted. * [GTP] Support generating GTPv1-C messages * [SMF] Add Gn interface support This patch introduces GTPv1C support to open5gs-smfd. With it, open5gs-becomes a GGSN too, where SGSN can connect to, hence supporting GERAN and UTRAN networks.
2022-02-18 13:23:45 +00:00
case OGS_PFCP_CAUSE_CONDITIONAL_IE_MISSING:
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
return OGS_GTP2_CAUSE_CONDITIONAL_IE_MISSING;
Introduce Gn interface (GTPv1C) Support to PGW (#1351) * [CORE] tlv: Store mode in ogs_tlv_t This allows specifying the format of the IE for each individual IE, hence allowing messages containing IEs formatted in different ways. This is needed in order to support parsing GTPv1-C, since messages contain IEs with different structure (TLV vs TV). Hence, this is a preparation patch to add support for parsing TVs in ogs-tlv.c/.h. * [CORE] tlv: Support parsing msg with both TLV and TV in it IEs of type TV are sometimes used in GTPv1-C. Current tlv parser/builder doesn't provide with ways to parse messages which contain TV formatted IEs. This patch adds the relevant types and ways to encode/decode them. Furthermore, the current parser/builder allows parsing/building messages containing the exact same format in all its IEs. A new parser function is added which allows parsing messages of different types (TV, TLV) mixed in the same message. In order to be able to do so, it uses the general msg_mode passed to it in order to know the general TLV format (in essence, the length of the Tag field, and also the length of the Length field if applicable each IE). Looking up the instance in the TLV description is left undone and hadcoded to 0, since the only user so far requiring this API is GTPv1-C, which has no instances. * [CORE] tlv: Support repeated tag+instance parsing TLV message In GTPv2C, repeated IEs (same tag) are easily differentiated by the Instance byte, which provides info to match different decoded structures. In GTPv1C though, there's no Instance byte, and we still encounter repeated IEs (like GSN Address in Create PDP Context Request). Hence, the TLV decoder needs to be updated to track count of IEs found (identified by tag+instance, where instance is always 0 in GTPv1C) and get the proper description index + offset into the decoded structure. * [GTP]: Move GTPv2-C specifics to its own libgtp subdir This will allow adding GTPv1-C code by the side. Most GTPv2 code is left in this patch as "gtp" instead of renaming it to "gtp2" in order to avoid massive changes. It can be done at a later stage if wanted. * [GTP] Support generating GTPv1-C messages * [SMF] Add Gn interface support This patch introduces GTPv1C support to open5gs-smfd. With it, open5gs-becomes a GGSN too, where SGSN can connect to, hence supporting GERAN and UTRAN networks.
2022-02-18 13:23:45 +00:00
case OGS_PFCP_CAUSE_INVALID_LENGTH:
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
return OGS_GTP2_CAUSE_INVALID_LENGTH;
Introduce Gn interface (GTPv1C) Support to PGW (#1351) * [CORE] tlv: Store mode in ogs_tlv_t This allows specifying the format of the IE for each individual IE, hence allowing messages containing IEs formatted in different ways. This is needed in order to support parsing GTPv1-C, since messages contain IEs with different structure (TLV vs TV). Hence, this is a preparation patch to add support for parsing TVs in ogs-tlv.c/.h. * [CORE] tlv: Support parsing msg with both TLV and TV in it IEs of type TV are sometimes used in GTPv1-C. Current tlv parser/builder doesn't provide with ways to parse messages which contain TV formatted IEs. This patch adds the relevant types and ways to encode/decode them. Furthermore, the current parser/builder allows parsing/building messages containing the exact same format in all its IEs. A new parser function is added which allows parsing messages of different types (TV, TLV) mixed in the same message. In order to be able to do so, it uses the general msg_mode passed to it in order to know the general TLV format (in essence, the length of the Tag field, and also the length of the Length field if applicable each IE). Looking up the instance in the TLV description is left undone and hadcoded to 0, since the only user so far requiring this API is GTPv1-C, which has no instances. * [CORE] tlv: Support repeated tag+instance parsing TLV message In GTPv2C, repeated IEs (same tag) are easily differentiated by the Instance byte, which provides info to match different decoded structures. In GTPv1C though, there's no Instance byte, and we still encounter repeated IEs (like GSN Address in Create PDP Context Request). Hence, the TLV decoder needs to be updated to track count of IEs found (identified by tag+instance, where instance is always 0 in GTPv1C) and get the proper description index + offset into the decoded structure. * [GTP]: Move GTPv2-C specifics to its own libgtp subdir This will allow adding GTPv1-C code by the side. Most GTPv2 code is left in this patch as "gtp" instead of renaming it to "gtp2" in order to avoid massive changes. It can be done at a later stage if wanted. * [GTP] Support generating GTPv1-C messages * [SMF] Add Gn interface support This patch introduces GTPv1C support to open5gs-smfd. With it, open5gs-becomes a GGSN too, where SGSN can connect to, hence supporting GERAN and UTRAN networks.
2022-02-18 13:23:45 +00:00
case OGS_PFCP_CAUSE_MANDATORY_IE_INCORRECT:
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
return OGS_GTP2_CAUSE_MANDATORY_IE_INCORRECT;
Introduce Gn interface (GTPv1C) Support to PGW (#1351) * [CORE] tlv: Store mode in ogs_tlv_t This allows specifying the format of the IE for each individual IE, hence allowing messages containing IEs formatted in different ways. This is needed in order to support parsing GTPv1-C, since messages contain IEs with different structure (TLV vs TV). Hence, this is a preparation patch to add support for parsing TVs in ogs-tlv.c/.h. * [CORE] tlv: Support parsing msg with both TLV and TV in it IEs of type TV are sometimes used in GTPv1-C. Current tlv parser/builder doesn't provide with ways to parse messages which contain TV formatted IEs. This patch adds the relevant types and ways to encode/decode them. Furthermore, the current parser/builder allows parsing/building messages containing the exact same format in all its IEs. A new parser function is added which allows parsing messages of different types (TV, TLV) mixed in the same message. In order to be able to do so, it uses the general msg_mode passed to it in order to know the general TLV format (in essence, the length of the Tag field, and also the length of the Length field if applicable each IE). Looking up the instance in the TLV description is left undone and hadcoded to 0, since the only user so far requiring this API is GTPv1-C, which has no instances. * [CORE] tlv: Support repeated tag+instance parsing TLV message In GTPv2C, repeated IEs (same tag) are easily differentiated by the Instance byte, which provides info to match different decoded structures. In GTPv1C though, there's no Instance byte, and we still encounter repeated IEs (like GSN Address in Create PDP Context Request). Hence, the TLV decoder needs to be updated to track count of IEs found (identified by tag+instance, where instance is always 0 in GTPv1C) and get the proper description index + offset into the decoded structure. * [GTP]: Move GTPv2-C specifics to its own libgtp subdir This will allow adding GTPv1-C code by the side. Most GTPv2 code is left in this patch as "gtp" instead of renaming it to "gtp2" in order to avoid massive changes. It can be done at a later stage if wanted. * [GTP] Support generating GTPv1-C messages * [SMF] Add Gn interface support This patch introduces GTPv1C support to open5gs-smfd. With it, open5gs-becomes a GGSN too, where SGSN can connect to, hence supporting GERAN and UTRAN networks.
2022-02-18 13:23:45 +00:00
case OGS_PFCP_CAUSE_INVALID_FORWARDING_POLICY:
case OGS_PFCP_CAUSE_INVALID_F_TEID_ALLOCATION_OPTION:
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
return OGS_GTP2_CAUSE_INVALID_MESSAGE_FORMAT;
Introduce Gn interface (GTPv1C) Support to PGW (#1351) * [CORE] tlv: Store mode in ogs_tlv_t This allows specifying the format of the IE for each individual IE, hence allowing messages containing IEs formatted in different ways. This is needed in order to support parsing GTPv1-C, since messages contain IEs with different structure (TLV vs TV). Hence, this is a preparation patch to add support for parsing TVs in ogs-tlv.c/.h. * [CORE] tlv: Support parsing msg with both TLV and TV in it IEs of type TV are sometimes used in GTPv1-C. Current tlv parser/builder doesn't provide with ways to parse messages which contain TV formatted IEs. This patch adds the relevant types and ways to encode/decode them. Furthermore, the current parser/builder allows parsing/building messages containing the exact same format in all its IEs. A new parser function is added which allows parsing messages of different types (TV, TLV) mixed in the same message. In order to be able to do so, it uses the general msg_mode passed to it in order to know the general TLV format (in essence, the length of the Tag field, and also the length of the Length field if applicable each IE). Looking up the instance in the TLV description is left undone and hadcoded to 0, since the only user so far requiring this API is GTPv1-C, which has no instances. * [CORE] tlv: Support repeated tag+instance parsing TLV message In GTPv2C, repeated IEs (same tag) are easily differentiated by the Instance byte, which provides info to match different decoded structures. In GTPv1C though, there's no Instance byte, and we still encounter repeated IEs (like GSN Address in Create PDP Context Request). Hence, the TLV decoder needs to be updated to track count of IEs found (identified by tag+instance, where instance is always 0 in GTPv1C) and get the proper description index + offset into the decoded structure. * [GTP]: Move GTPv2-C specifics to its own libgtp subdir This will allow adding GTPv1-C code by the side. Most GTPv2 code is left in this patch as "gtp" instead of renaming it to "gtp2" in order to avoid massive changes. It can be done at a later stage if wanted. * [GTP] Support generating GTPv1-C messages * [SMF] Add Gn interface support This patch introduces GTPv1C support to open5gs-smfd. With it, open5gs-becomes a GGSN too, where SGSN can connect to, hence supporting GERAN and UTRAN networks.
2022-02-18 13:23:45 +00:00
case OGS_PFCP_CAUSE_NO_ESTABLISHED_PFCP_ASSOCIATION:
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
return OGS_GTP2_CAUSE_REMOTE_PEER_NOT_RESPONDING;
Introduce Gn interface (GTPv1C) Support to PGW (#1351) * [CORE] tlv: Store mode in ogs_tlv_t This allows specifying the format of the IE for each individual IE, hence allowing messages containing IEs formatted in different ways. This is needed in order to support parsing GTPv1-C, since messages contain IEs with different structure (TLV vs TV). Hence, this is a preparation patch to add support for parsing TVs in ogs-tlv.c/.h. * [CORE] tlv: Support parsing msg with both TLV and TV in it IEs of type TV are sometimes used in GTPv1-C. Current tlv parser/builder doesn't provide with ways to parse messages which contain TV formatted IEs. This patch adds the relevant types and ways to encode/decode them. Furthermore, the current parser/builder allows parsing/building messages containing the exact same format in all its IEs. A new parser function is added which allows parsing messages of different types (TV, TLV) mixed in the same message. In order to be able to do so, it uses the general msg_mode passed to it in order to know the general TLV format (in essence, the length of the Tag field, and also the length of the Length field if applicable each IE). Looking up the instance in the TLV description is left undone and hadcoded to 0, since the only user so far requiring this API is GTPv1-C, which has no instances. * [CORE] tlv: Support repeated tag+instance parsing TLV message In GTPv2C, repeated IEs (same tag) are easily differentiated by the Instance byte, which provides info to match different decoded structures. In GTPv1C though, there's no Instance byte, and we still encounter repeated IEs (like GSN Address in Create PDP Context Request). Hence, the TLV decoder needs to be updated to track count of IEs found (identified by tag+instance, where instance is always 0 in GTPv1C) and get the proper description index + offset into the decoded structure. * [GTP]: Move GTPv2-C specifics to its own libgtp subdir This will allow adding GTPv1-C code by the side. Most GTPv2 code is left in this patch as "gtp" instead of renaming it to "gtp2" in order to avoid massive changes. It can be done at a later stage if wanted. * [GTP] Support generating GTPv1-C messages * [SMF] Add Gn interface support This patch introduces GTPv1C support to open5gs-smfd. With it, open5gs-becomes a GGSN too, where SGSN can connect to, hence supporting GERAN and UTRAN networks.
2022-02-18 13:23:45 +00:00
case OGS_PFCP_CAUSE_RULE_CREATION_MODIFICATION_FAILURE:
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
return OGS_GTP2_CAUSE_SEMANTIC_ERROR_IN_THE_TFT_OPERATION;
Introduce Gn interface (GTPv1C) Support to PGW (#1351) * [CORE] tlv: Store mode in ogs_tlv_t This allows specifying the format of the IE for each individual IE, hence allowing messages containing IEs formatted in different ways. This is needed in order to support parsing GTPv1-C, since messages contain IEs with different structure (TLV vs TV). Hence, this is a preparation patch to add support for parsing TVs in ogs-tlv.c/.h. * [CORE] tlv: Support parsing msg with both TLV and TV in it IEs of type TV are sometimes used in GTPv1-C. Current tlv parser/builder doesn't provide with ways to parse messages which contain TV formatted IEs. This patch adds the relevant types and ways to encode/decode them. Furthermore, the current parser/builder allows parsing/building messages containing the exact same format in all its IEs. A new parser function is added which allows parsing messages of different types (TV, TLV) mixed in the same message. In order to be able to do so, it uses the general msg_mode passed to it in order to know the general TLV format (in essence, the length of the Tag field, and also the length of the Length field if applicable each IE). Looking up the instance in the TLV description is left undone and hadcoded to 0, since the only user so far requiring this API is GTPv1-C, which has no instances. * [CORE] tlv: Support repeated tag+instance parsing TLV message In GTPv2C, repeated IEs (same tag) are easily differentiated by the Instance byte, which provides info to match different decoded structures. In GTPv1C though, there's no Instance byte, and we still encounter repeated IEs (like GSN Address in Create PDP Context Request). Hence, the TLV decoder needs to be updated to track count of IEs found (identified by tag+instance, where instance is always 0 in GTPv1C) and get the proper description index + offset into the decoded structure. * [GTP]: Move GTPv2-C specifics to its own libgtp subdir This will allow adding GTPv1-C code by the side. Most GTPv2 code is left in this patch as "gtp" instead of renaming it to "gtp2" in order to avoid massive changes. It can be done at a later stage if wanted. * [GTP] Support generating GTPv1-C messages * [SMF] Add Gn interface support This patch introduces GTPv1C support to open5gs-smfd. With it, open5gs-becomes a GGSN too, where SGSN can connect to, hence supporting GERAN and UTRAN networks.
