Stir/Shaken Refactor
Why do we need a refactor?
The original stir/shaken implementation was started over 3 years ago
when little was understood about practical implementation. The
result was an implementation that wouldn't actually interoperate
with any other stir-shaken implementations.
There were also a number of stir-shaken features and RFC
requirements that were never implemented such as TNAuthList
certificate validation, sending Reason headers in SIP responses
when verification failed but we wished to continue the call, and
the ability to send Media Key(mky) grants in the Identity header
when the call involved DTLS.
Finally, there were some performance concerns around outgoing
calls and selection of the correct certificate and private key.
The configuration was keyed by an arbitrary name which meant that
for every outgoing call, we had to scan the entire list of
configured TNs to find the correct cert to use. With only a few
TNs configured, this wasn't an issue but if you have a thousand,
it could be.
What's changed?
* Configuration objects have been refactored to be clearer about
their uses and to fix issues.
* The "general" object was renamed to "verification" since it
contains parameters specific to the incoming verification
process. It also never handled ca_path and crl_path
correctly.
* A new "attestation" object was added that controls the
outgoing attestation process. It sets default certificates,
keys, etc.
* The "certificate" object was renamed to "tn" and had it's key
change to telephone number since outgoing call attestation
needs to look up certificates by telephone number.
* The "profile" object had more parameters added to it that can
override default parameters specified in the "attestation"
and "verification" objects.
* The "store" object was removed altogther as it was never
implemented.
* We now use libjwt to create outgoing Identity headers and to
parse and validate signatures on incoming Identiy headers. Our
previous custom implementation was much of the source of the
interoperability issues.
* General code cleanup and refactor.
* Moved things to better places.
* Separated some of the complex functions to smaller ones.
* Using context objects rather than passing tons of parameters
in function calls.
* Removed some complexity and unneeded encapsuation from the
config objects.
Resolves: #351
Resolves: #46
UserNote: Asterisk's stir-shaken feature has been refactored to
correct interoperability, RFC compliance, and performance issues.
See https://docs.asterisk.org/Deployment/STIR-SHAKEN for more
information.
UpgradeNote: The stir-shaken refactor is a breaking change but since
it's not working now we don't think it matters. The
stir_shaken.conf file has changed significantly which means that
existing ones WILL need to be changed. The stir_shaken.conf.sample
file in configs/samples/ has quite a bit more information. This is
also an ABI breaking change since some of the existing objects
needed to be changed or removed, and new ones added. Additionally,
if res_stir_shaken is enabled in menuselect, you'll need to either
have the development package for libjwt v1.15.3 installed or use
the --with-libjwt-bundled option with ./configure.
2023-10-26 10:27:35 -06:00
|
|
|
/*
|
|
|
|
|
* Asterisk -- An open source telephony toolkit.
|
|
|
|
|
*
|
|
|
|
|
* Copyright (C) 2023, Sangoma Technologies Corporation
|
|
|
|
|
*
|
|
|
|
|
* George Joseph <gjoseph@sangoma.com>
|
|
|
|
|
*
|
|
|
|
|
* See http://www.asterisk.org for more information about
|
|
|
|
|
* the Asterisk project. Please do not directly contact
|
|
|
|
|
* any of the maintainers of this project for assistance;
|
|
|
|
|
* the project provides a web site, mailing lists and IRC
|
|
|
|
|
* channels for your use.
|
|
|
|
|
*
|
|
|
|
|
* This program is free software, distributed under the terms of
|
|
|
|
|
* the GNU General Public License Version 2. See the LICENSE file
|
|
|
|
|
* at the top of the source tree.
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
#include <jwt.h>
|
|
|
|
|
|
|
|
|
|
#define _TRACE_PREFIX_ "a",__LINE__, ""
|
|
|
|
|
|
|
|
|
|
#include "asterisk.h"
|
|
|
|
|
#include "asterisk/module.h"
|
|
|
|
|
#include "asterisk/uuid.h"
|
|
|
|
|
#include "asterisk/json.h"
|
|
|
|
|
#include "asterisk/channel.h"
|
|
|
|
|
|
|
|
|
|
#include "stir_shaken.h"
|
|
|
|
|
|
|
|
|
|
static const char *as_rc_map[] = {
|
|
|
|
|
[AST_STIR_SHAKEN_AS_SUCCESS] = "success",
|
|
|
|
|
[AST_STIR_SHAKEN_AS_DISABLED] = "disabled",
|
|
|
|
|
[AST_STIR_SHAKEN_AS_INVALID_ARGUMENTS] = "invalid_arguments",
|
|
|
|
|
[AST_STIR_SHAKEN_AS_MISSING_PARAMETERS] = "missing_parameters",
|
|
|
|
|
[AST_STIR_SHAKEN_AS_INTERNAL_ERROR] = "internal_error",
|
|
|
|
|
[AST_STIR_SHAKEN_AS_NO_TN_FOR_CALLERID] = "no_tn_for_callerid",
|
|
|
|
|
[AST_STIR_SHAKEN_AS_NO_PRIVATE_KEY_AVAIL] = "no_private_key_avail",
|
|
|
|
|
[AST_STIR_SHAKEN_AS_NO_PUBLIC_CERT_URL_AVAIL] = "no_public_cert_url_avail",
|
|
|
|
|
[AST_STIR_SHAKEN_AS_NO_ATTEST_LEVEL] = "no_attest_level",
|
|
|
|
|
[AST_STIR_SHAKEN_AS_IDENTITY_HDR_EXISTS] = "identity_header_exists",
|
|
|
|
|
[AST_STIR_SHAKEN_AS_NO_TO_HDR] = "no_to_hdr",
|
|
|
|
|
[AST_STIR_SHAKEN_AS_TO_HDR_BAD_URI] = "to_hdr_bad_uri",
|
|
|
|
|
[AST_STIR_SHAKEN_AS_SIGN_ENCODE_FAILURE] "sign_encode_failure",
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
const char *as_response_code_to_str(
|
|
|
|
|
enum ast_stir_shaken_as_response_code as_rc)
|
|
|
|
|
{
|
|
|
|
|
return ARRAY_IN_BOUNDS(as_rc, as_rc_map) ?
|
|
|
|
|
as_rc_map[as_rc] : NULL;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static void ctx_destructor(void *obj)
|
|
|
|
|
{
|
|
|
|
|
struct ast_stir_shaken_as_ctx *ctx = obj;
|
|
|
|
|
|
|
|
|
|
ao2_cleanup(ctx->etn);
|
|
|
|
|
ast_channel_cleanup(ctx->chan);
|
|
|
|
|
ast_string_field_free_memory(ctx);
|
|
|
|
|
AST_VECTOR_RESET(&ctx->fingerprints, ast_free);
|
|
|
|
|
AST_VECTOR_FREE(&ctx->fingerprints);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
enum ast_stir_shaken_as_response_code
|
|
|
|
|
ast_stir_shaken_as_ctx_create(const char *orig_tn,
|
|
|
|
|
const char *dest_tn, struct ast_channel *chan,
|
|
|
|
|
const char *profile_name,
|
|
|
|
|
const char *tag, struct ast_stir_shaken_as_ctx **ctxout)
|
|
|
|
|
{
|
|
|
|
|
RAII_VAR(struct ast_stir_shaken_as_ctx *, ctx, NULL, ao2_cleanup);
|
|
|
|
|
RAII_VAR(struct profile_cfg *, eprofile, NULL, ao2_cleanup);
|
|
|
|
|
RAII_VAR(struct attestation_cfg *, as_cfg, NULL, ao2_cleanup);
|
|
|
|
|
RAII_VAR(struct tn_cfg *, etn, NULL, ao2_cleanup);
|
2024-04-25 11:56:15 -06:00
|
|
|
RAII_VAR(char *, canon_dest_tn , canonicalize_tn_alloc(dest_tn), ast_free);
|
|
|
|
|
RAII_VAR(char *, canon_orig_tn , canonicalize_tn_alloc(orig_tn), ast_free);
|
|
|
|
|
|
2024-09-11 10:19:23 -06:00
|
|
|
const char *t = S_OR(tag, S_COR(chan, ast_channel_name(chan), ""));
|
|
|
|
|
SCOPE_ENTER(3, "%s: Enter\n", t);
|
Stir/Shaken Refactor
Why do we need a refactor?
