asterisk/res/res_pjsip_outbound_authenti...

604 lines
20 KiB
C
Raw Permalink Normal View History

/*
* Asterisk -- An open source telephony toolkit.
*
* Copyright (C) 2013, Digium, Inc.
*
* Mark Michelson <mmichelson@digium.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.
*/
/*** MODULEINFO
<depend>pjproject</depend>
<depend>res_pjsip</depend>
<support_level>core</support_level>
***/
#include "asterisk.h"
#include <pjsip.h>
#include "asterisk/res_pjsip.h"
#include "asterisk/logger.h"
#include "asterisk/module.h"
#include "asterisk/strings.h"
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
#include "asterisk/vector.h"
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
pj_str_t supported_digest_algorithms[] = {
{ "MD5", 3}
};
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
/*!
* \internal
* \brief Determine proper authenticate header
*
* We need to search for different headers depending on whether
* the response code from the UAS/Proxy was 401 or 407.
*/
static pjsip_hdr_e get_auth_search_type(pjsip_rx_data *challenge)
{
if (challenge->msg_info.msg->line.status.code == PJSIP_SC_UNAUTHORIZED) {
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
return PJSIP_H_WWW_AUTHENTICATE;
} else if (challenge->msg_info.msg->line.status.code == PJSIP_SC_PROXY_AUTHENTICATION_REQUIRED) {
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
return PJSIP_H_PROXY_AUTHENTICATE;
} else {
ast_log(LOG_ERROR,
"Status code %d was received when it should have been 401 or 407.\n",
challenge->msg_info.msg->line.status.code);
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
return PJSIP_H_OTHER;
}
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
}
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
/*!
* \internal
* \brief Determine if digest algorithm in the header is one we support
*
* \retval 1 If we support the algorithm
* \retval 0 If we do not
*
*/
static int is_digest_algorithm_supported(pjsip_www_authenticate_hdr *auth_hdr)
{
int digest;
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
/* An empty digest is assumed to be md5 */
if (pj_strlen(&auth_hdr->challenge.digest.algorithm) == 0) {
return 1;
}
for (digest = 0; digest < ARRAY_LEN(supported_digest_algorithms); digest++) {
if (pj_stricmp(&auth_hdr->challenge.digest.algorithm, &supported_digest_algorithms[digest]) == 0) {
return 1;
}
}
return 0;
}
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
/*!
* \internal
* \brief Initialize pjproject with a valid set of credentials
*
* RFC7616 and RFC8760 allow more than one WWW-Authenticate or
* Proxy-Authenticate header per realm, each with different digest
* algorithms (including new ones like SHA-256 and SHA-512-256). However,
* thankfully, a UAS can NOT send back multiple Authenticate headers for
* the same realm with the same digest algorithm. The UAS is also
* supposed to send the headers in order of preference with the first one
* being the most preferred.
*
* We're supposed to send an Authorization header for the first one we
* encounter for a realm that we can support.
*
* The UAS can also send multiple realms, especially when it's a proxy
* that has forked the request in which case the proxy will aggregate all
* of the Authenticate and then them all back to the UAC.
*
* It doesn't stop there though... Each realm can require a different
* username from the others. There's also nothing preventing each digest
* algorithm from having a unique password although I'm not sure if
* that adds any benefit.
*
* So now... For each Authenticate header we encounter, we have to
* determine if we support the digest algorithm and, if not, just skip the
* header. We then have to find an auth object that matches the realm AND
* the digest algorithm or find a wildcard object that matches the digest
* algorithm. If we find one, we add it to the results vector and read the
* next Authenticate header. If the next header is for the same realm AND
* we already added an auth object for that realm, we skip the header.
* Otherwise we repeat the process for the next header.
*
* In the end, we'll have accumulated a list of credentials we can pass to
* pjproject that it can use to add Authentication headers to a request.