2022-02-18 13:23:45 +00:00
case OGS_PFCP_CAUSE_PFCP_ENTITY_IN_CONGESTION:
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
return OGS_GTP2_CAUSE_GTP_C_ENTITY_CONGESTION;
Introduce Gn interface (GTPv1C) Support to PGW (#1351) * [CORE] tlv: Store mode in ogs_tlv_t This allows specifying the format of the IE for each individual IE, hence allowing messages containing IEs formatted in different ways. This is needed in order to support parsing GTPv1-C, since messages contain IEs with different structure (TLV vs TV). Hence, this is a preparation patch to add support for parsing TVs in ogs-tlv.c/.h. * [CORE] tlv: Support parsing msg with both TLV and TV in it IEs of type TV are sometimes used in GTPv1-C. Current tlv parser/builder doesn't provide with ways to parse messages which contain TV formatted IEs. This patch adds the relevant types and ways to encode/decode them. Furthermore, the current parser/builder allows parsing/building messages containing the exact same format in all its IEs. A new parser function is added which allows parsing messages of different types (TV, TLV) mixed in the same message. In order to be able to do so, it uses the general msg_mode passed to it in order to know the general TLV format (in essence, the length of the Tag field, and also the length of the Length field if applicable each IE). Looking up the instance in the TLV description is left undone and hadcoded to 0, since the only user so far requiring this API is GTPv1-C, which has no instances. * [CORE] tlv: Support repeated tag+instance parsing TLV message In GTPv2C, repeated IEs (same tag) are easily differentiated by the Instance byte, which provides info to match different decoded structures. In GTPv1C though, there's no Instance byte, and we still encounter repeated IEs (like GSN Address in Create PDP Context Request). Hence, the TLV decoder needs to be updated to track count of IEs found (identified by tag+instance, where instance is always 0 in GTPv1C) and get the proper description index + offset into the decoded structure. * [GTP]: Move GTPv2-C specifics to its own libgtp subdir This will allow adding GTPv1-C code by the side. Most GTPv2 code is left in this patch as "gtp" instead of renaming it to "gtp2" in order to avoid massive changes. It can be done at a later stage if wanted. * [GTP] Support generating GTPv1-C messages * [SMF] Add Gn interface support This patch introduces GTPv1C support to open5gs-smfd. With it, open5gs-becomes a GGSN too, where SGSN can connect to, hence supporting GERAN and UTRAN networks.
2022-02-18 13:23:45 +00:00
case OGS_PFCP_CAUSE_NO_RESOURCES_AVAILABLE:
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
return OGS_GTP2_CAUSE_NO_RESOURCES_AVAILABLE;
Introduce Gn interface (GTPv1C) Support to PGW (#1351) * [CORE] tlv: Store mode in ogs_tlv_t This allows specifying the format of the IE for each individual IE, hence allowing messages containing IEs formatted in different ways. This is needed in order to support parsing GTPv1-C, since messages contain IEs with different structure (TLV vs TV). Hence, this is a preparation patch to add support for parsing TVs in ogs-tlv.c/.h. * [CORE] tlv: Support parsing msg with both TLV and TV in it IEs of type TV are sometimes used in GTPv1-C. Current tlv parser/builder doesn't provide with ways to parse messages which contain TV formatted IEs. This patch adds the relevant types and ways to encode/decode them. Furthermore, the current parser/builder allows parsing/building messages containing the exact same format in all its IEs. A new parser function is added which allows parsing messages of different types (TV, TLV) mixed in the same message. In order to be able to do so, it uses the general msg_mode passed to it in order to know the general TLV format (in essence, the length of the Tag field, and also the length of the Length field if applicable each IE). Looking up the instance in the TLV description is left undone and hadcoded to 0, since the only user so far requiring this API is GTPv1-C, which has no instances. * [CORE] tlv: Support repeated tag+instance parsing TLV message In GTPv2C, repeated IEs (same tag) are easily differentiated by the Instance byte, which provides info to match different decoded structures. In GTPv1C though, there's no Instance byte, and we still encounter repeated IEs (like GSN Address in Create PDP Context Request). Hence, the TLV decoder needs to be updated to track count of IEs found (identified by tag+instance, where instance is always 0 in GTPv1C) and get the proper description index + offset into the decoded structure. * [GTP]: Move GTPv2-C specifics to its own libgtp subdir This will allow adding GTPv1-C code by the side. Most GTPv2 code is left in this patch as "gtp" instead of renaming it to "gtp2" in order to avoid massive changes. It can be done at a later stage if wanted. * [GTP] Support generating GTPv1-C messages * [SMF] Add Gn interface support This patch introduces GTPv1C support to open5gs-smfd. With it, open5gs-becomes a GGSN too, where SGSN can connect to, hence supporting GERAN and UTRAN networks.
2022-02-18 13:23:45 +00:00
case OGS_PFCP_CAUSE_SERVICE_NOT_SUPPORTED:
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
return OGS_GTP2_CAUSE_SERVICE_NOT_SUPPORTED;
Introduce Gn interface (GTPv1C) Support to PGW (#1351) * [CORE] tlv: Store mode in ogs_tlv_t This allows specifying the format of the IE for each individual IE, hence allowing messages containing IEs formatted in different ways. This is needed in order to support parsing GTPv1-C, since messages contain IEs with different structure (TLV vs TV). Hence, this is a preparation patch to add support for parsing TVs in ogs-tlv.c/.h. * [CORE] tlv: Support parsing msg with both TLV and TV in it IEs of type TV are sometimes used in GTPv1-C. Current tlv parser/builder doesn't provide with ways to parse messages which contain TV formatted IEs. This patch adds the relevant types and ways to encode/decode them. Furthermore, the current parser/builder allows parsing/building messages containing the exact same format in all its IEs. A new parser function is added which allows parsing messages of different types (TV, TLV) mixed in the same message. In order to be able to do so, it uses the general msg_mode passed to it in order to know the general TLV format (in essence, the length of the Tag field, and also the length of the Length field if applicable each IE). Looking up the instance in the TLV description is left undone and hadcoded to 0, since the only user so far requiring this API is GTPv1-C, which has no instances. * [CORE] tlv: Support repeated tag+instance parsing TLV message In GTPv2C, repeated IEs (same tag) are easily differentiated by the Instance byte, which provides info to match different decoded structures. In GTPv1C though, there's no Instance byte, and we still encounter repeated IEs (like GSN Address in Create PDP Context Request). Hence, the TLV decoder needs to be updated to track count of IEs found (identified by tag+instance, where instance is always 0 in GTPv1C) and get the proper description index + offset into the decoded structure. * [GTP]: Move GTPv2-C specifics to its own libgtp subdir This will allow adding GTPv1-C code by the side. Most GTPv2 code is left in this patch as "gtp" instead of renaming it to "gtp2" in order to avoid massive changes. It can be done at a later stage if wanted. * [GTP] Support generating GTPv1-C messages * [SMF] Add Gn interface support This patch introduces GTPv1C support to open5gs-smfd. With it, open5gs-becomes a GGSN too, where SGSN can connect to, hence supporting GERAN and UTRAN networks.
2022-02-18 13:23:45 +00:00
case OGS_PFCP_CAUSE_SYSTEM_FAILURE:
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
return OGS_GTP2_CAUSE_SYSTEM_FAILURE;
Introduce Gn interface (GTPv1C) Support to PGW (#1351) * [CORE] tlv: Store mode in ogs_tlv_t This allows specifying the format of the IE for each individual IE, hence allowing messages containing IEs formatted in different ways. This is needed in order to support parsing GTPv1-C, since messages contain IEs with different structure (TLV vs TV). Hence, this is a preparation patch to add support for parsing TVs in ogs-tlv.c/.h. * [CORE] tlv: Support parsing msg with both TLV and TV in it IEs of type TV are sometimes used in GTPv1-C. Current tlv parser/builder doesn't provide with ways to parse messages which contain TV formatted IEs. This patch adds the relevant types and ways to encode/decode them. Furthermore, the current parser/builder allows parsing/building messages containing the exact same format in all its IEs. A new parser function is added which allows parsing messages of different types (TV, TLV) mixed in the same message. In order to be able to do so, it uses the general msg_mode passed to it in order to know the general TLV format (in essence, the length of the Tag field, and also the length of the Length field if applicable each IE). Looking up the instance in the TLV description is left undone and hadcoded to 0, since the only user so far requiring this API is GTPv1-C, which has no instances. * [CORE] tlv: Support repeated tag+instance parsing TLV message In GTPv2C, repeated IEs (same tag) are easily differentiated by the Instance byte, which provides info to match different decoded structures. In GTPv1C though, there's no Instance byte, and we still encounter repeated IEs (like GSN Address in Create PDP Context Request). Hence, the TLV decoder needs to be updated to track count of IEs found (identified by tag+instance, where instance is always 0 in GTPv1C) and get the proper description index + offset into the decoded structure. * [GTP]: Move GTPv2-C specifics to its own libgtp subdir This will allow adding GTPv1-C code by the side. Most GTPv2 code is left in this patch as "gtp" instead of renaming it to "gtp2" in order to avoid massive changes. It can be done at a later stage if wanted. * [GTP] Support generating GTPv1-C messages * [SMF] Add Gn interface support This patch introduces GTPv1C support to open5gs-smfd. With it, open5gs-becomes a GGSN too, where SGSN can connect to, hence supporting GERAN and UTRAN networks.
2022-02-18 13:23:45 +00:00
case OGS_PFCP_CAUSE_ALL_DYNAMIC_ADDRESS_ARE_OCCUPIED:
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
return OGS_GTP2_CAUSE_ALL_DYNAMIC_ADDRESSES_ARE_OCCUPIED;
Introduce Gn interface (GTPv1C) Support to PGW (#1351) * [CORE] tlv: Store mode in ogs_tlv_t This allows specifying the format of the IE for each individual IE, hence allowing messages containing IEs formatted in different ways. This is needed in order to support parsing GTPv1-C, since messages contain IEs with different structure (TLV vs TV). Hence, this is a preparation patch to add support for parsing TVs in ogs-tlv.c/.h. * [CORE] tlv: Support parsing msg with both TLV and TV in it IEs of type TV are sometimes used in GTPv1-C. Current tlv parser/builder doesn't provide with ways to parse messages which contain TV formatted IEs. This patch adds the relevant types and ways to encode/decode them. Furthermore, the current parser/builder allows parsing/building messages containing the exact same format in all its IEs. A new parser function is added which allows parsing messages of different types (TV, TLV) mixed in the same message. In order to be able to do so, it uses the general msg_mode passed to it in order to know the general TLV format (in essence, the length of the Tag field, and also the length of the Length field if applicable each IE). Looking up the instance in the TLV description is left undone and hadcoded to 0, since the only user so far requiring this API is GTPv1-C, which has no instances. * [CORE] tlv: Support repeated tag+instance parsing TLV message In GTPv2C, repeated IEs (same tag) are easily differentiated by the Instance byte, which provides info to match different decoded structures. In GTPv1C though, there's no Instance byte, and we still encounter repeated IEs (like GSN Address in Create PDP Context Request). Hence, the TLV decoder needs to be updated to track count of IEs found (identified by tag+instance, where instance is always 0 in GTPv1C) and get the proper description index + offset into the decoded structure. * [GTP]: Move GTPv2-C specifics to its own libgtp subdir This will allow adding GTPv1-C code by the side. Most GTPv2 code is left in this patch as "gtp" instead of renaming it to "gtp2" in order to avoid massive changes. It can be done at a later stage if wanted. * [GTP] Support generating GTPv1-C messages * [SMF] Add Gn interface support This patch introduces GTPv1C support to open5gs-smfd. With it, open5gs-becomes a GGSN too, where SGSN can connect to, hence supporting GERAN and UTRAN networks.
2022-02-18 13:23:45 +00:00
default:
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
return OGS_GTP2_CAUSE_SYSTEM_FAILURE;
Introduce Gn interface (GTPv1C) Support to PGW (#1351) * [CORE] tlv: Store mode in ogs_tlv_t This allows specifying the format of the IE for each individual IE, hence allowing messages containing IEs formatted in different ways. This is needed in order to support parsing GTPv1-C, since messages contain IEs with different structure (TLV vs TV). Hence, this is a preparation patch to add support for parsing TVs in ogs-tlv.c/.h. * [CORE] tlv: Support parsing msg with both TLV and TV in it IEs of type TV are sometimes used in GTPv1-C. Current tlv parser/builder doesn't provide with ways to parse messages which contain TV formatted IEs. This patch adds the relevant types and ways to encode/decode them. Furthermore, the current parser/builder allows parsing/building messages containing the exact same format in all its IEs. A new parser function is added which allows parsing messages of different types (TV, TLV) mixed in the same message. In order to be able to do so, it uses the general msg_mode passed to it in order to know the general TLV format (in essence, the length of the Tag field, and also the length of the Length field if applicable each IE). Looking up the instance in the TLV description is left undone and hadcoded to 0, since the only user so far requiring this API is GTPv1-C, which has no instances. * [CORE] tlv: Support repeated tag+instance parsing TLV message In GTPv2C, repeated IEs (same tag) are easily differentiated by the Instance byte, which provides info to match different decoded structures. In GTPv1C though, there's no Instance byte, and we still encounter repeated IEs (like GSN Address in Create PDP Context Request). Hence, the TLV decoder needs to be updated to track count of IEs found (identified by tag+instance, where instance is always 0 in GTPv1C) and get the proper description index + offset into the decoded structure. * [GTP]: Move GTPv2-C specifics to its own libgtp subdir This will allow adding GTPv1-C code by the side. Most GTPv2 code is left in this patch as "gtp" instead of renaming it to "gtp2" in order to avoid massive changes. It can be done at a later stage if wanted. * [GTP] Support generating GTPv1-C messages * [SMF] Add Gn interface support This patch introduces GTPv1C support to open5gs-smfd. With it, open5gs-becomes a GGSN too, where SGSN can connect to, hence supporting GERAN and UTRAN networks.