The original stir/shaken implementation was started over 3 years ago
when little was understood about practical implementation. The
result was an implementation that wouldn't actually interoperate
with any other stir-shaken implementations.
There were also a number of stir-shaken features and RFC
requirements that were never implemented such as TNAuthList
certificate validation, sending Reason headers in SIP responses
when verification failed but we wished to continue the call, and
the ability to send Media Key(mky) grants in the Identity header
when the call involved DTLS.
Finally, there were some performance concerns around outgoing
calls and selection of the correct certificate and private key.
The configuration was keyed by an arbitrary name which meant that
for every outgoing call, we had to scan the entire list of
configured TNs to find the correct cert to use. With only a few
TNs configured, this wasn't an issue but if you have a thousand,
it could be.
What's changed?
* Configuration objects have been refactored to be clearer about
their uses and to fix issues.
* The "general" object was renamed to "verification" since it
contains parameters specific to the incoming verification
process. It also never handled ca_path and crl_path
correctly.
* A new "attestation" object was added that controls the
outgoing attestation process. It sets default certificates,
keys, etc.
* The "certificate" object was renamed to "tn" and had it's key
change to telephone number since outgoing call attestation
needs to look up certificates by telephone number.
* The "profile" object had more parameters added to it that can
override default parameters specified in the "attestation"
and "verification" objects.
* The "store" object was removed altogther as it was never
implemented.
* We now use libjwt to create outgoing Identity headers and to
parse and validate signatures on incoming Identiy headers. Our
previous custom implementation was much of the source of the
interoperability issues.
* General code cleanup and refactor.
* Moved things to better places.
* Separated some of the complex functions to smaller ones.
* Using context objects rather than passing tons of parameters
in function calls.
* Removed some complexity and unneeded encapsuation from the
config objects.
Resolves: #351
Resolves: #46
UserNote: Asterisk's stir-shaken feature has been refactored to
correct interoperability, RFC compliance, and performance issues.
See https://docs.asterisk.org/Deployment/STIR-SHAKEN for more
information.
UpgradeNote: The stir-shaken refactor is a breaking change but since
it's not working now we don't think it matters. The
stir_shaken.conf file has changed significantly which means that
existing ones WILL need to be changed. The stir_shaken.conf.sample
file in configs/samples/ has quite a bit more information. This is
also an ABI breaking change since some of the existing objects
needed to be changed or removed, and new ones added. Additionally,
if res_stir_shaken is enabled in menuselect, you'll need to either
have the development package for libjwt v1.15.3 installed or use
the --with-libjwt-bundled option with ./configure.
2023-10-26 10:27:35 -06:00
|
|
|
|
2024-09-11 10:19:23 -06:00
|
|
|
as_cfg = as_get_cfg();
|
|
|
|
|
if (as_cfg->global_disable) {
|
Stir/Shaken Refactor
Why do we need a refactor?
The original stir/shaken implementation was started over 3 years ago
when little was understood about practical implementation. The
result was an implementation that wouldn't actually interoperate
with any other stir-shaken implementations.
There were also a number of stir-shaken features and RFC
requirements that were never implemented such as TNAuthList
certificate validation, sending Reason headers in SIP responses
when verification failed but we wished to continue the call, and
the ability to send Media Key(mky) grants in the Identity header
when the call involved DTLS.
Finally, there were some performance concerns around outgoing
calls and selection of the correct certificate and private key.
The configuration was keyed by an arbitrary name which meant that
for every outgoing call, we had to scan the entire list of
configured TNs to find the correct cert to use. With only a few
TNs configured, this wasn't an issue but if you have a thousand,
it could be.
What's changed?
* Configuration objects have been refactored to be clearer about
their uses and to fix issues.
* The "general" object was renamed to "verification" since it
contains parameters specific to the incoming verification
process. It also never handled ca_path and crl_path
correctly.
* A new "attestation" object was added that controls the
outgoing attestation process. It sets default certificates,
keys, etc.
* The "certificate" object was renamed to "tn" and had it's key
change to telephone number since outgoing call attestation
needs to look up certificates by telephone number.
* The "profile" object had more parameters added to it that can
override default parameters specified in the "attestation"
and "verification" objects.
* The "store" object was removed altogther as it was never
implemented.
* We now use libjwt to create outgoing Identity headers and to
parse and validate signatures on incoming Identiy headers. Our
previous custom implementation was much of the source of the
interoperability issues.
* General code cleanup and refactor.
* Moved things to better places.
* Separated some of the complex functions to smaller ones.
* Using context objects rather than passing tons of parameters
in function calls.
* Removed some complexity and unneeded encapsuation from the
config objects.
Resolves: #351
Resolves: #46
UserNote: Asterisk's stir-shaken feature has been refactored to
correct interoperability, RFC compliance, and performance issues.
See https://docs.asterisk.org/Deployment/STIR-SHAKEN for more
information.
UpgradeNote: The stir-shaken refactor is a breaking change but since
it's not working now we don't think it matters. The
stir_shaken.conf file has changed significantly which means that
existing ones WILL need to be changed. The stir_shaken.conf.sample
file in configs/samples/ has quite a bit more information. This is
also an ABI breaking change since some of the existing objects
needed to be changed or removed, and new ones added. Additionally,
if res_stir_shaken is enabled in menuselect, you'll need to either
have the development package for libjwt v1.15.3 installed or use
the --with-libjwt-bundled option with ./configure.
2023-10-26 10:27:35 -06:00
|
|
|
SCOPE_EXIT_RTN_VALUE(AST_STIR_SHAKEN_AS_DISABLED,
|
2024-09-11 10:19:23 -06:00
|
|
|
"%s: Globally disabled\n", t);
|
Stir/Shaken Refactor
Why do we need a refactor?
The original stir/shaken implementation was started over 3 years ago
when little was understood about practical implementation. The
result was an implementation that wouldn't actually interoperate
with any other stir-shaken implementations.
There were also a number of stir-shaken features and RFC
requirements that were never implemented such as TNAuthList
certificate validation, sending Reason headers in SIP responses
when verification failed but we wished to continue the call, and
the ability to send Media Key(mky) grants in the Identity header
when the call involved DTLS.
Finally, there were some performance concerns around outgoing
calls and selection of the correct certificate and private key.
The configuration was keyed by an arbitrary name which meant that
for every outgoing call, we had to scan the entire list of
configured TNs to find the correct cert to use. With only a few
TNs configured, this wasn't an issue but if you have a thousand,
it could be.
What's changed?
* Configuration objects have been refactored to be clearer about
their uses and to fix issues.
* The "general" object was renamed to "verification" since it
contains parameters specific to the incoming verification
process. It also never handled ca_path and crl_path
correctly.