*
* \note: Neither we nor pjproject can currently handle digest algorithms
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
* other than MD5. We don't even have a place for it in the ast_sip_auth
* object. For this reason, we just skip processing any Authenticate
* header that's not MD5. When we support the others, we'll move the
* check into the loop that searches the objects.
*/
static pj_status_t set_outbound_authentication_credentials(pjsip_auth_clt_sess *auth_sess,
const struct ast_sip_auth_objects_vector *auth_objects_vector, pjsip_rx_data *challenge,
struct ast_str **realms)
{
int i;
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
size_t auth_object_count;
pjsip_www_authenticate_hdr *auth_hdr = NULL;
pj_status_t res = PJ_SUCCESS;
pjsip_hdr_e search_type;
size_t cred_count;
pjsip_cred_info *creds_array;
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
/*
* Normally vector elements are pointers to something else, usually
* structures. In this case however, the elements are the
* structures themselves instead of pointers to them. This is due
* to the fact that pjsip_auth_clt_set_credentials() expects an
* array of structures, not an array of pointers to structures.
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
* Thankfully, vectors allow you to "steal" their underlying
* arrays, in this case an array of pjsip_cred_info structures,
* which we'll pass to pjsip_auth_clt_set_credentials() at the
* end.
*/
AST_VECTOR(cred_info, pjsip_cred_info) auth_creds;
search_type = get_auth_search_type(challenge);
if (search_type == PJSIP_H_OTHER) {
/*
* The status code on the response wasn't 401 or 407
* so there are no WWW-Authenticate or Proxy-Authenticate
* headers to process.
*/
return PJ_ENOTSUP;
}
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
auth_object_count = AST_VECTOR_SIZE(auth_objects_vector);
if (auth_object_count == 0) {
/* This shouldn't happen but we'll check anyway. */
return PJ_EINVAL;
}
/*
* The number of pjsip_cred_infos we send to pjproject can
* vary based on the number of acceptable headers received
* and the number of acceptable auth objects on the endpoint
* so we just use a vector to accumulate them.
*
* NOTE: You have to call AST_VECTOR_FREE() on the vector
* but you don't have to free the elements because they're
* actual structures, not pointers to structures.
*/
if (AST_VECTOR_INIT(&auth_creds, 5) != 0) {
return PJ_ENOMEM;
}
/*
* It's going to be rare that we actually have more than one
* WWW-Authentication header or more than one auth object to
* match to it so the following nested loop should be fine.
*/
while ((auth_hdr = pjsip_msg_find_hdr(challenge->msg_info.msg,
search_type, auth_hdr ? auth_hdr->next : NULL))) {
int exact_match_index = -1;
int wildcard_match_index = -1;
int match_index = 0;
pjsip_cred_info auth_cred;
struct ast_sip_auth *auth = NULL;
memset(&auth_cred, 0, sizeof(auth_cred));
/*
* Since we only support the MD5 algorithm at the current time,
* there's no sense searching for auth objects that match the algorithm.
* In fact, the auth_object structure doesn't even have a member
* for it.
*
* When we do support more algorithms, this check will need to be
* moved inside the auth object loop below.
*
* Note: The header may not have specified an algorithm at all in which
* case it's assumed to be MD5. is_digest_algorithm_supported() returns
* true for that case.
*/
if (!is_digest_algorithm_supported(auth_hdr)) {
ast_debug(3, "Skipping header with realm '%.*s' and unsupported '%.*s' algorithm \n",
(int)auth_hdr->challenge.digest.realm.slen, auth_hdr->challenge.digest.realm.ptr,
(int)auth_hdr->challenge.digest.algorithm.slen, auth_hdr->challenge.digest.algorithm.ptr);
continue;
}
/*
* Appending the realms is strictly so digest_create_request_with_auth()
* can display good error messages. Since we only support one algorithm,
* there can't be more than one header with the same realm. No need to worry
* about duplicate realms until then.