2022-02-18 13:23:45 +00:00
}
break;
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
}
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
return OGS_GTP2_CAUSE_SYSTEM_FAILURE;
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
}
5gc: Paging was added
2021-01-18 16:48:35 +00:00
Fix the SMF crash bug [#487]
2020-07-04 03:14:48 +00:00
static int sbi_status_from_pfcp(uint8_t pfcp_cause)
{
switch (pfcp_cause) {
case OGS_PFCP_CAUSE_REQUEST_ACCEPTED:
return OGS_SBI_HTTP_STATUS_OK;
case OGS_PFCP_CAUSE_REQUEST_REJECTED:
return OGS_SBI_HTTP_STATUS_FORBIDDEN;
case OGS_PFCP_CAUSE_SESSION_CONTEXT_NOT_FOUND:
return OGS_SBI_HTTP_STATUS_NOT_FOUND;
case OGS_PFCP_CAUSE_MANDATORY_IE_MISSING:
case OGS_PFCP_CAUSE_CONDITIONAL_IE_MISSING:
case OGS_PFCP_CAUSE_INVALID_LENGTH:
case OGS_PFCP_CAUSE_MANDATORY_IE_INCORRECT:
case OGS_PFCP_CAUSE_INVALID_FORWARDING_POLICY:
case OGS_PFCP_CAUSE_INVALID_F_TEID_ALLOCATION_OPTION:
case OGS_PFCP_CAUSE_RULE_CREATION_MODIFICATION_FAILURE:
case OGS_PFCP_CAUSE_PFCP_ENTITY_IN_CONGESTION:
case OGS_PFCP_CAUSE_NO_RESOURCES_AVAILABLE:
return OGS_SBI_HTTP_STATUS_BAD_REQUEST;
case OGS_PFCP_CAUSE_NO_ESTABLISHED_PFCP_ASSOCIATION:
return OGS_SBI_HTTP_STATUS_GATEWAY_TIMEOUT;
case OGS_PFCP_CAUSE_SERVICE_NOT_SUPPORTED:
return OGS_SBI_HTTP_STATUS_SERVICE_UNAVAILABLE;
case OGS_PFCP_CAUSE_SYSTEM_FAILURE:
return OGS_SBI_HTTP_STATUS_INTERNAL_SERVER_ERROR;
default:
return OGS_SBI_HTTP_STATUS_INTERNAL_SERVER_ERROR;
}
return OGS_SBI_HTTP_STATUS_INTERNAL_SERVER_ERROR;
}
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
[SMF] Apply 5G Core into GSM (#1498)
2022-05-14 21:27:54 +00:00
/* Returns OGS_PFCP_CAUSE_REQUEST_ACCEPTED on success,
* other cause value on failure */
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
uint8_t smf_5gc_n4_handle_session_establishment_response(
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
smf_sess_t *sess, ogs_pfcp_xact_t *xact,
ogs_pfcp_session_establishment_response_t *rsp)
{
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
int i;
Use HTTP/2 instead of HTTP/1.1 in 5G Core SBI
2020-11-27 02:44:37 +00:00
ogs_sbi_stream_t *stream = NULL;
Re-factoring 5GC for solving crash issue
2020-08-03 03:22:41 +00:00
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
uint8_t cause_value = OGS_PFCP_CAUSE_REQUEST_ACCEPTED;
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
uint8_t offending_ie_value = 0;
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
ogs_pfcp_f_seid_t *up_f_seid = NULL;
[PFCP] F-TEID allocation by SMF (#1466)
2022-04-16 05:08:56 +00:00
ogs_pfcp_pdr_t *pdr = NULL;
ogs_pfcp_far_t *far = NULL;
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
ogs_assert(sess);
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
ogs_assert(xact);
ogs_assert(rsp);
logging improvements (#1348) * [SMF] Fix typo in debug message * [SMF] Add debug messages to follow messages received in N4
2022-02-02 14:23:12 +00:00
ogs_debug("Session Establishment Response [5gc]");
Use HTTP/2 instead of HTTP/1.1 in 5G Core SBI
2020-11-27 02:44:37 +00:00
stream = xact->assoc_stream;
ogs_assert(stream);
Re-factoring 5GC for solving crash issue
2020-08-03 03:22:41 +00:00
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
ogs_pfcp_xact_commit(xact);
if (rsp->up_f_seid.presence == 0) {
ogs_error("No UP F-SEID");
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
cause_value = OGS_PFCP_CAUSE_MANDATORY_IE_MISSING;
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
}
if (rsp->cause.presence) {
if (rsp->cause.u8 != OGS_PFCP_CAUSE_REQUEST_ACCEPTED) {
Fix the SMF crash bug [#487]
2020-07-04 03:14:48 +00:00
ogs_error("PFCP Cause [%d] : Not Accepted", rsp->cause.u8);
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
cause_value = rsp->cause.u8;
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
}
} else {
ogs_error("No Cause");
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
cause_value = OGS_PFCP_CAUSE_MANDATORY_IE_MISSING;
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
}
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
if (cause_value != OGS_PFCP_CAUSE_REQUEST_ACCEPTED)
return cause_value;
Fix the SMF crash bug [#487]
2020-07-04 03:14:48 +00:00
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
for (i = 0; i < OGS_MAX_NUM_OF_PDR; i++) {
pdr = ogs_pfcp_handle_created_pdr(
&sess->pfcp, &rsp->created_pdr[i],
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
&cause_value, &offending_ie_value);
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
if (!pdr)
break;
[PFCP] F-TEID allocation by SMF (#1466)
2022-04-16 05:08:56 +00:00
}
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
[PFCP] F-TEID allocation by SMF (#1466)
2022-04-16 05:08:56 +00:00
ogs_list_for_each(&sess->pfcp.pdr_list, pdr) {
Fixes UE IPv6 BUG (#808)
2021-03-15 01:01:55 +00:00
far = pdr->far;
ogs_assert(far);
if (pdr->src_if == OGS_PFCP_INTERFACE_ACCESS) {
if (far->dst_if == OGS_PFCP_INTERFACE_CP_FUNCTION)
ogs_pfcp_far_teid_hash_set(far);
ogs_assert(sess->pfcp_node);
if (sess->pfcp_node->up_function_features.ftup &&
pdr->f_teid_len) {
if (sess->upf_n3_addr)
ogs_freeaddrinfo(sess->upf_n3_addr);
if (sess->upf_n3_addr6)
ogs_freeaddrinfo(sess->upf_n3_addr6);
Add more memory log
2021-06-06 13:35:46 +00:00
ogs_assert(OGS_OK ==
ogs_pfcp_f_teid_to_sockaddr(
Fixes UE IPv6 BUG (#808)
2021-03-15 01:01:55 +00:00
&pdr->f_teid, pdr->f_teid_len,
Add more memory log
2021-06-06 13:35:46 +00:00
&sess->upf_n3_addr, &sess->upf_n3_addr6));
Fixes UE IPv6 BUG (#808)
2021-03-15 01:01:55 +00:00
sess->upf_n3_teid = pdr->f_teid.teid;
}
} else if (pdr->src_if == OGS_PFCP_INTERFACE_CP_FUNCTION) {
Add more memory log
2021-06-06 13:35:46 +00:00
ogs_assert(OGS_ERROR != ogs_pfcp_setup_pdr_gtpu_node(pdr));
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
}
}
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
if (cause_value != OGS_PFCP_CAUSE_REQUEST_ACCEPTED) {
ogs_error("PFCP Cause [%d] : Not Accepted", cause_value);
return cause_value;
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
}
if (sess->upf_n3_addr == NULL && sess->upf_n3_addr6 == NULL) {
ogs_error("No UP F-TEID");
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
return OGS_PFCP_CAUSE_SESSION_CONTEXT_NOT_FOUND;
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
}
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
/* UP F-SEID */
up_f_seid = rsp->up_f_seid.data;
ogs_assert(up_f_seid);
sess->upf_n4_seid = be64toh(up_f_seid->seid);
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
return OGS_PFCP_CAUSE_REQUEST_ACCEPTED;
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
}
void smf_5gc_n4_handle_session_modification_response(
smf_sess_t *sess, ogs_pfcp_xact_t *xact,
ogs_pfcp_session_modification_response_t *rsp)
{
Fix the SMF crash bug [#487]
2020-07-04 03:14:48 +00:00
int status = 0;
Re-factoring 5GC for solving crash issue
2020-08-03 03:22:41 +00:00
uint64_t flags = 0;
Use HTTP/2 instead of HTTP/1.1 in 5G Core SBI
2020-11-27 02:44:37 +00:00
ogs_sbi_stream_t *stream = NULL;
feat: Add dedicated QoS flow
2021-01-01 02:07:08 +00:00
smf_bearer_t *qos_flow = NULL;
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
[PFCP] F-TEID allocation by SMF (#1466)
2022-04-16 05:08:56 +00:00
OGS_LIST(pdr_to_create_list);
[GTP2] Fixed handling multi-bearer messages(#1498)
2022-04-29 12:28:16 +00:00
ogs_debug("Session Modification Response [5gc]");
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
ogs_assert(xact);
ogs_assert(rsp);
Re-factoring 5GC for solving crash issue
2020-08-03 03:22:41 +00:00
flags = xact->modify_flags;
ogs_assert(flags);
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
feat: Add dedicated QoS flow
2021-01-01 02:07:08 +00:00
/* 'stream' could be NULL in smf_qos_flow_binding() */
stream = xact->assoc_stream;
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
if (flags & OGS_PFCP_MODIFY_SESSION) {
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
/* If smf_5gc_pfcp_send_pdr_modification_request() is called */
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
} else {
/* If smf_5gc_pfcp_send_qos_flow_modification_request() is called */
qos_flow = xact->data;
ogs_assert(qos_flow);
}
feat: Add dedicated QoS flow
2021-01-01 02:07:08 +00:00
[PFCP] F-TEID allocation by SMF (#1466)
2022-04-16 05:08:56 +00:00
ogs_list_copy(&pdr_to_create_list, &xact->pdr_to_create_list);
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
ogs_pfcp_xact_commit(xact);
Fix the SMF crash bug [#487]
2020-07-04 03:14:48 +00:00
status = OGS_SBI_HTTP_STATUS_OK;
if (!sess) {
ogs_warn("No Context");
status = OGS_SBI_HTTP_STATUS_NOT_FOUND;
}
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
if (rsp->cause.presence) {
if (rsp->cause.u8 != OGS_PFCP_CAUSE_REQUEST_ACCEPTED) {
Fix the SMF crash bug [#487]
2020-07-04 03:14:48 +00:00
ogs_warn("PFCP Cause [%d] : Not Accepted", rsp->cause.u8);
status = sbi_status_from_pfcp(rsp->cause.u8);
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
}
} else {
ogs_error("No Cause");
Fix the SMF crash bug [#487]
2020-07-04 03:14:48 +00:00
status = OGS_SBI_HTTP_STATUS_BAD_REQUEST;
}
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
if (status == OGS_SBI_HTTP_STATUS_OK) {
int i;
uint8_t pfcp_cause_value = OGS_PFCP_CAUSE_REQUEST_ACCEPTED;
uint8_t offending_ie_value = 0;
[PFCP] F-TEID allocation by SMF (#1466)
2022-04-16 05:08:56 +00:00
ogs_pfcp_pdr_t *pdr = NULL;
ogs_pfcp_far_t *far = NULL;
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
ogs_assert(sess);
for (i = 0; i < OGS_MAX_NUM_OF_PDR; i++) {
pdr = ogs_pfcp_handle_created_pdr(
&sess->pfcp, &rsp->created_pdr[i],
&pfcp_cause_value, &offending_ie_value);
if (!pdr)
break;
[PFCP] F-TEID allocation by SMF (#1466)
2022-04-16 05:08:56 +00:00
}
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
[PFCP] F-TEID allocation by SMF (#1466)
2022-04-16 05:08:56 +00:00
ogs_list_for_each_entry(&pdr_to_create_list, pdr, to_create_node) {
Fixes UE IPv6 BUG (#808)
2021-03-15 01:01:55 +00:00
far = pdr->far;
ogs_assert(far);
if (pdr->src_if == OGS_PFCP_INTERFACE_ACCESS) {
if (far->dst_if == OGS_PFCP_INTERFACE_CP_FUNCTION)
ogs_pfcp_far_teid_hash_set(far);
ogs_assert(sess->pfcp_node);
if (sess->pfcp_node->up_function_features.ftup &&
pdr->f_teid_len) {
if (far->dst_if == OGS_PFCP_INTERFACE_CORE) {
if (sess->upf_n3_addr)
ogs_freeaddrinfo(sess->upf_n3_addr);
if (sess->upf_n3_addr6)
ogs_freeaddrinfo(sess->upf_n3_addr6);
Add more memory log
2021-06-06 13:35:46 +00:00
ogs_assert(OGS_OK ==
ogs_pfcp_f_teid_to_sockaddr(
Fixes UE IPv6 BUG (#808)
2021-03-15 01:01:55 +00:00
&pdr->f_teid, pdr->f_teid_len,
Add more memory log
2021-06-06 13:35:46 +00:00
&sess->upf_n3_addr, &sess->upf_n3_addr6));
Fixes UE IPv6 BUG (#808)
2021-03-15 01:01:55 +00:00
sess->upf_n3_teid = pdr->f_teid.teid;
} else if (far->dst_if == OGS_PFCP_INTERFACE_ACCESS) {
if (sess->handover.upf_dl_addr)
ogs_freeaddrinfo(sess->handover.upf_dl_addr);
if (sess->handover.upf_dl_addr6)
ogs_freeaddrinfo(sess->handover.upf_dl_addr6);
Add more memory log
2021-06-06 13:35:46 +00:00
ogs_assert(OGS_OK ==
ogs_pfcp_f_teid_to_sockaddr(
Fixes UE IPv6 BUG (#808)
2021-03-15 01:01:55 +00:00
&pdr->f_teid, pdr->f_teid_len,
&sess->handover.upf_dl_addr,
Add more memory log
2021-06-06 13:35:46 +00:00
&sess->handover.upf_dl_addr6));
Fixes UE IPv6 BUG (#808)
2021-03-15 01:01:55 +00:00
sess->handover.upf_dl_teid = pdr->f_teid.teid;
}
feat: Indirect Data Forwarding in N2-Handover
2021-02-01 04:01:15 +00:00
}
Fixes UE IPv6 BUG (#808)
2021-03-15 01:01:55 +00:00
} else if (pdr->src_if == OGS_PFCP_INTERFACE_CP_FUNCTION) {
Add more memory log
2021-06-06 13:35:46 +00:00
ogs_assert(OGS_ERROR != ogs_pfcp_setup_pdr_gtpu_node(pdr));
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
}
}
status = sbi_status_from_pfcp(pfcp_cause_value);
}
Fix the SMF crash bug [#487]
2020-07-04 03:14:48 +00:00
if (status != OGS_SBI_HTTP_STATUS_OK) {
char *strerror = ogs_msprintf(
"PFCP Cause [%d] : Not Accepted", rsp->cause.u8);
feat: Add dedicated QoS flow
2021-01-01 02:07:08 +00:00
if (stream)
smf_sbi_send_sm_context_update_error(
stream, status, strerror, NULL, NULL, NULL);
[SMF] Exception handling in session release(#1155) - UE changes the airplain-mode. However, it does not send PDU session complete message - Added exception handling for PDU session release
2021-09-04 07:14:18 +00:00
ogs_error("%s", strerror);
Fix the SMF crash bug [#487]
2020-07-04 03:14:48 +00:00
ogs_free(strerror);
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
return;
}
Fix the SMF crash bug [#487]
2020-07-04 03:14:48 +00:00
ogs_assert(sess);
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
if (sess->upf_n3_addr == NULL && sess->upf_n3_addr6 == NULL) {
feat: Add dedicated QoS flow
2021-01-01 02:07:08 +00:00
if (stream)
smf_sbi_send_sm_context_update_error(
stream, status, "No UP F_TEID", NULL, NULL, NULL);
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
return;
}
Split SGW into SGW-C/SGW-U
2020-08-13 00:31:22 +00:00
if (flags & OGS_PFCP_MODIFY_ACTIVATE) {
feat: Add N2-Handover
2021-01-28 19:23:54 +00:00
if (flags & OGS_PFCP_MODIFY_XN_HANDOVER) {
feat: Add 5G Core Xn Handover (#726)
2021-01-03 05:50:59 +00:00
ogs_pkbuf_t *n2smbuf =
ngap_build_path_switch_request_ack_transfer(sess);
ogs_assert(n2smbuf);
5gc: Paging was added
2021-01-18 16:48:35 +00:00
smf_sbi_send_sm_context_updated_data_n2smbuf(sess, stream,
feat: Add 5G Core Xn Handover (#726)
2021-01-03 05:50:59 +00:00
OpenAPI_n2_sm_info_type_PATH_SWITCH_REQ_ACK, n2smbuf);
feat: Add N2-Handover
2021-01-28 19:23:54 +00:00
} else if (flags & OGS_PFCP_MODIFY_N2_HANDOVER) {
feat: Indirect Data Forwarding in N2-Handover
2021-02-01 04:01:15 +00:00
if (smf_sess_have_indirect_data_forwarding(sess) == true) {
[ALL] Add protection of message encoding (#991) All process will be forcely exited if it failed to encode the S1AP/NGAP/GTP/PFCP message. It is to make sure there was no problem with the encoding of open5gs.