* A new "attestation" object was added that controls the
outgoing attestation process. It sets default certificates,
keys, etc.
* The "certificate" object was renamed to "tn" and had it's key
change to telephone number since outgoing call attestation
needs to look up certificates by telephone number.
* The "profile" object had more parameters added to it that can
override default parameters specified in the "attestation"
and "verification" objects.
* The "store" object was removed altogther as it was never
implemented.
* We now use libjwt to create outgoing Identity headers and to
parse and validate signatures on incoming Identiy headers. Our
previous custom implementation was much of the source of the
interoperability issues.
* General code cleanup and refactor.
* Moved things to better places.
* Separated some of the complex functions to smaller ones.
* Using context objects rather than passing tons of parameters
in function calls.
* Removed some complexity and unneeded encapsuation from the
config objects.
Resolves: #351
Resolves: #46
UserNote: Asterisk's stir-shaken feature has been refactored to
correct interoperability, RFC compliance, and performance issues.
See https://docs.asterisk.org/Deployment/STIR-SHAKEN for more
information.
UpgradeNote: The stir-shaken refactor is a breaking change but since
it's not working now we don't think it matters. The
stir_shaken.conf file has changed significantly which means that
existing ones WILL need to be changed. The stir_shaken.conf.sample
file in configs/samples/ has quite a bit more information. This is
also an ABI breaking change since some of the existing objects
needed to be changed or removed, and new ones added. Additionally,
if res_stir_shaken is enabled in menuselect, you'll need to either
have the development package for libjwt v1.15.3 installed or use
the --with-libjwt-bundled option with ./configure.
2023-10-26 10:27:35 -06:00
|
|
|
}
|
|
|
|
|
|
2024-09-11 10:19:23 -06:00
|
|
|
if (ast_strlen_zero(profile_name)) {
|
Stir/Shaken Refactor
Why do we need a refactor?
The original stir/shaken implementation was started over 3 years ago
when little was understood about practical implementation. The
result was an implementation that wouldn't actually interoperate
with any other stir-shaken implementations.
There were also a number of stir-shaken features and RFC
requirements that were never implemented such as TNAuthList
certificate validation, sending Reason headers in SIP responses
when verification failed but we wished to continue the call, and
the ability to send Media Key(mky) grants in the Identity header
when the call involved DTLS.
Finally, there were some performance concerns around outgoing
calls and selection of the correct certificate and private key.
The configuration was keyed by an arbitrary name which meant that
for every outgoing call, we had to scan the entire list of
configured TNs to find the correct cert to use. With only a few
TNs configured, this wasn't an issue but if you have a thousand,
it could be.
What's changed?
* Configuration objects have been refactored to be clearer about
their uses and to fix issues.
* The "general" object was renamed to "verification" since it
contains parameters specific to the incoming verification
process. It also never handled ca_path and crl_path
correctly.
* A new "attestation" object was added that controls the
outgoing attestation process. It sets default certificates,
keys, etc.
* The "certificate" object was renamed to "tn" and had it's key
change to telephone number since outgoing call attestation
needs to look up certificates by telephone number.
* The "profile" object had more parameters added to it that can
override default parameters specified in the "attestation"
and "verification" objects.
* The "store" object was removed altogther as it was never
implemented.
* We now use libjwt to create outgoing Identity headers and to
parse and validate signatures on incoming Identiy headers. Our
previous custom implementation was much of the source of the
interoperability issues.
* General code cleanup and refactor.
* Moved things to better places.
* Separated some of the complex functions to smaller ones.
* Using context objects rather than passing tons of parameters
in function calls.
* Removed some complexity and unneeded encapsuation from the
config objects.
Resolves: #351
Resolves: #46
UserNote: Asterisk's stir-shaken feature has been refactored to
correct interoperability, RFC compliance, and performance issues.
See https://docs.asterisk.org/Deployment/STIR-SHAKEN for more
information.
UpgradeNote: The stir-shaken refactor is a breaking change but since
it's not working now we don't think it matters. The
stir_shaken.conf file has changed significantly which means that
existing ones WILL need to be changed. The stir_shaken.conf.sample
file in configs/samples/ has quite a bit more information. This is
also an ABI breaking change since some of the existing objects
needed to be changed or removed, and new ones added. Additionally,
if res_stir_shaken is enabled in menuselect, you'll need to either
have the development package for libjwt v1.15.3 installed or use
the --with-libjwt-bundled option with ./configure.
2023-10-26 10:27:35 -06:00
|
|
|
SCOPE_EXIT_RTN_VALUE(AST_STIR_SHAKEN_AS_DISABLED,
|
2024-09-11 10:19:23 -06:00
|
|
|
"%s: Disabled due to missing profile name\n", t);
|
Stir/Shaken Refactor
Why do we need a refactor?
The original stir/shaken implementation was started over 3 years ago
when little was understood about practical implementation. The
result was an implementation that wouldn't actually interoperate
with any other stir-shaken implementations.
There were also a number of stir-shaken features and RFC
requirements that were never implemented such as TNAuthList
certificate validation, sending Reason headers in SIP responses
when verification failed but we wished to continue the call, and
the ability to send Media Key(mky) grants in the Identity header
when the call involved DTLS.
Finally, there were some performance concerns around outgoing
calls and selection of the correct certificate and private key.
The configuration was keyed by an arbitrary name which meant that
for every outgoing call, we had to scan the entire list of
configured TNs to find the correct cert to use. With only a few
TNs configured, this wasn't an issue but if you have a thousand,
it could be.
What's changed?
* Configuration objects have been refactored to be clearer about
their uses and to fix issues.
* The "general" object was renamed to "verification" since it
contains parameters specific to the incoming verification
process. It also never handled ca_path and crl_path
correctly.
* A new "attestation" object was added that controls the
outgoing attestation process. It sets default certificates,
keys, etc.
* The "certificate" object was renamed to "tn" and had it's key
change to telephone number since outgoing call attestation
needs to look up certificates by telephone number.
* The "profile" object had more parameters added to it that can
override default parameters specified in the "attestation"
and "verification" objects.
* The "store" object was removed altogther as it was never
implemented.
* We now use libjwt to create outgoing Identity headers and to
parse and validate signatures on incoming Identiy headers. Our
previous custom implementation was much of the source of the
interoperability issues.
* General code cleanup and refactor.
* Moved things to better places.
* Separated some of the complex functions to smaller ones.
* Using context objects rather than passing tons of parameters
in function calls.
* Removed some complexity and unneeded encapsuation from the
config objects.
Resolves: #351
Resolves: #46
UserNote: Asterisk's stir-shaken feature has been refactored to
correct interoperability, RFC compliance, and performance issues.
See https://docs.asterisk.org/Deployment/STIR-SHAKEN for more
information.
UpgradeNote: The stir-shaken refactor is a breaking change but since
it's not working now we don't think it matters. The
stir_shaken.conf file has changed significantly which means that
existing ones WILL need to be changed. The stir_shaken.conf.sample
file in configs/samples/ has quite a bit more information. This is
also an ABI breaking change since some of the existing objects
needed to be changed or removed, and new ones added. Additionally,
if res_stir_shaken is enabled in menuselect, you'll need to either
have the development package for libjwt v1.15.3 installed or use
the --with-libjwt-bundled option with ./configure.