*/
if (*realms) {
ast_str_append(realms, 0, "%.*s, ",
(int)auth_hdr->challenge.digest.realm.slen, auth_hdr->challenge.digest.realm.ptr);
}
ast_debug(3, "Searching auths to find matching ones for header with realm '%.*s' and algorithm '%.*s'\n",
(int)auth_hdr->challenge.digest.realm.slen, auth_hdr->challenge.digest.realm.ptr,
(int)auth_hdr->challenge.digest.algorithm.slen, auth_hdr->challenge.digest.algorithm.ptr);
/*
* Now that we have a valid header, we can loop over the auths available to
* find either an exact realm match or, failing that, a wildcard auth (an
* auth with an empty or "*" realm).
*
* NOTE: We never use the global default realm when we're the UAC responding
* to a 401 or 407. We only use that when we're the UAS (handled elsewhere)
* and the auth object didn't have a realm.
*/
for (i = 0; i < auth_object_count; ++i) {
auth = AST_VECTOR_GET(auth_objects_vector, i);
/*
* If this auth object's realm exactly matches the one
* from the header, we can just break out and use it.
*
* NOTE: If there's more than one auth object for an endpoint with
* a matching realm it's a misconfiguration. We'll only use the first.
*/
if (pj_stricmp2(&auth_hdr->challenge.digest.realm, auth->realm) == 0) {
ast_debug(3, "Found matching auth '%s' with realm '%s'\n", ast_sorcery_object_get_id(auth),
auth->realm);
exact_match_index = i;
/*
* If we found an exact realm match, there's no need to keep
* looking for a wildcard.
*/
break;
}
/*
* If this auth object's realm is empty or a "*", it's a wildcard
* auth object. We going to save its index but keep iterating over
* the vector in case we find an exact match later.
*
* NOTE: If there's more than one wildcard auth object for an endpoint
* it's a misconfiguration. We'll only use the first.
*/
if (wildcard_match_index < 0
&& (ast_strlen_zero(auth->realm) || ast_strings_equal(auth->realm, "*"))) {
ast_debug(3, "Found wildcard auth '%s' for realm '%.*s'\n", ast_sorcery_object_get_id(auth),
(int)auth_hdr->challenge.digest.realm.slen, auth_hdr->challenge.digest.realm.ptr);
wildcard_match_index = i;
}
}
if (exact_match_index < 0 && wildcard_match_index < 0) {
/*
* Didn't find either a wildcard or an exact realm match.
* Move on to the next header.
*/
ast_debug(3, "No auth matching realm or no wildcard found for realm '%.*s'\n",
(int)auth_hdr->challenge.digest.realm.slen, auth_hdr->challenge.digest.realm.ptr);
continue;
}
if (exact_match_index >= 0) {
/*
* If we found an exact match, we'll always prefer that.
*/
match_index = exact_match_index;
auth = AST_VECTOR_GET(auth_objects_vector, match_index);
ast_debug(3, "Using matched auth '%s' with realm '%.*s'\n", ast_sorcery_object_get_id(auth),
(int)auth_hdr->challenge.digest.realm.slen, auth_hdr->challenge.digest.realm.ptr);
} else {
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
/*
* We'll only use the wildcard if we didn't find an exact match.
*/
match_index = wildcard_match_index;
auth = AST_VECTOR_GET(auth_objects_vector, match_index);
ast_debug(3, "Using wildcard auth '%s' for realm '%.*s'\n", ast_sorcery_object_get_id(auth),
(int)auth_hdr->challenge.digest.realm.slen, auth_hdr->challenge.digest.realm.ptr);
}
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
/*
* Copy the fields from the auth_object to the
* pjsip_cred_info structure.