2021-05-16 03:22:10 +00:00
ogs_assert(OGS_OK ==
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
smf_5gc_pfcp_send_pdr_modification_request(
feat: Indirect Data Forwarding in N2-Handover
2021-02-01 04:01:15 +00:00
sess, stream,
OGS_PFCP_MODIFY_INDIRECT|OGS_PFCP_MODIFY_REMOVE,
[ALL] Add protection of message encoding (#991) All process will be forcely exited if it failed to encode the S1AP/NGAP/GTP/PFCP message. It is to make sure there was no problem with the encoding of open5gs.
2021-05-16 03:22:10 +00:00
ogs_app()->time.handover.duration));
feat: Indirect Data Forwarding in N2-Handover
2021-02-01 04:01:15 +00:00
}
feat: Add N2-Handover
2021-01-28 19:23:54 +00:00
smf_sbi_send_sm_context_updated_data_ho_state(
sess, stream, OpenAPI_ho_state_COMPLETED);
5gc: Paging was added
2021-01-18 16:48:35 +00:00
} else {
sess->paging.ue_requested_pdu_session_establishment_done = true;
Add more memory log
2021-06-06 13:35:46 +00:00
ogs_assert(true == ogs_sbi_send_http_status_no_content(stream));
feat: Add 5G Core Xn Handover (#726)
2021-01-03 05:50:59 +00:00
}
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
Split SGW into SGW-C/SGW-U
2020-08-13 00:31:22 +00:00
} else if (flags & OGS_PFCP_MODIFY_DEACTIVATE) {
5gc: Paging was added
2021-01-18 16:48:35 +00:00
if (flags & OGS_PFCP_MODIFY_ERROR_INDICATION) {
smf_n1_n2_message_transfer_param_t param;
memset(&param, 0, sizeof(param));
param.state = SMF_ERROR_INDICATON_RECEIVED_FROM_5G_AN;
param.n2smbuf =
ngap_build_pdu_session_resource_release_command_transfer(
sess, SMF_NGAP_STATE_ERROR_INDICATION_RECEIVED_FROM_5G_AN,
NGAP_Cause_PR_nas, NGAP_CauseNas_normal_release);
ogs_assert(param.n2smbuf);
param.skip_ind = true;
smf_namf_comm_send_n1_n2_message_transfer(sess, &param);
} else {
smf_sbi_send_sm_context_updated_data_up_cnx_state(
sess, stream, OpenAPI_up_cnx_state_DEACTIVATED);
}
feat: Indirect Data Forwarding in N2-Handover
2021-02-01 04:01:15 +00:00
/*
* You should not change the following order to support
* OGS_PFCP_MODIFY_REMOVE|OGS_PFCP_MODIFY_CREATE.
*
* 1. if (flags & OGS_PFCP_MODIFY_REMOVE) {
* 2. } else if (flags & OGS_PFCP_MODIFY_CREATE) {
* }
*/
} else if (flags & OGS_PFCP_MODIFY_REMOVE) {
if (flags & OGS_PFCP_MODIFY_INDIRECT) {
smf_sess_delete_indirect_data_forwarding(sess);
/*
* OGS_PFCP_MODIFY_CREATE remains.
* So now we do some extra work to create an indirect tunnel.
*/
if (flags & OGS_PFCP_MODIFY_CREATE) {
smf_sess_create_indirect_data_forwarding(sess);
[ALL] Add protection of message encoding (#991) All process will be forcely exited if it failed to encode the S1AP/NGAP/GTP/PFCP message. It is to make sure there was no problem with the encoding of open5gs.
2021-05-16 03:22:10 +00:00
ogs_assert(OGS_OK ==
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
smf_5gc_pfcp_send_pdr_modification_request(
feat: Indirect Data Forwarding in N2-Handover
2021-02-01 04:01:15 +00:00
sess, stream,
OGS_PFCP_MODIFY_INDIRECT|OGS_PFCP_MODIFY_CREATE,
[ALL] Add protection of message encoding (#991) All process will be forcely exited if it failed to encode the S1AP/NGAP/GTP/PFCP message. It is to make sure there was no problem with the encoding of open5gs.
2021-05-16 03:22:10 +00:00
0));
feat: Indirect Data Forwarding in N2-Handover
2021-02-01 04:01:15 +00:00
} else if (flags & OGS_PFCP_MODIFY_HANDOVER_CANCEL) {
smf_sbi_send_sm_context_updated_data_ho_state(
sess, stream, OpenAPI_ho_state_CANCELLED);
}
VoNR added but not tested!
2021-11-14 12:07:56 +00:00
} else if (flags & OGS_PFCP_MODIFY_NETWORK_REQUESTED) {
smf_n1_n2_message_transfer_param_t param;
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
smf_bearer_t *next = NULL;
VoNR added but not tested!
2021-11-14 12:07:56 +00:00
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
ogs_assert(flags & OGS_PFCP_MODIFY_SESSION);
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
VoNR added but not tested!
2021-11-14 12:07:56 +00:00
memset(&param, 0, sizeof(param));
param.state = SMF_NETWORK_REQUESTED_QOS_FLOW_MODIFICATION;
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
param.n1smbuf = gsm_build_pdu_session_modification_command(
sess,
VoNR added but not tested!
2021-11-14 12:07:56 +00:00
OGS_NAS_PROCEDURE_TRANSACTION_IDENTITY_UNASSIGNED,
OGS_NAS_QOS_CODE_DELETE_EXISTING_QOS_RULE,
OGS_NAS_DELETE_NEW_QOS_FLOW_DESCRIPTION);
ogs_assert(param.n1smbuf);
param.n2smbuf =
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
ngap_build_pdu_session_resource_release_request_transfer(
sess,
VoNR added but not tested!
2021-11-14 12:07:56 +00:00
NGAP_Cause_PR_nas, NGAP_CauseNas_normal_release);
ogs_assert(param.n2smbuf);
smf_namf_comm_send_n1_n2_message_transfer(sess, &param);
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
ogs_list_for_each_entry_safe(&sess->qos_flow_to_modify_list,
next, qos_flow, to_modify_node)
smf_bearer_remove(qos_flow);
VoNR added but not tested!
2021-11-14 12:07:56 +00:00
} else if (flags & OGS_PFCP_MODIFY_UE_REQUESTED) {
ogs_pkbuf_t *n1smbuf = NULL, *n2smbuf = NULL;
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
smf_bearer_t *next = NULL;
VoNR added but not tested!
2021-11-14 12:07:56 +00:00
ogs_assert(stream);
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
ogs_assert(flags & OGS_PFCP_MODIFY_SESSION);
n1smbuf = gsm_build_pdu_session_modification_command(
sess, sess->pti,
VoNR added but not tested!
2021-11-14 12:07:56 +00:00
OGS_NAS_QOS_CODE_DELETE_EXISTING_QOS_RULE,
OGS_NAS_DELETE_NEW_QOS_FLOW_DESCRIPTION);
ogs_assert(n1smbuf);
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
n2smbuf = ngap_build_pdu_session_resource_release_request_transfer(
sess,
VoNR added but not tested!
2021-11-14 12:07:56 +00:00
NGAP_Cause_PR_nas, NGAP_CauseNas_normal_release);
ogs_assert(n2smbuf);
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
smf_sbi_send_sm_context_updated_data_n1_n2_message(
sess, stream, n1smbuf,
OpenAPI_n2_sm_info_type_PDU_RES_MOD_REQ, n2smbuf);
ogs_list_for_each_entry_safe(&sess->qos_flow_to_modify_list,
next, qos_flow, to_modify_node)
smf_bearer_remove(qos_flow);
VoNR added but not tested!
2021-11-14 12:07:56 +00:00
} else {
ogs_fatal("Unknown flags [0x%llx]", (long long)flags);
ogs_assert_if_reached();
feat: Indirect Data Forwarding in N2-Handover
2021-02-01 04:01:15 +00:00
}
feat: Add dedicated QoS flow
2021-01-01 02:07:08 +00:00
} else if (flags & OGS_PFCP_MODIFY_CREATE) {
feat: Indirect Data Forwarding in N2-Handover
2021-02-01 04:01:15 +00:00
if (flags & OGS_PFCP_MODIFY_INDIRECT) {
ogs_pkbuf_t *n2smbuf = ngap_build_handover_command_transfer(sess);
ogs_assert(n2smbuf);
smf_sbi_send_sm_context_updated_data(
sess, stream, 0, OpenAPI_ho_state_PREPARED,
NULL, OpenAPI_n2_sm_info_type_HANDOVER_CMD, n2smbuf);
VoNR added but not tested!
2021-11-14 12:07:56 +00:00
} else if (flags & OGS_PFCP_MODIFY_NETWORK_REQUESTED) {
feat: Indirect Data Forwarding in N2-Handover
2021-02-01 04:01:15 +00:00
smf_n1_n2_message_transfer_param_t param;
feat: Add dedicated QoS flow
2021-01-01 02:07:08 +00:00
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
ogs_assert(flags & OGS_PFCP_MODIFY_SESSION);
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
feat: Indirect Data Forwarding in N2-Handover
2021-02-01 04:01:15 +00:00
memset(&param, 0, sizeof(param));
param.state = SMF_NETWORK_REQUESTED_QOS_FLOW_MODIFICATION;
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
param.n1smbuf = gsm_build_pdu_session_modification_command(
sess,
VoNR added but not tested!
2021-11-14 12:07:56 +00:00
OGS_NAS_PROCEDURE_TRANSACTION_IDENTITY_UNASSIGNED,
OGS_NAS_QOS_CODE_CREATE_NEW_QOS_RULE,
OGS_NAS_CREATE_NEW_QOS_FLOW_DESCRIPTION);
ogs_assert(param.n1smbuf);
param.n2smbuf =
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
ngap_build_pdu_session_resource_modify_request_transfer(
sess, true);
VoNR added but not tested!
2021-11-14 12:07:56 +00:00
ogs_assert(param.n2smbuf);
smf_namf_comm_send_n1_n2_message_transfer(sess, &param);
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
VoNR added but not tested!
2021-11-14 12:07:56 +00:00
} else {
ogs_fatal("Unknown flags [0x%llx]", (long long)flags);
ogs_assert_if_reached();
}
} else if (flags &
(OGS_PFCP_MODIFY_TFT_NEW|OGS_PFCP_MODIFY_TFT_ADD|
OGS_PFCP_MODIFY_TFT_REPLACE|OGS_PFCP_MODIFY_TFT_DELETE|
OGS_PFCP_MODIFY_QOS_MODIFY)) {
smf_n1_n2_message_transfer_param_t param;
uint8_t qos_rule_code = 0;
uint8_t qos_flow_description_code = 0;
if (flags & OGS_PFCP_MODIFY_TFT_NEW) {
qos_rule_code = OGS_NAS_QOS_CODE_CREATE_NEW_QOS_RULE;
} else if (flags & OGS_PFCP_MODIFY_TFT_ADD) {
qos_rule_code = OGS_NAS_QOS_CODE_MODIFY_EXISTING_QOS_RULE_AND_ADD_PACKET_FILTERS;
} else if (flags & OGS_PFCP_MODIFY_TFT_REPLACE) {
qos_rule_code = OGS_NAS_QOS_CODE_MODIFY_EXISTING_QOS_RULE_AND_REPLACE_PACKET_FILTERS;
} else if (flags & OGS_PFCP_MODIFY_TFT_DELETE) {
qos_rule_code = OGS_NAS_QOS_CODE_MODIFY_EXISTING_QOS_RULE_AND_DELETE_PACKET_FILTERS;
}
if (flags & OGS_PFCP_MODIFY_QOS_CREATE) {
ogs_assert_if_reached();
} else if (flags & OGS_PFCP_MODIFY_QOS_MODIFY) {
qos_flow_description_code = OGS_NAS_MODIFY_NEW_QOS_FLOW_DESCRIPTION;
} else if (flags & OGS_PFCP_MODIFY_QOS_DELETE) {
ogs_assert_if_reached();
}
if (flags & OGS_PFCP_MODIFY_NETWORK_REQUESTED) {
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
ogs_assert(flags & OGS_PFCP_MODIFY_SESSION);
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
VoNR added but not tested!