2023-10-26 10:27:35 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
eprofile = eprofile_get_cfg(profile_name);
|
|
|
|
|
if (!eprofile) {
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_DISABLED,
|
|
|
|
|
LOG_ERROR, "%s: No profile for profile name '%s'. Call will continue\n", tag,
|
|
|
|
|
profile_name);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (!PROFILE_ALLOW_ATTEST(eprofile)) {
|
|
|
|
|
SCOPE_EXIT_RTN_VALUE(AST_STIR_SHAKEN_AS_DISABLED,
|
2024-09-11 10:19:23 -06:00
|
|
|
"%s: Disabled by profile '%s'\n", t, profile_name);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (ast_strlen_zero(tag)) {
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_INVALID_ARGUMENTS,
|
|
|
|
|
LOG_ERROR, "%s: Must provide tag\n", t);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (!canon_orig_tn) {
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_INVALID_ARGUMENTS,
|
|
|
|
|
LOG_ERROR, "%s: Must provide caller_id/orig_tn\n", tag);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (!canon_dest_tn) {
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_INVALID_ARGUMENTS,
|
|
|
|
|
LOG_ERROR, "%s: Must provide dest_tn\n", tag);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (!ctxout) {
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_INVALID_ARGUMENTS,
|
|
|
|
|
LOG_ERROR, "%s: Must provide ctxout\n", tag);
|
Stir/Shaken Refactor
Why do we need a refactor?
The original stir/shaken implementation was started over 3 years ago
when little was understood about practical implementation. The
result was an implementation that wouldn't actually interoperate
with any other stir-shaken implementations.
There were also a number of stir-shaken features and RFC
requirements that were never implemented such as TNAuthList
certificate validation, sending Reason headers in SIP responses
when verification failed but we wished to continue the call, and
the ability to send Media Key(mky) grants in the Identity header
when the call involved DTLS.
Finally, there were some performance concerns around outgoing
calls and selection of the correct certificate and private key.
The configuration was keyed by an arbitrary name which meant that
for every outgoing call, we had to scan the entire list of
configured TNs to find the correct cert to use. With only a few
TNs configured, this wasn't an issue but if you have a thousand,
it could be.
What's changed?
* Configuration objects have been refactored to be clearer about
their uses and to fix issues.
* The "general" object was renamed to "verification" since it
contains parameters specific to the incoming verification
process. It also never handled ca_path and crl_path
correctly.
* A new "attestation" object was added that controls the
outgoing attestation process. It sets default certificates,
keys, etc.
* The "certificate" object was renamed to "tn" and had it's key
change to telephone number since outgoing call attestation
needs to look up certificates by telephone number.
* The "profile" object had more parameters added to it that can
override default parameters specified in the "attestation"
and "verification" objects.
* The "store" object was removed altogther as it was never
implemented.
* We now use libjwt to create outgoing Identity headers and to
parse and validate signatures on incoming Identiy headers. Our
previous custom implementation was much of the source of the
interoperability issues.
* General code cleanup and refactor.
* Moved things to better places.
* Separated some of the complex functions to smaller ones.
* Using context objects rather than passing tons of parameters
in function calls.
* Removed some complexity and unneeded encapsuation from the
config objects.
Resolves: #351
Resolves: #46
UserNote: Asterisk's stir-shaken feature has been refactored to
correct interoperability, RFC compliance, and performance issues.
See https://docs.asterisk.org/Deployment/STIR-SHAKEN for more
information.
UpgradeNote: The stir-shaken refactor is a breaking change but since
it's not working now we don't think it matters. The
stir_shaken.conf file has changed significantly which means that
existing ones WILL need to be changed. The stir_shaken.conf.sample
file in configs/samples/ has quite a bit more information. This is
also an ABI breaking change since some of the existing objects
needed to be changed or removed, and new ones added. Additionally,
if res_stir_shaken is enabled in menuselect, you'll need to either
have the development package for libjwt v1.15.3 installed or use
the --with-libjwt-bundled option with ./configure.
2023-10-26 10:27:35 -06:00
|
|
|
}
|
|
|
|
|
|
2024-04-25 11:56:15 -06:00
|
|
|
etn = tn_get_etn(canon_orig_tn, eprofile);
|
Stir/Shaken Refactor
Why do we need a refactor?
The original stir/shaken implementation was started over 3 years ago
when little was understood about practical implementation. The
result was an implementation that wouldn't actually interoperate
with any other stir-shaken implementations.
There were also a number of stir-shaken features and RFC
requirements that were never implemented such as TNAuthList
certificate validation, sending Reason headers in SIP responses
when verification failed but we wished to continue the call, and
the ability to send Media Key(mky) grants in the Identity header
when the call involved DTLS.
Finally, there were some performance concerns around outgoing
calls and selection of the correct certificate and private key.
The configuration was keyed by an arbitrary name which meant that
for every outgoing call, we had to scan the entire list of
configured TNs to find the correct cert to use. With only a few
TNs configured, this wasn't an issue but if you have a thousand,
it could be.
What's changed?
* Configuration objects have been refactored to be clearer about
their uses and to fix issues.
* The "general" object was renamed to "verification" since it
contains parameters specific to the incoming verification
process. It also never handled ca_path and crl_path
correctly.
* A new "attestation" object was added that controls the
outgoing attestation process. It sets default certificates,
keys, etc.
* The "certificate" object was renamed to "tn" and had it's key
change to telephone number since outgoing call attestation
needs to look up certificates by telephone number.
* The "profile" object had more parameters added to it that can
override default parameters specified in the "attestation"
and "verification" objects.
* The "store" object was removed altogther as it was never
implemented.
* We now use libjwt to create outgoing Identity headers and to
parse and validate signatures on incoming Identiy headers. Our
previous custom implementation was much of the source of the
interoperability issues.
* General code cleanup and refactor.
* Moved things to better places.
* Separated some of the complex functions to smaller ones.
* Using context objects rather than passing tons of parameters
in function calls.
* Removed some complexity and unneeded encapsuation from the
config objects.
Resolves: #351
Resolves: #46
UserNote: Asterisk's stir-shaken feature has been refactored to
correct interoperability, RFC compliance, and performance issues.
See https://docs.asterisk.org/Deployment/STIR-SHAKEN for more
information.
UpgradeNote: The stir-shaken refactor is a breaking change but since
it's not working now we don't think it matters. The
stir_shaken.conf file has changed significantly which means that
existing ones WILL need to be changed. The stir_shaken.conf.sample
file in configs/samples/ has quite a bit more information. This is
also an ABI breaking change since some of the existing objects
needed to be changed or removed, and new ones added. Additionally,
if res_stir_shaken is enabled in menuselect, you'll need to either
have the development package for libjwt v1.15.3 installed or use
the --with-libjwt-bundled option with ./configure.
2023-10-26 10:27:35 -06:00
|
|
|
if (!etn) {
|
|
|
|
|
SCOPE_EXIT_RTN_VALUE(AST_STIR_SHAKEN_AS_DISABLED,
|
2024-04-25 11:56:15 -06:00
|
|
|
"%s: No tn for orig_tn '%s'\n", tag, canon_orig_tn);
|
Stir/Shaken Refactor
Why do we need a refactor?
The original stir/shaken implementation was started over 3 years ago
when little was understood about practical implementation. The
result was an implementation that wouldn't actually interoperate
with any other stir-shaken implementations.
There were also a number of stir-shaken features and RFC
requirements that were never implemented such as TNAuthList
certificate validation, sending Reason headers in SIP responses
when verification failed but we wished to continue the call, and
the ability to send Media Key(mky) grants in the Identity header
when the call involved DTLS.
Finally, there were some performance concerns around outgoing
calls and selection of the correct certificate and private key.
The configuration was keyed by an arbitrary name which meant that
for every outgoing call, we had to scan the entire list of
configured TNs to find the correct cert to use. With only a few
TNs configured, this wasn't an issue but if you have a thousand,
it could be.
What's changed?
* Configuration objects have been refactored to be clearer about
their uses and to fix issues.
* The "general" object was renamed to "verification" since it
contains parameters specific to the incoming verification
process. It also never handled ca_path and crl_path
correctly.
* A new "attestation" object was added that controls the
outgoing attestation process. It sets default certificates,
keys, etc.
* The "certificate" object was renamed to "tn" and had it's key
change to telephone number since outgoing call attestation
needs to look up certificates by telephone number.
* The "profile" object had more parameters added to it that can
override default parameters specified in the "attestation"
and "verification" objects.
* The "store" object was removed altogther as it was never
implemented.
* We now use libjwt to create outgoing Identity headers and to
parse and validate signatures on incoming Identiy headers. Our
previous custom implementation was much of the source of the
interoperability issues.
* General code cleanup and refactor.
* Moved things to better places.
* Separated some of the complex functions to smaller ones.
* Using context objects rather than passing tons of parameters
in function calls.
* Removed some complexity and unneeded encapsuation from the
config objects.
Resolves: #351
Resolves: #46
UserNote: Asterisk's stir-shaken feature has been refactored to
correct interoperability, RFC compliance, and performance issues.
See https://docs.asterisk.org/Deployment/STIR-SHAKEN for more
information.
UpgradeNote: The stir-shaken refactor is a breaking change but since
it's not working now we don't think it matters. The
stir_shaken.conf file has changed significantly which means that
existing ones WILL need to be changed. The stir_shaken.conf.sample
file in configs/samples/ has quite a bit more information. This is
also an ABI breaking change since some of the existing objects
needed to be changed or removed, and new ones added. Additionally,
if res_stir_shaken is enabled in menuselect, you'll need to either
have the development package for libjwt v1.15.3 installed or use
the --with-libjwt-bundled option with ./configure.
2023-10-26 10:27:35 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* We don't need eprofile or as_cfg anymore so let's clean em up */
|
|
|
|
|
ao2_cleanup(as_cfg);
|
|
|
|
|
as_cfg = NULL;
|
|
|
|
|
ao2_cleanup(eprofile);
|
|
|
|
|
eprofile = NULL;
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
if (etn->acfg_common.attest_level == attest_level_NOT_SET) {
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_MISSING_PARAMETERS,
|
|
|
|
|
LOG_ERROR,
|
|
|
|
|
"'%s': No attest_level specified in tn, profile or attestation objects\n",
|
|
|
|
|
tag);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (ast_strlen_zero(etn->acfg_common.public_cert_url)) {
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_NO_PUBLIC_CERT_URL_AVAIL,
|
|
|
|
|
LOG_ERROR, "%s: No public cert url in tn %s, profile or attestation objects\n",
|
2024-04-25 11:56:15 -06:00
|
|
|
tag, canon_orig_tn);
|
Stir/Shaken Refactor
Why do we need a refactor?
The original stir/shaken implementation was started over 3 years ago
when little was understood about practical implementation. The
result was an implementation that wouldn't actually interoperate
with any other stir-shaken implementations.
There were also a number of stir-shaken features and RFC
requirements that were never implemented such as TNAuthList
certificate validation, sending Reason headers in SIP responses
when verification failed but we wished to continue the call, and
the ability to send Media Key(mky) grants in the Identity header
when the call involved DTLS.
Finally, there were some performance concerns around outgoing
calls and selection of the correct certificate and private key.
The configuration was keyed by an arbitrary name which meant that
for every outgoing call, we had to scan the entire list of
configured TNs to find the correct cert to use. With only a few
TNs configured, this wasn't an issue but if you have a thousand,
it could be.
What's changed?
* Configuration objects have been refactored to be clearer about
their uses and to fix issues.
* The "general" object was renamed to "verification" since it
contains parameters specific to the incoming verification
process. It also never handled ca_path and crl_path
correctly.
* A new "attestation" object was added that controls the
outgoing attestation process. It sets default certificates,
keys, etc.
* The "certificate" object was renamed to "tn" and had it's key
change to telephone number since outgoing call attestation
needs to look up certificates by telephone number.
* The "profile" object had more parameters added to it that can
override default parameters specified in the "attestation"
and "verification" objects.
* The "store" object was removed altogther as it was never
implemented.
* We now use libjwt to create outgoing Identity headers and to
parse and validate signatures on incoming Identiy headers. Our
previous custom implementation was much of the source of the
interoperability issues.
* General code cleanup and refactor.
* Moved things to better places.
* Separated some of the complex functions to smaller ones.
* Using context objects rather than passing tons of parameters
in function calls.
* Removed some complexity and unneeded encapsuation from the
config objects.
Resolves: #351
Resolves: #46
UserNote: Asterisk's stir-shaken feature has been refactored to
correct interoperability, RFC compliance, and performance issues.
See https://docs.asterisk.org/Deployment/STIR-SHAKEN for more
information.
UpgradeNote: The stir-shaken refactor is a breaking change but since
it's not working now we don't think it matters. The
stir_shaken.conf file has changed significantly which means that
existing ones WILL need to be changed. The stir_shaken.conf.sample
file in configs/samples/ has quite a bit more information. This is
also an ABI breaking change since some of the existing objects
needed to be changed or removed, and new ones added. Additionally,
if res_stir_shaken is enabled in menuselect, you'll need to either
have the development package for libjwt v1.15.3 installed or use
the --with-libjwt-bundled option with ./configure.
2023-10-26 10:27:35 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (etn->acfg_common.raw_key_length == 0) {
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_NO_PRIVATE_KEY_AVAIL,
|
|
|
|
|
LOG_ERROR, "%s: No private key in tn %s, profile or attestation objects\n",
|
2024-04-25 11:56:15 -06:00
|
|
|
canon_orig_tn, tag);
|
Stir/Shaken Refactor
Why do we need a refactor?
The original stir/shaken implementation was started over 3 years ago
when little was understood about practical implementation. The
result was an implementation that wouldn't actually interoperate
with any other stir-shaken implementations.
There were also a number of stir-shaken features and RFC
requirements that were never implemented such as TNAuthList
certificate validation, sending Reason headers in SIP responses
when verification failed but we wished to continue the call, and
the ability to send Media Key(mky) grants in the Identity header
when the call involved DTLS.
Finally, there were some performance concerns around outgoing
calls and selection of the correct certificate and private key.
The configuration was keyed by an arbitrary name which meant that
for every outgoing call, we had to scan the entire list of
configured TNs to find the correct cert to use. With only a few
TNs configured, this wasn't an issue but if you have a thousand,
it could be.
What's changed?
* Configuration objects have been refactored to be clearer about
their uses and to fix issues.
* The "general" object was renamed to "verification" since it
contains parameters specific to the incoming verification
process. It also never handled ca_path and crl_path
correctly.
* A new "attestation" object was added that controls the
outgoing attestation process. It sets default certificates,
keys, etc.
* The "certificate" object was renamed to "tn" and had it's key
change to telephone number since outgoing call attestation
needs to look up certificates by telephone number.
* The "profile" object had more parameters added to it that can
override default parameters specified in the "attestation"
and "verification" objects.
* The "store" object was removed altogther as it was never
implemented.