2021-11-14 12:07:56 +00:00
memset(&param, 0, sizeof(param));
param.state = SMF_NETWORK_REQUESTED_QOS_FLOW_MODIFICATION;
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
param.n1smbuf = gsm_build_pdu_session_modification_command(
sess,
VoNR added but not tested!
2021-11-14 12:07:56 +00:00
OGS_NAS_PROCEDURE_TRANSACTION_IDENTITY_UNASSIGNED,
qos_rule_code, qos_flow_description_code);
feat: Indirect Data Forwarding in N2-Handover
2021-02-01 04:01:15 +00:00
ogs_assert(param.n1smbuf);
param.n2smbuf =
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
ngap_build_pdu_session_resource_modify_request_transfer(
sess,
(flags & OGS_PFCP_MODIFY_QOS_MODIFY) ? true : false);
feat: Indirect Data Forwarding in N2-Handover
2021-02-01 04:01:15 +00:00
ogs_assert(param.n2smbuf);
feat: Add dedicated QoS flow
2021-01-01 02:07:08 +00:00
feat: Indirect Data Forwarding in N2-Handover
2021-02-01 04:01:15 +00:00
smf_namf_comm_send_n1_n2_message_transfer(sess, &param);
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
VoNR added but not tested!
2021-11-14 12:07:56 +00:00
} else if (flags & OGS_PFCP_MODIFY_UE_REQUESTED) {
ogs_pkbuf_t *n1smbuf = NULL, *n2smbuf = NULL;
ogs_assert(stream);
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
ogs_assert(flags & OGS_PFCP_MODIFY_SESSION);
n1smbuf = gsm_build_pdu_session_modification_command(
sess, sess->pti,
VoNR added but not tested!
2021-11-14 12:07:56 +00:00
qos_rule_code, qos_flow_description_code);
ogs_assert(n1smbuf);
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
n2smbuf = ngap_build_pdu_session_resource_modify_request_transfer(
sess,
VoNR added but not tested!
2021-11-14 12:07:56 +00:00
(flags & OGS_PFCP_MODIFY_QOS_MODIFY) ? true : false);
ogs_assert(n2smbuf);
smf_sbi_send_sm_context_updated_data_n1_n2_message(sess, stream,
n1smbuf, OpenAPI_n2_sm_info_type_PDU_RES_MOD_REQ, n2smbuf);
} else {
ogs_fatal("Unknown flags [0x%llx]", (long long)flags);
ogs_assert_if_reached();
feat: Indirect Data Forwarding in N2-Handover
2021-02-01 04:01:15 +00:00
}
Re-factoring 5GC for solving crash issue
2020-08-03 03:22:41 +00:00
}
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
}
[SMF] Apply 5G Core into GSM (#1498)
2022-05-14 21:27:54 +00:00
int smf_5gc_n4_handle_session_deletion_response(
smf_sess_t *sess, ogs_sbi_stream_t *stream, int trigger,
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
ogs_pfcp_session_deletion_response_t *rsp)
{
Fix the SMF crash bug [#487]
2020-07-04 03:14:48 +00:00
int status = 0;
simple 5GC registration is done
2020-06-22 03:07:14 +00:00
logging improvements (#1348) * [SMF] Fix typo in debug message * [SMF] Add debug messages to follow messages received in N4
2022-02-02 14:23:12 +00:00
ogs_debug("Session Deletion Response [5gc]");
[SMF] Apply 5G Core into GSM (#1498)
2022-05-14 21:27:54 +00:00
ogs_assert(rsp);
Add PDU Session Release [#488]
2020-07-02 05:50:23 +00:00
ogs_assert(trigger);
Fix the SMF crash bug [#487]
2020-07-04 03:14:48 +00:00
status = OGS_SBI_HTTP_STATUS_OK;
if (!sess) {
ogs_warn("No Context");
status = OGS_SBI_HTTP_STATUS_NOT_FOUND;
}
simple 5GC registration is done
2020-06-22 03:07:14 +00:00
if (rsp->cause.presence) {
if (rsp->cause.u8 != OGS_PFCP_CAUSE_REQUEST_ACCEPTED) {
Fix the SMF crash bug [#487]
2020-07-04 03:14:48 +00:00
ogs_warn("PFCP Cause [%d] : Not Accepted", rsp->cause.u8);
status = sbi_status_from_pfcp(rsp->cause.u8);
simple 5GC registration is done
2020-06-22 03:07:14 +00:00
}
} else {
ogs_error("No Cause");
Fix the SMF crash bug [#487]
2020-07-04 03:14:48 +00:00
status = OGS_SBI_HTTP_STATUS_BAD_REQUEST;
}
if (status != OGS_SBI_HTTP_STATUS_OK) {
char *strerror = ogs_msprintf(
"PFCP Cause [%d] : Not Accepted", rsp->cause.u8);
VoNR added but not tested!
2021-11-14 12:07:56 +00:00
if (trigger == OGS_PFCP_DELETE_TRIGGER_UE_REQUESTED ||
trigger == OGS_PFCP_DELETE_TRIGGER_AMF_UPDATE_SM_CONTEXT) {
ogs_assert(stream);
smf_sbi_send_sm_context_update_error(
[SMF] fix crash when BSF is not running [#1149]
2021-08-28 15:02:39 +00:00
stream, status, strerror, NULL, NULL, NULL);
VoNR added but not tested!
2021-11-14 12:07:56 +00:00
} else if (trigger == OGS_PFCP_DELETE_TRIGGER_AMF_RELEASE_SM_CONTEXT) {
ogs_assert(stream);
ogs_assert(true ==
ogs_sbi_server_send_error(
stream, status, NULL, strerror, NULL));
} else if (trigger == OGS_PFCP_DELETE_TRIGGER_PCF_INITIATED) {
/* No stream - Nothing */
} else {
ogs_fatal("Unknown trigger [%d]", trigger);
ogs_assert_if_reached();
}
[SMF] Exception handling in session release(#1155) - UE changes the airplain-mode. However, it does not send PDU session complete message - Added exception handling for PDU session release
2021-09-04 07:14:18 +00:00
ogs_error("%s", strerror);
Fix the SMF crash bug [#487]
2020-07-04 03:14:48 +00:00
ogs_free(strerror);
[SMF] Apply 5G Core into GSM (#1498)
2022-05-14 21:27:54 +00:00
return status;
simple 5GC registration is done
2020-06-22 03:07:14 +00:00
}
Fix the SMF crash bug [#487]
2020-07-04 03:14:48 +00:00
ogs_assert(sess);
[SMF] Apply 5G Core into GSM (#1498)
2022-05-14 21:27:54 +00:00
return status;
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
}
[SMF] Apply 5G Core into GSM (#1498)
2022-05-14 21:27:54 +00:00
/* Returns OGS_PFCP_CAUSE_REQUEST_ACCEPTED on success,
* other cause value on failure */
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
uint8_t smf_epc_n4_handle_session_establishment_response(
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
smf_sess_t *sess, ogs_pfcp_xact_t *xact,
ogs_pfcp_session_establishment_response_t *rsp)
{
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
uint8_t cause_value = OGS_PFCP_CAUSE_REQUEST_ACCEPTED;
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
smf_bearer_t *bearer = NULL;
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
ogs_gtp_xact_t *gtp_xact = NULL;
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
ogs_pfcp_f_seid_t *up_f_seid = NULL;
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
ogs_assert(sess);
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
ogs_assert(xact);
ogs_assert(rsp);
logging improvements (#1348) * [SMF] Fix typo in debug message * [SMF] Add debug messages to follow messages received in N4
2022-02-02 14:23:12 +00:00
ogs_debug("Session Establishment Response [epc]");
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
gtp_xact = xact->assoc_xact;
ogs_assert(gtp_xact);
ogs_pfcp_xact_commit(xact);
if (rsp->up_f_seid.presence == 0) {
ogs_error("No UP F-SEID");
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
cause_value = OGS_PFCP_CAUSE_MANDATORY_IE_MISSING;
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
}
if (rsp->cause.presence) {
if (rsp->cause.u8 != OGS_PFCP_CAUSE_REQUEST_ACCEPTED) {
Fix the SMF crash bug [#487]
2020-07-04 03:14:48 +00:00
ogs_warn("PFCP Cause [%d] : Not Accepted", rsp->cause.u8);
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
cause_value = rsp->cause.u8;
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
}
} else {
ogs_error("No Cause");
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
cause_value = OGS_PFCP_CAUSE_MANDATORY_IE_MISSING;
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
}
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
if (cause_value == OGS_PFCP_CAUSE_REQUEST_ACCEPTED) {
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
int i;
uint8_t offending_ie_value = 0;
[PFCP] F-TEID allocation by SMF (#1466)
2022-04-16 05:08:56 +00:00
ogs_pfcp_pdr_t *pdr = NULL;
ogs_pfcp_far_t *far = NULL;
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
[PFCP] F-TEID allocation by SMF (#1466)
2022-04-16 05:08:56 +00:00
for (i = 0; i < OGS_MAX_NUM_OF_PDR; i++) {
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
pdr = ogs_pfcp_handle_created_pdr(
&sess->pfcp, &rsp->created_pdr[i],
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
&cause_value, &offending_ie_value);
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
if (!pdr)
break;
[PFCP] F-TEID allocation by SMF (#1466)
2022-04-16 05:08:56 +00:00
}
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
[PFCP] F-TEID allocation by SMF (#1466)
2022-04-16 05:08:56 +00:00
ogs_list_for_each(&sess->pfcp.pdr_list, pdr) {
Fixes UE IPv6 BUG (#808)
2021-03-15 01:01:55 +00:00
far = pdr->far;
ogs_assert(far);
if (pdr->src_if == OGS_PFCP_INTERFACE_ACCESS) {
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
smf_bearer_t *pdr_bearer = NULL;
Fixes UE IPv6 BUG (#808)
2021-03-15 01:01:55 +00:00
if (far->dst_if == OGS_PFCP_INTERFACE_CP_FUNCTION)
ogs_pfcp_far_teid_hash_set(far);
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
pdr_bearer = smf_bearer_find_by_pdr_id(sess, pdr->id);
if (pdr_bearer) {
Fixes UE IPv6 BUG (#808)
2021-03-15 01:01:55 +00:00
ogs_assert(sess->pfcp_node);
if (sess->pfcp_node->up_function_features.ftup &&
pdr->f_teid_len) {
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
if (pdr_bearer->pgw_s5u_addr)
ogs_freeaddrinfo(pdr_bearer->pgw_s5u_addr);
if (pdr_bearer->pgw_s5u_addr)
ogs_freeaddrinfo(pdr_bearer->pgw_s5u_addr6);
Fixes UE IPv6 BUG (#808)
2021-03-15 01:01:55 +00:00
Add more memory log
2021-06-06 13:35:46 +00:00
ogs_assert(OGS_OK ==
ogs_pfcp_f_teid_to_sockaddr(
Fixes UE IPv6 BUG (#808)
2021-03-15 01:01:55 +00:00
&pdr->f_teid, pdr->f_teid_len,
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
&pdr_bearer->pgw_s5u_addr,
&pdr_bearer->pgw_s5u_addr6));
pdr_bearer->pgw_s5u_teid = pdr->f_teid.teid;
Fixes UE IPv6 BUG (#808)
2021-03-15 01:01:55 +00:00
}
}
} else if (pdr->src_if == OGS_PFCP_INTERFACE_CP_FUNCTION) {
Add more memory log
2021-06-06 13:35:46 +00:00
ogs_assert(OGS_ERROR != ogs_pfcp_setup_pdr_gtpu_node(pdr));
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
}
}
}
Introduce Gn interface (GTPv1C) Support to PGW (#1351) * [CORE] tlv: Store mode in ogs_tlv_t This allows specifying the format of the IE for each individual IE, hence allowing messages containing IEs formatted in different ways. This is needed in order to support parsing GTPv1-C, since messages contain IEs with different structure (TLV vs TV). Hence, this is a preparation patch to add support for parsing TVs in ogs-tlv.c/.h. * [CORE] tlv: Support parsing msg with both TLV and TV in it IEs of type TV are sometimes used in GTPv1-C. Current tlv parser/builder doesn't provide with ways to parse messages which contain TV formatted IEs. This patch adds the relevant types and ways to encode/decode them. Furthermore, the current parser/builder allows parsing/building messages containing the exact same format in all its IEs. A new parser function is added which allows parsing messages of different types (TV, TLV) mixed in the same message. In order to be able to do so, it uses the general msg_mode passed to it in order to know the general TLV format (in essence, the length of the Tag field, and also the length of the Length field if applicable each IE). Looking up the instance in the TLV description is left undone and hadcoded to 0, since the only user so far requiring this API is GTPv1-C, which has no instances. * [CORE] tlv: Support repeated tag+instance parsing TLV message In GTPv2C, repeated IEs (same tag) are easily differentiated by the Instance byte, which provides info to match different decoded structures. In GTPv1C though, there's no Instance byte, and we still encounter repeated IEs (like GSN Address in Create PDP Context Request). Hence, the TLV decoder needs to be updated to track count of IEs found (identified by tag+instance, where instance is always 0 in GTPv1C) and get the proper description index + offset into the decoded structure. * [GTP]: Move GTPv2-C specifics to its own libgtp subdir This will allow adding GTPv1-C code by the side. Most GTPv2 code is left in this patch as "gtp" instead of renaming it to "gtp2" in order to avoid massive changes. It can be done at a later stage if wanted. * [GTP] Support generating GTPv1-C messages * [SMF] Add Gn interface support This patch introduces GTPv1C support to open5gs-smfd. With it, open5gs-becomes a GGSN too, where SGSN can connect to, hence supporting GERAN and UTRAN networks.
2022-02-18 13:23:45 +00:00
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
if (cause_value != OGS_PFCP_CAUSE_REQUEST_ACCEPTED)
return cause_value;
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
bearer = smf_default_bearer_in_sess(sess);
ogs_assert(bearer);
if (bearer->pgw_s5u_addr == NULL && bearer->pgw_s5u_addr6 == NULL) {
ogs_error("No UP F-TEID");
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
return OGS_PFCP_CAUSE_SESSION_CONTEXT_NOT_FOUND;
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
}
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
/* UP F-SEID */
up_f_seid = rsp->up_f_seid.data;
ogs_assert(up_f_seid);
sess->upf_n4_seid = be64toh(up_f_seid->seid);
[SMF] Inegrate session setup cycle into sess->sm (#1489) It allows for much better control on the lifecycle of the session, and already shows some missing tear down paths in case of errors. It also clarifies the existence of "sess" pointer in several paths. The code also becomes clearer overall, since all the transitions and logic to send next messages are put together. Tear down of the session will be integrated into gsm-sm in a follow-up patch. The 5gc session setup is only partially moved to gsm-sm, and left as an exercise for users wishin to improve 5gc support.