* We now use libjwt to create outgoing Identity headers and to
parse and validate signatures on incoming Identiy headers. Our
previous custom implementation was much of the source of the
interoperability issues.
* General code cleanup and refactor.
* Moved things to better places.
* Separated some of the complex functions to smaller ones.
* Using context objects rather than passing tons of parameters
in function calls.
* Removed some complexity and unneeded encapsuation from the
config objects.
Resolves: #351
Resolves: #46
UserNote: Asterisk's stir-shaken feature has been refactored to
correct interoperability, RFC compliance, and performance issues.
See https://docs.asterisk.org/Deployment/STIR-SHAKEN for more
information.
UpgradeNote: The stir-shaken refactor is a breaking change but since
it's not working now we don't think it matters. The
stir_shaken.conf file has changed significantly which means that
existing ones WILL need to be changed. The stir_shaken.conf.sample
file in configs/samples/ has quite a bit more information. This is
also an ABI breaking change since some of the existing objects
needed to be changed or removed, and new ones added. Additionally,
if res_stir_shaken is enabled in menuselect, you'll need to either
have the development package for libjwt v1.15.3 installed or use
the --with-libjwt-bundled option with ./configure.
2023-10-26 10:27:35 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
ctx = ao2_alloc_options(sizeof(*ctx), ctx_destructor,
|
|
|
|
|
AO2_ALLOC_OPT_LOCK_NOLOCK);
|
|
|
|
|
if (!ctx) {
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_INTERNAL_ERROR,
|
|
|
|
|
LOG_ERROR, "%s: Unable to allocate memory for ctx\n", tag);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (ast_string_field_init(ctx, 1024) != 0) {
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_INTERNAL_ERROR,
|
|
|
|
|
LOG_ERROR, "%s: Unable to allocate memory for ctx\n", tag);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (ast_string_field_set(ctx, tag, tag) != 0) {
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_INTERNAL_ERROR,
|
|
|
|
|
LOG_ERROR, "%s: Unable to allocate memory for ctx\n", tag);
|
|
|
|
|
}
|
|
|
|
|
|
2024-04-25 11:56:15 -06:00
|
|
|
if (ast_string_field_set(ctx, orig_tn, canon_orig_tn) != 0) {
|
Stir/Shaken Refactor
Why do we need a refactor?
The original stir/shaken implementation was started over 3 years ago
when little was understood about practical implementation. The
result was an implementation that wouldn't actually interoperate
with any other stir-shaken implementations.
There were also a number of stir-shaken features and RFC
requirements that were never implemented such as TNAuthList
certificate validation, sending Reason headers in SIP responses
when verification failed but we wished to continue the call, and
the ability to send Media Key(mky) grants in the Identity header
when the call involved DTLS.
Finally, there were some performance concerns around outgoing
calls and selection of the correct certificate and private key.
The configuration was keyed by an arbitrary name which meant that
for every outgoing call, we had to scan the entire list of
configured TNs to find the correct cert to use. With only a few
TNs configured, this wasn't an issue but if you have a thousand,
it could be.
What's changed?
* Configuration objects have been refactored to be clearer about
their uses and to fix issues.
* The "general" object was renamed to "verification" since it
contains parameters specific to the incoming verification
process. It also never handled ca_path and crl_path
correctly.
* A new "attestation" object was added that controls the
outgoing attestation process. It sets default certificates,
keys, etc.
* The "certificate" object was renamed to "tn" and had it's key
change to telephone number since outgoing call attestation
needs to look up certificates by telephone number.
* The "profile" object had more parameters added to it that can
override default parameters specified in the "attestation"
and "verification" objects.
* The "store" object was removed altogther as it was never
implemented.
* We now use libjwt to create outgoing Identity headers and to
parse and validate signatures on incoming Identiy headers. Our
previous custom implementation was much of the source of the
interoperability issues.
* General code cleanup and refactor.
* Moved things to better places.
* Separated some of the complex functions to smaller ones.
* Using context objects rather than passing tons of parameters
in function calls.
* Removed some complexity and unneeded encapsuation from the
config objects.
Resolves: #351
Resolves: #46
UserNote: Asterisk's stir-shaken feature has been refactored to
correct interoperability, RFC compliance, and performance issues.
See https://docs.asterisk.org/Deployment/STIR-SHAKEN for more
information.
UpgradeNote: The stir-shaken refactor is a breaking change but since
it's not working now we don't think it matters. The
stir_shaken.conf file has changed significantly which means that
existing ones WILL need to be changed. The stir_shaken.conf.sample
file in configs/samples/ has quite a bit more information. This is
also an ABI breaking change since some of the existing objects
needed to be changed or removed, and new ones added. Additionally,
if res_stir_shaken is enabled in menuselect, you'll need to either
have the development package for libjwt v1.15.3 installed or use
the --with-libjwt-bundled option with ./configure.
2023-10-26 10:27:35 -06:00
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_INTERNAL_ERROR,
|
|
|
|
|
LOG_ERROR, "%s: Unable to allocate memory for ctx\n", tag);
|
|
|
|
|
}
|
|
|
|
|
|
2024-04-25 11:56:15 -06:00
|
|
|
if (ast_string_field_set(ctx, dest_tn, canon_dest_tn)) {
|
Stir/Shaken Refactor
Why do we need a refactor?
The original stir/shaken implementation was started over 3 years ago
when little was understood about practical implementation. The
result was an implementation that wouldn't actually interoperate
with any other stir-shaken implementations.
There were also a number of stir-shaken features and RFC
requirements that were never implemented such as TNAuthList
certificate validation, sending Reason headers in SIP responses
when verification failed but we wished to continue the call, and
the ability to send Media Key(mky) grants in the Identity header
when the call involved DTLS.
Finally, there were some performance concerns around outgoing
calls and selection of the correct certificate and private key.
The configuration was keyed by an arbitrary name which meant that
for every outgoing call, we had to scan the entire list of
configured TNs to find the correct cert to use. With only a few
TNs configured, this wasn't an issue but if you have a thousand,
it could be.
What's changed?
* Configuration objects have been refactored to be clearer about
their uses and to fix issues.
* The "general" object was renamed to "verification" since it
contains parameters specific to the incoming verification
process. It also never handled ca_path and crl_path
correctly.
* A new "attestation" object was added that controls the
outgoing attestation process. It sets default certificates,
keys, etc.
* The "certificate" object was renamed to "tn" and had it's key
change to telephone number since outgoing call attestation
needs to look up certificates by telephone number.
* The "profile" object had more parameters added to it that can
override default parameters specified in the "attestation"
and "verification" objects.
* The "store" object was removed altogther as it was never
implemented.
* We now use libjwt to create outgoing Identity headers and to
parse and validate signatures on incoming Identiy headers. Our
previous custom implementation was much of the source of the
interoperability issues.
* General code cleanup and refactor.
* Moved things to better places.
* Separated some of the complex functions to smaller ones.
* Using context objects rather than passing tons of parameters
in function calls.
* Removed some complexity and unneeded encapsuation from the
config objects.
Resolves: #351
Resolves: #46
UserNote: Asterisk's stir-shaken feature has been refactored to
correct interoperability, RFC compliance, and performance issues.
See https://docs.asterisk.org/Deployment/STIR-SHAKEN for more
information.
UpgradeNote: The stir-shaken refactor is a breaking change but since
it's not working now we don't think it matters. The
stir_shaken.conf file has changed significantly which means that
existing ones WILL need to be changed. The stir_shaken.conf.sample
file in configs/samples/ has quite a bit more information. This is
also an ABI breaking change since some of the existing objects
needed to be changed or removed, and new ones added. Additionally,
if res_stir_shaken is enabled in menuselect, you'll need to either
have the development package for libjwt v1.15.3 installed or use
the --with-libjwt-bundled option with ./configure.
2023-10-26 10:27:35 -06:00
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_INTERNAL_ERROR,
|
|
|
|
|
LOG_ERROR, "%s: Unable to allocate memory for ctx\n", tag);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
ctx->chan = chan;
|
|
|
|
|
ast_channel_ref(ctx->chan);
|
|
|
|
|
|
|
|
|
|
if (AST_VECTOR_INIT(&ctx->fingerprints, 1) != 0) {
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_INTERNAL_ERROR,
|
|
|
|
|
LOG_ERROR, "%s: Unable to allocate memory for ctx\n", tag);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Transfer the references */
|
|
|
|
|
ctx->etn = etn;
|
|
|
|
|
etn = NULL;
|
|
|
|
|
*ctxout = ctx;
|
|
|
|
|
ctx = NULL;
|
|
|
|
|
|
|
|
|
|
SCOPE_EXIT_RTN_VALUE(AST_STIR_SHAKEN_AS_SUCCESS, "%s: Done\n", tag);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
int ast_stir_shaken_as_ctx_wants_fingerprints(struct ast_stir_shaken_as_ctx *ctx)
|
|
|
|
|
{
|
|
|
|
|
return ENUM_BOOL(ctx->etn->acfg_common.send_mky, send_mky);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
enum ast_stir_shaken_as_response_code
|
|
|
|
|
ast_stir_shaken_as_ctx_add_fingerprint(
|
|
|
|
|
struct ast_stir_shaken_as_ctx *ctx, const char *alg, const char *fingerprint)
|
|
|
|
|
{
|
|
|
|
|
char *compacted_fp = ast_alloca(strlen(fingerprint) + 1);
|
|
|
|
|
const char *f = fingerprint;
|
|
|
|
|
char *fp = compacted_fp;
|
|
|
|
|
char *combined;
|
|
|
|
|
int rc;
|
|
|
|
|
SCOPE_ENTER(4, "%s: Add fingerprint %s:%s\n", ctx ? ctx->tag : "",
|
|
|
|
|
alg, fingerprint);
|
|
|
|
|
|
|
|
|
|
if (!ctx || ast_strlen_zero(alg) || ast_strlen_zero(fingerprint)) {
|
|
|
|
|
SCOPE_EXIT_RTN_VALUE(AST_STIR_SHAKEN_AS_INVALID_ARGUMENTS,
|
|
|
|
|
"%s: Missing arguments\n", ctx->tag);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (!ENUM_BOOL(ctx->etn->acfg_common.send_mky, send_mky)) {
|
|
|
|
|
SCOPE_EXIT_RTN_VALUE(AST_STIR_SHAKEN_AS_DISABLED,
|
|
|
|
|
"%s: Not needed\n", ctx->tag);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* De-colonize */
|
|
|
|
|
while (*f != '\0') {
|
|
|
|
|
if (*f != ':') {
|
|
|
|
|
*fp++ = *f;
|
|
|
|
|
}
|
|
|
|
|
f++;
|
|
|
|
|
}
|
|
|
|
|
*fp = '\0';
|
|
|
|
|
rc = ast_asprintf(&combined, "%s:%s", alg, compacted_fp);
|
|
|
|
|
if (rc < 0) {
|
|
|
|
|
SCOPE_EXIT_RTN_VALUE(AST_STIR_SHAKEN_AS_INTERNAL_ERROR,
|
|
|
|
|
"%s: Can't allocate memory for comobined string\n", ctx->tag);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
rc = AST_VECTOR_ADD_SORTED(&ctx->fingerprints, combined, strcasecmp);
|
|
|
|
|
if (rc < 0) {
|
|
|
|
|
SCOPE_EXIT_RTN_VALUE(AST_STIR_SHAKEN_AS_INTERNAL_ERROR,
|
|
|
|
|
"%s: Can't add entry to vector\n", ctx->tag);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
SCOPE_EXIT_RTN_VALUE(AST_STIR_SHAKEN_AS_SUCCESS,
|
|
|
|
|
"%s: Done\n", ctx->tag);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* We have to construct the PASSporT payload manually instead of
|
|
|
|
|
* using ast_json_pack. These macros help make sure nothing
|
|
|
|
|
* leaks if there are errors creating the individual objects.
|
|
|
|
|
*/
|
|
|
|
|
#define CREATE_JSON_SET_OBJ(__val, __obj, __name) \
|
|
|
|
|
({ \
|
|
|
|
|
struct ast_json *__var; \
|
|
|
|
|
if (!(__var = __val)) {\
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_INTERNAL_ERROR, \
|
|
|
|
|
LOG_ERROR, "%s: Cannot allocate one of the JSON objects\n", \
|
|
|
|
|
ctx->tag); \
|
|
|
|
|
} else { \
|
|
|
|
|
if (ast_json_object_set(__obj, __name, __var)) { \
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_INTERNAL_ERROR, \
|
|
|
|
|
LOG_ERROR, "%s: Cannot set one of the JSON objects\n", \
|
|
|
|
|
ctx->tag); \
|
|
|
|
|
} \
|
|
|
|
|
} \
|
|
|
|
|
(__var); \
|
|
|
|
|
})
|
|
|
|
|
|
|
|
|
|
#define CREATE_JSON_APPEND_ARRAY(__val, __obj) \
|
|
|
|
|
({ \
|
|
|
|
|
struct ast_json *__var; \
|
|
|
|
|
if (!(__var = __val)) {\
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_INTERNAL_ERROR, \
|
|
|
|
|
LOG_ERROR, "%s: Cannot allocate one of the JSON objects\n", \
|
|
|
|
|
ctx->tag); \
|
|
|
|
|
} else { \
|
|
|
|
|
if (ast_json_array_append(__obj, __var)) { \
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_INTERNAL_ERROR, \
|
|
|
|
|
LOG_ERROR, "%s: Cannot set one of the JSON objects\n", \
|
|
|
|
|
ctx->tag); \
|
|
|
|
|
} \
|
|
|
|
|
} \
|
|
|
|
|
(__var); \
|
|
|
|
|
})
|
|
|
|
|
|
|
|
|
|
static enum ast_stir_shaken_as_response_code pack_payload(
|
|
|
|
|
struct ast_stir_shaken_as_ctx *ctx, jwt_t *jwt)
|
|
|
|
|
{
|
|
|
|
|
RAII_VAR(struct ast_json *, payload, ast_json_object_create(), ast_json_unref);
|
|
|
|
|
/*
|
|
|
|
|
* These don't need RAII because once they're added to payload,
|
|
|
|
|
* they'll get destroyed when payload gets unreffed.
|
|
|
|
|
*/
|
|
|
|
|
struct ast_json *dest;
|
|
|
|
|
struct ast_json *tns;
|
|
|
|
|
struct ast_json *orig;
|
|
|
|
|
char origid[AST_UUID_STR_LEN];
|
|
|
|
|
char *payload_str = NULL;
|
|
|
|
|
SCOPE_ENTER(3, "%s: Enter\n", ctx->tag);
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* All fields added need to be in alphabetical order
|
|
|
|
|
* and there must be no whitespace in the result.
|
|
|
|
|
*
|
|
|
|
|
* We can't use ast_json_pack here because the entries
|
|
|
|
|
* need to be kept in order and the "mky" array may
|
|
|
|
|
* not be present.