2022-04-14 01:30:58 +00:00
return OGS_PFCP_CAUSE_REQUEST_ACCEPTED;
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
}
Add only one 5GC scenario (call-flow)
2020-06-17 05:22:28 +00:00
void smf_epc_n4_handle_session_modification_response(
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
smf_sess_t *sess, ogs_pfcp_xact_t *xact,
X2 handover with SGW change (#1367, #1459)
2022-05-12 13:52:36 +00:00
ogs_gtp2_message_t *recv_message,
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
ogs_pfcp_session_modification_response_t *rsp)
{
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
int i;
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
smf_bearer_t *bearer = NULL;
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
ogs_gtp_xact_t *gtp_xact = NULL;
uint8_t gtp_pti = OGS_NAS_PROCEDURE_TRANSACTION_IDENTITY_UNASSIGNED;
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
uint8_t gtp_cause = OGS_GTP2_CAUSE_UNDEFINED_VALUE;
Split SGW into SGW-C/SGW-U
2020-08-13 00:31:22 +00:00
uint64_t flags = 0;
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
uint8_t pfcp_cause_value = OGS_PFCP_CAUSE_REQUEST_ACCEPTED;
uint8_t offending_ie_value = 0;
[PFCP] F-TEID allocation by SMF (#1466)
2022-04-16 05:08:56 +00:00
ogs_pfcp_pdr_t *pdr = NULL;
ogs_pfcp_far_t *far = NULL;
OGS_LIST(pdr_to_create_list);
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
ogs_assert(xact);
ogs_assert(rsp);
logging improvements (#1348) * [SMF] Fix typo in debug message * [SMF] Add debug messages to follow messages received in N4
2022-02-02 14:23:12 +00:00
ogs_debug("Session Modification Response [epc]");
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
if (flags & OGS_PFCP_MODIFY_SESSION) {
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
/* If smf_epc_pfcp_send_pdr_modification_request() is called */
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
} else {
/* If smf_epc_pfcp_send_bearer_modification_request() is called */
bearer = xact->data;
ogs_assert(bearer);
}
Split SGW into SGW-C/SGW-U
2020-08-13 00:31:22 +00:00
flags = xact->modify_flags;
ogs_assert(flags);
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
[SMF] Introduce optional Gy interface support (#1479) The use of the Gy interface (SMF acting as CTF towards an OCS node) is mandated through configuration file. Default value "enable: auto" will only make use of it in case a Diameter peer announcing support for the Credit-Control Application is found. Upon subscriber session creation, and after auth check over Gx, the SMF will create a Gy session with the OCS and only after that step the SMF will accept the session back to the subscriber. The OCS may then grant some traffic volumes/time and ask to be notified back with updated measurements. In order to get the measurements, the SMF relies on PFCP URR configured to the UPF through Session Repoort Request messages. When closing the subscriber session, the SMF will also terminate the Gy session at the OCS. So far only some specifics parts of the Gy interface as well as the PFCP side are implemented. Those should be enough to at least have volume/time thresholds granted by the OCS, which then will be able to track subsriber resource use. This patch doesn't implement the OCS side of the Gy interface, that's left as a future exercise. The interface was tested using an OCS emulator implemented in TTCN-3 [1] [1] https://cgit.osmocom.org/osmo-ttcn3-hacks/
2022-04-08 23:26:28 +00:00
/* OGS_PFCP_MODIFY_URR: Modification Response was originally triggered by
PFCP Session Report Request, xact->assoc_xact is not a gtp_xact. No
need to do anything. */
if (!(flags & OGS_PFCP_MODIFY_URR)) {
gtp_xact = xact->assoc_xact;
gtp_pti = xact->gtp_pti;
gtp_cause = xact->gtp_cause;
}
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
[PFCP] F-TEID allocation by SMF (#1466)
2022-04-16 05:08:56 +00:00
ogs_list_copy(&pdr_to_create_list, &xact->pdr_to_create_list);
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
ogs_pfcp_xact_commit(xact);
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
if (!sess) {
ogs_error("No Context");
return;
}
if (rsp->cause.presence) {
if (rsp->cause.u8 != OGS_PFCP_CAUSE_REQUEST_ACCEPTED) {
ogs_error("PFCP Cause [%d] : Not Accepted", rsp->cause.u8);
return;
}
} else {
ogs_error("No Cause");
return;
}
ogs_assert(sess);
pfcp_cause_value = OGS_PFCP_CAUSE_REQUEST_ACCEPTED;
for (i = 0; i < OGS_MAX_NUM_OF_PDR; i++) {
pdr = ogs_pfcp_handle_created_pdr(
&sess->pfcp, &rsp->created_pdr[i],
&pfcp_cause_value, &offending_ie_value);
if (!pdr)
break;
[PFCP] F-TEID allocation by SMF (#1466)
2022-04-16 05:08:56 +00:00
}
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
[PFCP] F-TEID allocation by SMF (#1466)
2022-04-16 05:08:56 +00:00
ogs_list_for_each_entry(&pdr_to_create_list, pdr, to_create_node) {
Fixes UE IPv6 BUG (#808)
2021-03-15 01:01:55 +00:00
far = pdr->far;
ogs_assert(far);
if (pdr->src_if == OGS_PFCP_INTERFACE_ACCESS) {
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
smf_bearer_t *pdr_bearer = NULL;
Fixes UE IPv6 BUG (#808)
2021-03-15 01:01:55 +00:00
if (far->dst_if == OGS_PFCP_INTERFACE_CP_FUNCTION)
ogs_pfcp_far_teid_hash_set(far);
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
pdr_bearer = smf_bearer_find_by_pdr_id(sess, pdr->id);
if (pdr_bearer) {
ogs_assert(sess->pfcp_node);
if (sess->pfcp_node->up_function_features.ftup &&
pdr->f_teid_len) {
if (pdr_bearer->pgw_s5u_addr)
ogs_freeaddrinfo(pdr_bearer->pgw_s5u_addr);
if (pdr_bearer->pgw_s5u_addr)
ogs_freeaddrinfo(pdr_bearer->pgw_s5u_addr6);
ogs_assert(OGS_OK ==
ogs_pfcp_f_teid_to_sockaddr(
&pdr->f_teid, pdr->f_teid_len,
&pdr_bearer->pgw_s5u_addr,
&pdr_bearer->pgw_s5u_addr6));
pdr_bearer->pgw_s5u_teid = pdr->f_teid.teid;
}
Fixes UE IPv6 BUG (#808)
2021-03-15 01:01:55 +00:00
}
} else if (pdr->src_if == OGS_PFCP_INTERFACE_CP_FUNCTION) {
Add more memory log
2021-06-06 13:35:46 +00:00
ogs_assert(OGS_ERROR != ogs_pfcp_setup_pdr_gtpu_node(pdr));
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
}
}
if (pfcp_cause_value != OGS_PFCP_CAUSE_REQUEST_ACCEPTED) {
ogs_error("PFCP Cause [%d] : Not Accepted", pfcp_cause_value);
return;
}
feat: Indirect Data Forwarding in N2-Handover
2021-02-01 04:01:15 +00:00
if (flags & OGS_PFCP_MODIFY_REMOVE) {
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
ogs_assert(bearer);
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
smf_bearer_remove(bearer);
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
feat: Indirect Data Forwarding in N2-Handover
2021-02-01 04:01:15 +00:00
} else if (flags & OGS_PFCP_MODIFY_CREATE) {
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
ogs_assert(bearer);
X2 handover with SGW change (#1367, #1459)
2022-05-12 13:52:36 +00:00
ogs_assert(OGS_OK == smf_gtp2_send_create_bearer_request(bearer));
feat: Indirect Data Forwarding in N2-Handover
2021-02-01 04:01:15 +00:00
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
} else if (flags & OGS_PFCP_MODIFY_DEACTIVATE) {
/*
* TS23.214
* 6.3.1.7 Procedures with modification of bearer
* p50
* 2. ...
* For "PGW/MME initiated bearer deactivation procedure",
* PGW-C shall indicate PGW-U to stop counting and stop
* forwarding downlink packets for the affected bearer(s).
*/
if (gtp_xact) {
/*
* 1. MME sends Bearer Resource Command to SGW/SMF.
* 2. SMF sends Delete Bearer Request(DEDICATED BEARER) to SGW/MME
* 3. MME sends Delete Bearer Response(DEDICATED BEARER) to SGW/SMF
*/
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
ogs_gtp2_header_t h;
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
ogs_pkbuf_t *pkbuf = NULL;
int rv;
ogs_assert(bearer);
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
memset(&h, 0, sizeof(ogs_gtp2_header_t));
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
h.teid = sess->sgw_s5c_teid;
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
h.type = OGS_GTP2_DELETE_BEARER_REQUEST_TYPE;
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
pkbuf = smf_s5c_build_delete_bearer_request(
h.type, bearer, gtp_pti, gtp_cause);
ogs_expect_or_return(pkbuf);
rv = ogs_gtp_xact_update_tx(gtp_xact, &h, pkbuf);
ogs_expect_or_return(rv == OGS_OK);
/* IMPORTANT:
*
* When initiaited by Bearer Resource Command,
* there must be bearer context in the Transaction.
* Otherwise, the bearer context cannot be found
* in GTP response message.
*
* For example,
* 1. MME sends Bearer Resource Command to SGW-C, SMF.
* 2. SMF sends Update/Delete Bearer Request to the SGW-C, MME.
* 3. MME sends Update/Delete Bearer Response to thw SGW-C, SMF.
*
* On number 3 step, if MME sends Response without Bearer Context,
* we need a way to find Bearer context.
*
* To do this, I saved Bearer Context in Transaction Context.
*/
gtp_xact->data = bearer;
rv = ogs_gtp_xact_commit(gtp_xact);
ogs_expect(rv == OGS_OK);
} else {
if (flags & OGS_PFCP_MODIFY_SESSION) {
/*
* 1. SMF sends Delete Bearer Request(DEFAULT BEARER) to SGW/MME.
* 2. MME sends Delete Bearer Response to SGW/SMF.
*
* OR
*
* 1. SMF sends Delete Bearer Request(DEFAULT BEARER) to ePDG.
* 2. ePDG sends Delete Bearer Response(DEFAULT BEARER) to SMF.
*/
smf_bearer_t *linked_bearer =
ogs_list_first(&sess->bearer_list);
ogs_assert(linked_bearer);
ogs_assert(OGS_OK ==
X2 handover with SGW change (#1367, #1459)
2022-05-12 13:52:36 +00:00
smf_gtp2_send_delete_bearer_request(
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
linked_bearer, gtp_pti, gtp_cause));
} else {
/*
[GTP2] Fixed handling multi-bearer messages(#1498)
2022-04-29 12:28:16 +00:00
* 1. RX : Session-Termination Request
* 2. GX : Re-Auth-Request(Charging-Rule-Remove)
* 3. RX : Session-Termination Answer
* 4. PFCP Session Modification Request(Buffering)
* 5. PFCP Session Modification Answer(Buffering)
* 6. SMF sends Delete Bearer Request(DEDICATED BEARER) to SGW/MME.
* 7. E-RABReleaseCommand + Deactivate EPS bearer context request
* 8. E-RABReleaseResponse
* 9. UplinkNASTransport + Deactivate EPS bearer context accept
* 10. MME sends Delete Bearer Response(DEDICATED BEARER) to SGW-C
* 11. PFCP Session Modification Request/Response(Remove PDR/FAR)
* 12. SGW-C sends Delete Bearer Response(DECIATED BEARER) to SMF
* 13. PFCP Session Modification Request/Response(Remove PDR/FAR)
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
*/
ogs_assert(bearer);
ogs_assert(OGS_OK ==
X2 handover with SGW change (#1367, #1459)
2022-05-12 13:52:36 +00:00
smf_gtp2_send_delete_bearer_request(
[SMF] fix the crash during IMS calls (#1238) If PFCP is not deactivated, some eNBs send an GTP error indication. To solve the problem, SMF deactivates the bearer that will be deleted.
2021-11-23 05:19:30 +00:00
bearer,
OGS_NAS_PROCEDURE_TRANSACTION_IDENTITY_UNASSIGNED,
gtp_cause));
}
}
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
} else if (flags & OGS_PFCP_MODIFY_ACTIVATE) {
[SMF] fix the crash in VoLTE test (#1383)
2022-02-23 11:31:28 +00:00
if (gtp_xact) {
X2 handover with SGW change (#1367, #1459)
2022-05-12 13:52:36 +00:00
[SMF] fix the crash in VoLTE test (#1383)
2022-02-23 11:31:28 +00:00
/* SMF send Update PDP Context Response (GTPv1C) to SGSN */
if (gtp_xact->gtp_version == 1) {
X2 handover with SGW change (#1367, #1459)
2022-05-12 13:52:36 +00:00
[SMF] fix the crash in VoLTE test (#1383)
2022-02-23 11:31:28 +00:00
bearer = gtp_xact->data;
smf_gtp1_send_update_pdp_context_response(bearer, gtp_xact);
X2 handover with SGW change (#1367, #1459)
2022-05-12 13:52:36 +00:00
[SMF] fix the crash in VoLTE test (#1383)
2022-02-23 11:31:28 +00:00
} else {
X2 handover with SGW change (#1367, #1459)
2022-05-12 13:52:36 +00:00
/* SMF send Modify Bearer Response (GTPv2C) to SGW-C */
ogs_gtp2_indication_t *indication = NULL;
ogs_assert(recv_message);
ogs_gtp2_modify_bearer_request_t *gtp_req =
&recv_message->modify_bearer_request;
ogs_assert(OGS_OK == smf_gtp2_send_modify_bearer_response(
sess, gtp_xact, gtp_req, true));
/* Check if Handover from Non-3GPP to 3GPP */
if (gtp_req->indication_flags.presence &&
gtp_req->indication_flags.data &&
gtp_req->indication_flags.len) {
indication = gtp_req->indication_flags.data;
}
if (indication && indication->handover_indication) {
ogs_assert(OGS_OK == smf_epc_pfcp_send_deactivation(sess,
OGS_GTP2_CAUSE_ACCESS_CHANGED_FROM_NON_3GPP_TO_3GPP));
}
[SMF] fix the crash in VoLTE test (#1383)
2022-02-23 11:31:28 +00:00
}
[SMF] UpdatePDPContext: forward update of remote TEID+IPaddr to UPF (#1383) * [SMF] typo fixes in commented code * [SMF] Fix early err return handling UpdatePDPContextRequest * [SMF] UpdatePDPContext: forward update of remote TEID+IPaddr to UPF Updating the remote GTP-U IP address and/or TEID on the GGSN is a common practice, used for instance by an SGSN in a UTRAN network to connect an HNB(GW) to exchange GTP-U directly with the GGSN. It is also used in general when doing handovers. When receiving a UpdatePDPContext with the new address, we need to forward the update to the UPF so that it takes it into account when forwarding packets. This patch only implements updating the information towards the UPF when GTPv1C is used. Similar approach for GTPv2C (upon receival of Modify Bearer Request) is still unimplemented. Related: https://github.com/open5gs/open5gs/issues/1367
2022-02-23 11:21:33 +00:00
} else {
[SMF] fix the crash in VoLTE test (#1383)
2022-02-23 11:31:28 +00:00
/* Nothing */
[SMF] UpdatePDPContext: forward update of remote TEID+IPaddr to UPF (#1383) * [SMF] typo fixes in commented code * [SMF] Fix early err return handling UpdatePDPContextRequest * [SMF] UpdatePDPContext: forward update of remote TEID+IPaddr to UPF Updating the remote GTP-U IP address and/or TEID on the GGSN is a common practice, used for instance by an SGSN in a UTRAN network to connect an HNB(GW) to exchange GTP-U directly with the GGSN. It is also used in general when doing handovers. When receiving a UpdatePDPContext with the new address, we need to forward the update to the UPF so that it takes it into account when forwarding packets. This patch only implements updating the information towards the UPF when GTPv1C is used. Similar approach for GTPv2C (upon receival of Modify Bearer Request) is still unimplemented. Related: https://github.com/open5gs/open5gs/issues/1367
2022-02-23 11:21:33 +00:00
}
Now, F-TEID is allocated in the UP function.