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* The order of the calls matters. We want to add an object
|
|
|
|
|
* to its parent as soon as it's created, then add things
|
|
|
|
|
* to it. This way if something later fails, the whole thing
|
|
|
|
|
* will get destroyed when its parent gets destroyed.
|
|
|
|
|
*/
|
|
|
|
|
CREATE_JSON_SET_OBJ(ast_json_string_create(
|
|
|
|
|
attest_level_to_str(ctx->etn->acfg_common.attest_level)),
|
|
|
|
|
payload, "attest");
|
|
|
|
|
|
|
|
|
|
dest = CREATE_JSON_SET_OBJ(ast_json_object_create(), payload, "dest");
|
|
|
|
|
tns = CREATE_JSON_SET_OBJ(ast_json_array_create(), dest, "tn");
|
|
|
|
|
CREATE_JSON_APPEND_ARRAY(ast_json_string_create(ctx->dest_tn), tns);
|
|
|
|
|
|
|
|
|
|
CREATE_JSON_SET_OBJ(ast_json_integer_create(time(NULL)), payload, "iat");
|
|
|
|
|
|
|
|
|
|
if (AST_VECTOR_SIZE(&ctx->fingerprints)
|
|
|
|
|
&& ENUM_BOOL(ctx->etn->acfg_common.send_mky, send_mky)) {
|
|
|
|
|
struct ast_json *mky;
|
|
|
|
|
int i;
|
|
|
|
|
|
|
|
|
|
mky = CREATE_JSON_SET_OBJ(ast_json_array_create(), payload, "mky");
|
|
|
|
|
|
|
|
|
|
for (i = 0; i < AST_VECTOR_SIZE(&ctx->fingerprints); i++) {
|
|
|
|
|
struct ast_json *mk;
|
|
|
|
|
char *afp = AST_VECTOR_GET(&ctx->fingerprints, i);
|
|
|
|
|
char *fp = strchr(afp, ':');
|
|
|
|
|
*fp++ = '\0';
|
|
|
|
|
|
|
|
|
|
mk = CREATE_JSON_APPEND_ARRAY(ast_json_object_create(), mky);
|
|
|
|
|
CREATE_JSON_SET_OBJ(ast_json_string_create(afp), mk, "alg");
|
|
|
|
|
CREATE_JSON_SET_OBJ(ast_json_string_create(fp), mk, "dig");
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
orig = CREATE_JSON_SET_OBJ(ast_json_object_create(), payload, "orig");
|
|
|
|
|
CREATE_JSON_SET_OBJ(ast_json_string_create(ctx->orig_tn), orig, "tn");
|
|
|
|
|
|
|
|
|
|
ast_uuid_generate_str(origid, sizeof(origid));
|
|
|
|
|
CREATE_JSON_SET_OBJ(ast_json_string_create(origid), payload, "origid");
|
|
|
|
|
|
|
|
|
|
payload_str = ast_json_dump_string_format(payload, AST_JSON_COMPACT);
|
|
|
|
|
ast_trace(2, "Payload: %s\n", payload_str);
|
|
|
|
|
jwt_add_grants_json(jwt, payload_str);
|
|
|
|
|
ast_json_free(payload_str);
|
|
|
|
|
|
|
|
|
|
SCOPE_EXIT_RTN_VALUE(AST_STIR_SHAKEN_AS_SUCCESS, "Done\n");
|
|
|
|
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
enum ast_stir_shaken_as_response_code ast_stir_shaken_attest(
|
|
|
|
|
struct ast_stir_shaken_as_ctx *ctx, char **header)
|
|
|
|
|
{
|
|
|
|
|
RAII_VAR(jwt_t *, jwt, NULL, jwt_free);
|
|
|
|
|
jwt_alg_t alg;
|
|
|
|
|
char *encoded = NULL;
|
|
|
|
|
enum ast_stir_shaken_as_response_code as_rc;
|
|
|
|
|
int rc = 0;
|
|
|
|
|
SCOPE_ENTER(3, "%s: Attestation: orig: %s dest: %s\n",
|
|
|
|
|
ctx ? ctx->tag : "NULL", ctx ? ctx->orig_tn : "NULL",
|
|
|
|
|
ctx ? ctx->dest_tn : "NULL");
|
|
|
|
|
|
|
|
|
|
if (!ctx) {
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_VS_INTERNAL_ERROR, LOG_ERROR,
|
|
|
|
|
"%s: No context object!\n", "NULL");
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (header == NULL) {
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_INVALID_ARGUMENTS,
|
|
|
|
|
LOG_ERROR, "%s: Header buffer was NULL\n", ctx->tag);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
rc = jwt_new(&jwt);
|
|
|
|
|
if (rc != 0) {
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_INTERNAL_ERROR,
|
|
|
|
|
LOG_ERROR, "%s: Cannot create JWT\n", ctx->tag);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* All headers added need to be in alphabetical order!
|
|
|
|
|
*/
|
|
|
|
|
alg = jwt_str_alg(STIR_SHAKEN_ENCRYPTION_ALGORITHM);
|
|
|
|
|
jwt_set_alg(jwt, alg, (const unsigned char *)ctx->etn->acfg_common.raw_key,
|
|
|
|
|
ctx->etn->acfg_common.raw_key_length);
|
|
|
|
|
jwt_add_header(jwt, "ppt", STIR_SHAKEN_PPT);
|
|
|
|
|
jwt_add_header(jwt, "typ", STIR_SHAKEN_TYPE);
|
|
|
|
|
jwt_add_header(jwt, "x5u", ctx->etn->acfg_common.public_cert_url);
|
|
|
|
|
|
|
|
|
|
as_rc = pack_payload(ctx, jwt);
|
|
|
|
|
if (as_rc != AST_STIR_SHAKEN_AS_SUCCESS) {
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(as_rc,
|
|
|
|
|
LOG_ERROR, "%s: Cannot pack payload\n", ctx->tag);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
encoded = jwt_encode_str(jwt);
|
|
|
|
|
if (!encoded) {
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_SIGN_ENCODE_FAILURE,
|
|
|
|
|
LOG_ERROR, "%s: Unable to sign/encode JWT\n", ctx->tag);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
rc = ast_asprintf(header, "%s;info=<%s>;alg=%s;ppt=%s",
|
|
|
|
|
encoded, ctx->etn->acfg_common.public_cert_url, jwt_alg_str(alg),
|
|
|
|
|
STIR_SHAKEN_PPT);
|
|
|
|
|
ast_std_free(encoded);
|
|
|
|
|
if (rc < 0) {
|
|
|
|
|
SCOPE_EXIT_LOG_RTN_VALUE(AST_STIR_SHAKEN_AS_INTERNAL_ERROR,
|
|
|
|
|
LOG_ERROR, "%s: Unable to allocate memory for identity header\n",
|
|
|
|
|
ctx->tag);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
SCOPE_EXIT_RTN_VALUE(AST_STIR_SHAKEN_AS_SUCCESS, "%s: Done\n", ctx->tag);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
int as_reload()
|
|
|
|
|
{
|
|
|
|
|
as_config_reload();
|
|
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
int as_unload()
|
|
|
|
|
{
|
|
|
|
|
as_config_unload();
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
int as_load()
|
|
|
|
|
{
|
|
|
|
|
if (as_config_load()) {
|
|
|
|
|
return AST_MODULE_LOAD_DECLINE;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return AST_MODULE_LOAD_SUCCESS;
|
|
|
|
|
}
|