2020-10-26 02:43:53 +00:00
}
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
}
[SMF] Integrate session tear down cycle into sess->sm (#1500) * [SMF] smf_sm.c Fix indentation * [SMF] gsm-sm: log fsm events * [SMF] Integrate session tear down cycle into sess->sm
2022-04-20 12:42:18 +00:00
uint8_t smf_epc_n4_handle_session_deletion_response(
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
smf_sess_t *sess, ogs_pfcp_xact_t *xact,
ogs_pfcp_session_deletion_response_t *rsp)
{
[SMF] Introduce optional Gy interface support (#1479) The use of the Gy interface (SMF acting as CTF towards an OCS node) is mandated through configuration file. Default value "enable: auto" will only make use of it in case a Diameter peer announcing support for the Credit-Control Application is found. Upon subscriber session creation, and after auth check over Gx, the SMF will create a Gy session with the OCS and only after that step the SMF will accept the session back to the subscriber. The OCS may then grant some traffic volumes/time and ask to be notified back with updated measurements. In order to get the measurements, the SMF relies on PFCP URR configured to the UPF through Session Repoort Request messages. When closing the subscriber session, the SMF will also terminate the Gy session at the OCS. So far only some specifics parts of the Gy interface as well as the PFCP side are implemented. Those should be enough to at least have volume/time thresholds granted by the OCS, which then will be able to track subsriber resource use. This patch doesn't implement the OCS side of the Gy interface, that's left as a future exercise. The interface was tested using an OCS emulator implemented in TTCN-3 [1] [1] https://cgit.osmocom.org/osmo-ttcn3-hacks/
2022-04-08 23:26:28 +00:00
smf_bearer_t *bearer = NULL;
unsigned int i;
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
[SMF] Integrate session tear down cycle into sess->sm (#1500) * [SMF] smf_sm.c Fix indentation * [SMF] gsm-sm: log fsm events * [SMF] Integrate session tear down cycle into sess->sm
2022-04-20 12:42:18 +00:00
ogs_assert(sess);
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
ogs_assert(xact);
ogs_assert(rsp);
logging improvements (#1348) * [SMF] Fix typo in debug message * [SMF] Add debug messages to follow messages received in N4
2022-02-02 14:23:12 +00:00
ogs_debug("Session Deletion Response [epc]");
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
ogs_pfcp_xact_commit(xact);
[SMF] Integrate session tear down cycle into sess->sm (#1500) * [SMF] smf_sm.c Fix indentation * [SMF] gsm-sm: log fsm events * [SMF] Integrate session tear down cycle into sess->sm
2022-04-20 12:42:18 +00:00
if (!rsp->cause.presence) {
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
ogs_error("No Cause");
[SMF] Integrate session tear down cycle into sess->sm (#1500) * [SMF] smf_sm.c Fix indentation * [SMF] gsm-sm: log fsm events * [SMF] Integrate session tear down cycle into sess->sm
2022-04-20 12:42:18 +00:00
return OGS_PFCP_CAUSE_MANDATORY_IE_MISSING;
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
}
[SMF] Integrate session tear down cycle into sess->sm (#1500) * [SMF] smf_sm.c Fix indentation * [SMF] gsm-sm: log fsm events * [SMF] Integrate session tear down cycle into sess->sm
2022-04-20 12:42:18 +00:00
if (rsp->cause.u8 != OGS_PFCP_CAUSE_REQUEST_ACCEPTED) {
ogs_warn("PFCP Cause[%d] : Not Accepted", rsp->cause.u8);
return rsp->cause.u8;
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
}
ogs_assert(sess);
[SMF] Introduce optional Gy interface support (#1479) The use of the Gy interface (SMF acting as CTF towards an OCS node) is mandated through configuration file. Default value "enable: auto" will only make use of it in case a Diameter peer announcing support for the Credit-Control Application is found. Upon subscriber session creation, and after auth check over Gx, the SMF will create a Gy session with the OCS and only after that step the SMF will accept the session back to the subscriber. The OCS may then grant some traffic volumes/time and ask to be notified back with updated measurements. In order to get the measurements, the SMF relies on PFCP URR configured to the UPF through Session Repoort Request messages. When closing the subscriber session, the SMF will also terminate the Gy session at the OCS. So far only some specifics parts of the Gy interface as well as the PFCP side are implemented. Those should be enough to at least have volume/time thresholds granted by the OCS, which then will be able to track subsriber resource use. This patch doesn't implement the OCS side of the Gy interface, that's left as a future exercise. The interface was tested using an OCS emulator implemented in TTCN-3 [1] [1] https://cgit.osmocom.org/osmo-ttcn3-hacks/
2022-04-08 23:26:28 +00:00
bearer = smf_default_bearer_in_sess(sess);
for (i = 0; i < OGS_ARRAY_SIZE(rsp->usage_report); i++) {
[SMF] Apply 5G Core into GSM (#1498)
2022-05-14 21:27:54 +00:00
ogs_pfcp_tlv_usage_report_session_deletion_response_t *use_rep =
&rsp->usage_report[i];
[SMF] Introduce optional Gy interface support (#1479) The use of the Gy interface (SMF acting as CTF towards an OCS node) is mandated through configuration file. Default value "enable: auto" will only make use of it in case a Diameter peer announcing support for the Credit-Control Application is found. Upon subscriber session creation, and after auth check over Gx, the SMF will create a Gy session with the OCS and only after that step the SMF will accept the session back to the subscriber. The OCS may then grant some traffic volumes/time and ask to be notified back with updated measurements. In order to get the measurements, the SMF relies on PFCP URR configured to the UPF through Session Repoort Request messages. When closing the subscriber session, the SMF will also terminate the Gy session at the OCS. So far only some specifics parts of the Gy interface as well as the PFCP side are implemented. Those should be enough to at least have volume/time thresholds granted by the OCS, which then will be able to track subsriber resource use. This patch doesn't implement the OCS side of the Gy interface, that's left as a future exercise. The interface was tested using an OCS emulator implemented in TTCN-3 [1] [1] https://cgit.osmocom.org/osmo-ttcn3-hacks/
2022-04-08 23:26:28 +00:00
uint32_t urr_id;
ogs_pfcp_volume_measurement_t volume;
if (use_rep->presence == 0)
break;
if (use_rep->urr_id.presence == 0)
continue;
urr_id = use_rep->urr_id.u32;
if (!bearer || !bearer->urr || bearer->urr->id != urr_id)
continue;
[SMF] Apply 5G Core into GSM (#1498)
2022-05-14 21:27:54 +00:00
ogs_pfcp_parse_volume_measurement(
&volume, &use_rep->volume_measurement);
[SMF] Introduce optional Gy interface support (#1479) The use of the Gy interface (SMF acting as CTF towards an OCS node) is mandated through configuration file. Default value "enable: auto" will only make use of it in case a Diameter peer announcing support for the Credit-Control Application is found. Upon subscriber session creation, and after auth check over Gx, the SMF will create a Gy session with the OCS and only after that step the SMF will accept the session back to the subscriber. The OCS may then grant some traffic volumes/time and ask to be notified back with updated measurements. In order to get the measurements, the SMF relies on PFCP URR configured to the UPF through Session Repoort Request messages. When closing the subscriber session, the SMF will also terminate the Gy session at the OCS. So far only some specifics parts of the Gy interface as well as the PFCP side are implemented. Those should be enough to at least have volume/time thresholds granted by the OCS, which then will be able to track subsriber resource use. This patch doesn't implement the OCS side of the Gy interface, that's left as a future exercise. The interface was tested using an OCS emulator implemented in TTCN-3 [1] [1] https://cgit.osmocom.org/osmo-ttcn3-hacks/
2022-04-08 23:26:28 +00:00
if (volume.ulvol)
sess->gy.ul_octets += volume.uplink_volume;
if (volume.dlvol)
sess->gy.dl_octets += volume.downlink_volume;
sess->gy.duration += use_rep->duration_measurement.u32;
}
[SMF] Integrate session tear down cycle into sess->sm (#1500) * [SMF] smf_sm.c Fix indentation * [SMF] gsm-sm: log fsm events * [SMF] Integrate session tear down cycle into sess->sm
2022-04-20 12:42:18 +00:00
return OGS_PFCP_CAUSE_REQUEST_ACCEPTED;
Initial proposal CUPS using PFCP
2020-04-26 19:36:05 +00:00
}
5gc: Paging was added
2021-01-18 16:48:35 +00:00
void smf_n4_handle_session_report_request(
smf_sess_t *sess, ogs_pfcp_xact_t *pfcp_xact,
ogs_pfcp_session_report_request_t *pfcp_req)
{
smf_bearer_t *qos_flow = NULL;
[SMF] Introduce optional Gy interface support (#1479) The use of the Gy interface (SMF acting as CTF towards an OCS node) is mandated through configuration file. Default value "enable: auto" will only make use of it in case a Diameter peer announcing support for the Credit-Control Application is found. Upon subscriber session creation, and after auth check over Gx, the SMF will create a Gy session with the OCS and only after that step the SMF will accept the session back to the subscriber. The OCS may then grant some traffic volumes/time and ask to be notified back with updated measurements. In order to get the measurements, the SMF relies on PFCP URR configured to the UPF through Session Repoort Request messages. When closing the subscriber session, the SMF will also terminate the Gy session at the OCS. So far only some specifics parts of the Gy interface as well as the PFCP side are implemented. Those should be enough to at least have volume/time thresholds granted by the OCS, which then will be able to track subsriber resource use. This patch doesn't implement the OCS side of the Gy interface, that's left as a future exercise. The interface was tested using an OCS emulator implemented in TTCN-3 [1] [1] https://cgit.osmocom.org/osmo-ttcn3-hacks/
2022-04-08 23:26:28 +00:00
smf_bearer_t *bearer = NULL;
5gc: Paging was added
2021-01-18 16:48:35 +00:00
ogs_pfcp_pdr_t *pdr = NULL;
ogs_pfcp_report_type_t report_type;
uint8_t cause_value = 0;
uint16_t pdr_id = 0;
[SMF] Introduce optional Gy interface support (#1479) The use of the Gy interface (SMF acting as CTF towards an OCS node) is mandated through configuration file. Default value "enable: auto" will only make use of it in case a Diameter peer announcing support for the Credit-Control Application is found. Upon subscriber session creation, and after auth check over Gx, the SMF will create a Gy session with the OCS and only after that step the SMF will accept the session back to the subscriber. The OCS may then grant some traffic volumes/time and ask to be notified back with updated measurements. In order to get the measurements, the SMF relies on PFCP URR configured to the UPF through Session Repoort Request messages. When closing the subscriber session, the SMF will also terminate the Gy session at the OCS. So far only some specifics parts of the Gy interface as well as the PFCP side are implemented. Those should be enough to at least have volume/time thresholds granted by the OCS, which then will be able to track subsriber resource use. This patch doesn't implement the OCS side of the Gy interface, that's left as a future exercise. The interface was tested using an OCS emulator implemented in TTCN-3 [1] [1] https://cgit.osmocom.org/osmo-ttcn3-hacks/
2022-04-08 23:26:28 +00:00
unsigned int i;
5gc: Paging was added
2021-01-18 16:48:35 +00:00
ogs_assert(pfcp_xact);
ogs_assert(pfcp_req);
logging improvements (#1348) * [SMF] Fix typo in debug message * [SMF] Add debug messages to follow messages received in N4
2022-02-02 14:23:12 +00:00
ogs_debug("Session Report Request");
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
cause_value = OGS_GTP2_CAUSE_REQUEST_ACCEPTED;
5gc: Paging was added
2021-01-18 16:48:35 +00:00
if (!sess) {
ogs_warn("No Context");
cause_value = OGS_PFCP_CAUSE_SESSION_CONTEXT_NOT_FOUND;
}
if (pfcp_req->report_type.presence == 0) {
ogs_error("No Report Type");
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
cause_value = OGS_GTP2_CAUSE_MANDATORY_IE_MISSING;
5gc: Paging was added
2021-01-18 16:48:35 +00:00
}
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
if (cause_value != OGS_GTP2_CAUSE_REQUEST_ACCEPTED) {
5gc: Paging was added
2021-01-18 16:48:35 +00:00
ogs_pfcp_send_error_message(pfcp_xact, 0,
OGS_PFCP_SESSION_REPORT_RESPONSE_TYPE,
cause_value, 0);
return;
}
ogs_assert(sess);
report_type.value = pfcp_req->report_type.u8;
if (report_type.downlink_data_report) {
ogs_pfcp_downlink_data_service_information_t *info = NULL;
uint8_t paging_policy_indication_value = 0;
uint8_t qfi = 0;
if (pfcp_req->downlink_data_report.presence) {
if (pfcp_req->downlink_data_report.
downlink_data_service_information.presence) {
info = pfcp_req->downlink_data_report.
downlink_data_service_information.data;
if (info) {
if (info->qfii && info->ppi) {
paging_policy_indication_value =
info->paging_policy_indication_value;
qfi = info->qfi;
} else if (info->qfii) {
qfi = info->qfi;
} else if (info->ppi) {
paging_policy_indication_value =
info->paging_policy_indication_value;
} else {
ogs_error("Invalid Downlink Data Service Information");
}
if (paging_policy_indication_value) {
ogs_warn("Not implement - "
"Paging Policy Indication Value");
ogs_pfcp_send_error_message(pfcp_xact, 0,
OGS_PFCP_SESSION_REPORT_RESPONSE_TYPE,
Rename GTPv2C specifics to gtp2 prefix (#1485) In the past only GTPv2C was supported, and had the "gtp" generic prefix. Later on, GTPv1C support was added, and "gtp1" prefix was used. Let's move GTPv2C specific bits to have "gtp2" prefix too, and leave "gtp" prefix for generic stuff among different GTP versions.
2022-04-12 22:07:39 +00:00
OGS_GTP2_CAUSE_SERVICE_NOT_SUPPORTED, 0);
5gc: Paging was added
2021-01-18 16:48:35 +00:00
return;
}
if (qfi) {
qos_flow = smf_qos_flow_find_by_qfi(sess, qfi);
if (!qos_flow)
ogs_error("Cannot find the QoS Flow[%d]", qfi);
}
} else {
ogs_error("No Info");
}
}
if (pfcp_req->downlink_data_report.pdr_id.presence) {
pdr = ogs_pfcp_pdr_find(&sess->pfcp,
pfcp_req->downlink_data_report.pdr_id.u16);
if (!pdr)
ogs_error("Cannot find the PDR-ID[%d]", pdr_id);
} else {
ogs_error("No PDR-ID");
}
} else {
ogs_error("No Downlink Data Report");
}
if (!pdr || !qos_flow) {
ogs_error("No Context [%p:%p]", pdr, qos_flow);
ogs_pfcp_send_error_message(pfcp_xact, 0,
OGS_PFCP_SESSION_REPORT_RESPONSE_TYPE,
cause_value, 0);
[SMF/UPF] fix the crash bug (#1034, #1006)
2021-06-08 12:14:18 +00:00
return;
5gc: Paging was added
2021-01-18 16:48:35 +00:00
}
if (sess->paging.ue_requested_pdu_session_establishment_done == true) {
smf_n1_n2_message_transfer_param_t param;
memset(&param, 0, sizeof(param));
param.state = SMF_NETWORK_TRIGGERED_SERVICE_REQUEST;
param.n2smbuf =
[5GC] Added BSF(Binding Support Function)
2021-05-29 06:56:12 +00:00
ngap_build_pdu_session_resource_setup_request_transfer(sess);
5gc: Paging was added
2021-01-18 16:48:35 +00:00
ogs_assert(param.n2smbuf);
param.n1n2_failure_txf_notif_uri = true;
smf_namf_comm_send_n1_n2_message_transfer(sess, &param);
}
[SMF] PFCP Session Report Req: Allow receiving multiple bits in Report Type (#1442) 3GPP TS 29.244 sec 8.2.21 "Report Type" states: At least one bit shall be set to "1". Several bits may be set to "1".
2022-03-17 07:23:43 +00:00
}
5gc: Paging was added
2021-01-18 16:48:35 +00:00
[SMF] PFCP Session Report Req: Allow receiving multiple bits in Report Type (#1442) 3GPP TS 29.244 sec 8.2.21 "Report Type" states: At least one bit shall be set to "1". Several bits may be set to "1".
2022-03-17 07:23:43 +00:00
if (report_type.error_indication_report) {
5gc: Paging was added
2021-01-18 16:48:35 +00:00
smf_ue_t *smf_ue = sess->smf_ue;
smf_sess_t *error_indication_session = NULL;
ogs_assert(smf_ue);
error_indication_session = smf_sess_find_by_error_indication_report(
smf_ue, &pfcp_req->error_indication_report);
[SMF] PFCP Session Report Req: Allow receiving multiple bits in Report Type (#1442) 3GPP TS 29.244 sec 8.2.21 "Report Type" states: At least one bit shall be set to "1". Several bits may be set to "1".
2022-03-17 07:23:43 +00:00
if (error_indication_session) {
ogs_assert(OGS_OK ==
[SMF] Optimiza Session Modification Message For video in VoNR, multiple QosFlow is required in one session. In the past, since only one Session Modification Message was supported, QoS-Flow was put into several Session Messages one by one and processed. Now that multiple QoS-Flows can be supported, it is optimized to process one session modification message.
2022-05-15 14:35:41 +00:00
smf_5gc_pfcp_send_pdr_modification_request(
[SMF] PFCP Session Report Req: Allow receiving multiple bits in Report Type (#1442) 3GPP TS 29.244 sec 8.2.21 "Report Type" states: At least one bit shall be set to "1". Several bits may be set to "1".
2022-03-17 07:23:43 +00:00
error_indication_session, NULL,
OGS_PFCP_MODIFY_DL_ONLY|OGS_PFCP_MODIFY_DEACTIVATE|
OGS_PFCP_MODIFY_ERROR_INDICATION,
0));
}
}
5gc: Paging was added
2021-01-18 16:48:35 +00:00
[SMF] Introduce optional Gy interface support (#1479) The use of the Gy interface (SMF acting as CTF towards an OCS node) is mandated through configuration file. Default value "enable: auto" will only make use of it in case a Diameter peer announcing support for the Credit-Control Application is found. Upon subscriber session creation, and after auth check over Gx, the SMF will create a Gy session with the OCS and only after that step the SMF will accept the session back to the subscriber. The OCS may then grant some traffic volumes/time and ask to be notified back with updated measurements. In order to get the measurements, the SMF relies on PFCP URR configured to the UPF through Session Repoort Request messages. When closing the subscriber session, the SMF will also terminate the Gy session at the OCS. So far only some specifics parts of the Gy interface as well as the PFCP side are implemented. Those should be enough to at least have volume/time thresholds granted by the OCS, which then will be able to track subsriber resource use. This patch doesn't implement the OCS side of the Gy interface, that's left as a future exercise. The interface was tested using an OCS emulator implemented in TTCN-3 [1] [1] https://cgit.osmocom.org/osmo-ttcn3-hacks/
2022-04-08 23:26:28 +00:00
if (report_type.usage_report) {
bearer = smf_default_bearer_in_sess(sess);
for (i = 0; i < OGS_ARRAY_SIZE(pfcp_req->usage_report); i++) {
[SMF] Apply 5G Core into GSM (#1498)
2022-05-14 21:27:54 +00:00
ogs_pfcp_tlv_usage_report_session_report_request_t *use_rep =
&pfcp_req->usage_report[i];
[SMF] Introduce optional Gy interface support (#1479) The use of the Gy interface (SMF acting as CTF towards an OCS node) is mandated through configuration file. Default value "enable: auto" will only make use of it in case a Diameter peer announcing support for the Credit-Control Application is found. Upon subscriber session creation, and after auth check over Gx, the SMF will create a Gy session with the OCS and only after that step the SMF will accept the session back to the subscriber. The OCS may then grant some traffic volumes/time and ask to be notified back with updated measurements. In order to get the measurements, the SMF relies on PFCP URR configured to the UPF through Session Repoort Request messages. When closing the subscriber session, the SMF will also terminate the Gy session at the OCS. So far only some specifics parts of the Gy interface as well as the PFCP side are implemented. Those should be enough to at least have volume/time thresholds granted by the OCS, which then will be able to track subsriber resource use. This patch doesn't implement the OCS side of the Gy interface, that's left as a future exercise. The interface was tested using an OCS emulator implemented in TTCN-3 [1] [1] https://cgit.osmocom.org/osmo-ttcn3-hacks/
2022-04-08 23:26:28 +00:00
uint32_t urr_id;
ogs_pfcp_volume_measurement_t volume;
if (use_rep->presence == 0)
break;
if (use_rep->urr_id.presence == 0)
continue;
urr_id = use_rep->urr_id.u32;
if (!bearer || !bearer->urr || bearer->urr->id != urr_id)
continue;
[SMF] Apply 5G Core into GSM (#1498)
2022-05-14 21:27:54 +00:00
ogs_pfcp_parse_volume_measurement(
&volume, &use_rep->volume_measurement);
[SMF] Introduce optional Gy interface support (#1479) The use of the Gy interface (SMF acting as CTF towards an OCS node) is mandated through configuration file. Default value "enable: auto" will only make use of it in case a Diameter peer announcing support for the Credit-Control Application is found. Upon subscriber session creation, and after auth check over Gx, the SMF will create a Gy session with the OCS and only after that step the SMF will accept the session back to the subscriber. The OCS may then grant some traffic volumes/time and ask to be notified back with updated measurements. In order to get the measurements, the SMF relies on PFCP URR configured to the UPF through Session Repoort Request messages. When closing the subscriber session, the SMF will also terminate the Gy session at the OCS. So far only some specifics parts of the Gy interface as well as the PFCP side are implemented. Those should be enough to at least have volume/time thresholds granted by the OCS, which then will be able to track subsriber resource use. This patch doesn't implement the OCS side of the Gy interface, that's left as a future exercise. The interface was tested using an OCS emulator implemented in TTCN-3 [1] [1] https://cgit.osmocom.org/osmo-ttcn3-hacks/
2022-04-08 23:26:28 +00:00
if (volume.ulvol)
sess->gy.ul_octets += volume.uplink_volume;
if (volume.dlvol)
sess->gy.dl_octets += volume.downlink_volume;
sess->gy.duration += use_rep->duration_measurement.u32;
}
switch(smf_use_gy_iface()) {
case 1:
[SMF] Apply 5G Core into GSM (#1498)
2022-05-14 21:27:54 +00:00
smf_gy_send_ccr(sess, pfcp_xact,
OGS_DIAM_GY_CC_REQUEST_TYPE_UPDATE_REQUEST);
[SMF] Introduce optional Gy interface support (#1479) The use of the Gy interface (SMF acting as CTF towards an OCS node) is mandated through configuration file. Default value "enable: auto" will only make use of it in case a Diameter peer announcing support for the Credit-Control Application is found. Upon subscriber session creation, and after auth check over Gx, the SMF will create a Gy session with the OCS and only after that step the SMF will accept the session back to the subscriber. The OCS may then grant some traffic volumes/time and ask to be notified back with updated measurements. In order to get the measurements, the SMF relies on PFCP URR configured to the UPF through Session Repoort Request messages. When closing the subscriber session, the SMF will also terminate the Gy session at the OCS. So far only some specifics parts of the Gy interface as well as the PFCP side are implemented. Those should be enough to at least have volume/time thresholds granted by the OCS, which then will be able to track subsriber resource use. This patch doesn't implement the OCS side of the Gy interface, that's left as a future exercise. The interface was tested using an OCS emulator implemented in TTCN-3 [1] [1] https://cgit.osmocom.org/osmo-ttcn3-hacks/
2022-04-08 23:26:28 +00:00
break;
case -1:
ogs_error("No Gy Diameter Peer");
/* TODO: terminate connection */
break;
/* default: continue below */
}
}
[SMF] PFCP Session Report Req: Allow receiving multiple bits in Report Type (#1442) 3GPP TS 29.244 sec 8.2.21 "Report Type" states: At least one bit shall be set to "1". Several bits may be set to "1".
2022-03-17 07:23:43 +00:00
/* TS 29.244 sec 8.2.21: At least one bit shall be set to "1". Several bits may be set to "1". */
[SMF] Introduce optional Gy interface support (#1479) The use of the Gy interface (SMF acting as CTF towards an OCS node) is mandated through configuration file. Default value "enable: auto" will only make use of it in case a Diameter peer announcing support for the Credit-Control Application is found. Upon subscriber session creation, and after auth check over Gx, the SMF will create a Gy session with the OCS and only after that step the SMF will accept the session back to the subscriber. The OCS may then grant some traffic volumes/time and ask to be notified back with updated measurements. In order to get the measurements, the SMF relies on PFCP URR configured to the UPF through Session Repoort Request messages. When closing the subscriber session, the SMF will also terminate the Gy session at the OCS. So far only some specifics parts of the Gy interface as well as the PFCP side are implemented. Those should be enough to at least have volume/time thresholds granted by the OCS, which then will be able to track subsriber resource use. This patch doesn't implement the OCS side of the Gy interface, that's left as a future exercise. The interface was tested using an OCS emulator implemented in TTCN-3 [1] [1] https://cgit.osmocom.org/osmo-ttcn3-hacks/
2022-04-08 23:26:28 +00:00
if (report_type.downlink_data_report ||
report_type.error_indication_report ||
report_type.usage_report) {
[ALL] Add protection of message encoding (#991) All process will be forcely exited if it failed to encode the S1AP/NGAP/GTP/PFCP message. It is to make sure there was no problem with the encoding of open5gs.
2021-05-16 03:22:10 +00:00
ogs_assert(OGS_OK ==
[SMF] PFCP Session Report Req: Allow receiving multiple bits in Report Type (#1442) 3GPP TS 29.244 sec 8.2.21 "Report Type" states: At least one bit shall be set to "1". Several bits may be set to "1".
2022-03-17 07:23:43 +00:00
smf_pfcp_send_session_report_response(
pfcp_xact, sess, OGS_PFCP_CAUSE_REQUEST_ACCEPTED));
5gc: Paging was added
2021-01-18 16:48:35 +00:00
} else {
ogs_error("Not supported Report Type[%d]", report_type.value);
[ALL] Add protection of message encoding (#991) All process will be forcely exited if it failed to encode the S1AP/NGAP/GTP/PFCP message. It is to make sure there was no problem with the encoding of open5gs.
2021-05-16 03:22:10 +00:00
ogs_assert(OGS_OK ==
smf_pfcp_send_session_report_response(
pfcp_xact, sess, OGS_PFCP_CAUSE_SYSTEM_FAILURE));
5gc: Paging was added
2021-01-18 16:48:35 +00:00
}
}