*/
auth_cred.realm = auth_hdr->challenge.common.realm;
pj_cstr(&auth_cred.username, auth->auth_user);
pj_cstr(&auth_cred.scheme, "digest");
switch (auth->type) {
case AST_SIP_AUTH_TYPE_USER_PASS:
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
pj_cstr(&auth_cred.data, auth->auth_pass);
auth_cred.data_type = PJSIP_CRED_DATA_PLAIN_PASSWD;
break;
case AST_SIP_AUTH_TYPE_MD5:
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
pj_cstr(&auth_cred.data, auth->md5_creds);
auth_cred.data_type = PJSIP_CRED_DATA_DIGEST;
break;
res_pjsip: Implement additional SIP RFCs for Google Voice trunk compatability This change implements a few different generic things which were brought on by Google Voice SIP. 1. The concept of flow transports have been introduced. These are configurable transports in pjsip.conf which can be used to reference a flow of signaling to a target. These have runtime configuration that can be changed by the signaling itself (such as Service-Routes and P-Preferred-Identity). When used these guarantee an individual connection (in the case of TCP or TLS) even if multiple flow transports exist to the same target. 2. Service-Routes (RFC 3608) support has been added to the outbound registration module which when received will be stored on the flow transport and used for requests referencing it. 3. P-Associated-URI / P-Preferred-Identity (RFC 3325) support has been added to the outbound registration module. If a P-Associated-URI header is received it will be used on requests as the P-Preferred-Identity. 4. Configurable outbound extension support has been added to the outbound registration module. When set the extension will be placed in the Supported header. 5. Header parameters can now be configured on an outbound registration which will be placed in the Contact header. 6. Google specific OAuth / Bearer token authentication (draft-ietf-sipcore-sip-authn-02) has been added to the outbound registration module. All functionality changes are controlled by pjsip.conf configuration options and do not affect non-configured pjsip endpoints otherwise. ASTERISK-27971 #close Change-Id: Id214c2d1c550a41fcf564b7df8f3da7be565bd58
2018-07-18 12:45:26 +00:00
case AST_SIP_AUTH_TYPE_GOOGLE_OAUTH:
/* nothing to do. handled seperately in res_pjsip_outbound_registration */
break;
case AST_SIP_AUTH_TYPE_ARTIFICIAL:
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
ast_log(LOG_ERROR,
"Trying to set artificial outbound auth credentials shouldn't happen.\n");
continue;
} /* End auth object loop */
/*
* Because the vector contains actual structures and not pointers
* to structures, the call to AST_VECTOR_APPEND results in a simple
* assign of one structure to another, effectively copying the auth_cred
* structure contents to the array element.
*
* Also note that the calls to pj_cstr above set their respective
* auth_cred fields to the _pointers_ of their corresponding auth
* object fields. This is safe because the call to
* pjsip_auth_clt_set_credentials() below strdups them before we
* return to the calling function which decrements the reference
* counts.
*/
res = AST_VECTOR_APPEND(&auth_creds, auth_cred);
if (res != PJ_SUCCESS) {
res = PJ_ENOMEM;
goto cleanup;
}
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
} /* End header loop */
if (*realms && ast_str_strlen(*realms)) {
/*
* Again, this is strictly so digest_create_request_with_auth()
* can display good error messages.
*
* Chop off the trailing ", " on the last realm.
*/
ast_str_truncate(*realms, ast_str_strlen(*realms) - 2);
}
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
if (AST_VECTOR_SIZE(&auth_creds) == 0) {
/* No matching auth objects were found. */
res = PJSIP_ENOCREDENTIAL;
goto cleanup;
}
/*
* Here's where we steal the cred info structures from the vector.
*
* The steal effectively returns a pointer to the underlying
* array of pjsip_cred_info structures which is exactly what we need
* to pass to pjsip_auth_clt_set_credentials().
*
* <struct cred info><struct cred info>...<struct cred info>
* ^pointer
*
* Since we stole the array from the vector, we have to free it ourselves.
*
* We also have to copy the size before we steal because stealing
* resets the vector size to 0.
*/
cred_count = AST_VECTOR_SIZE(&auth_creds);
creds_array = AST_VECTOR_STEAL_ELEMENTS(&auth_creds);
res = pjsip_auth_clt_set_credentials(auth_sess, cred_count, creds_array);
ast_free(creds_array);
if (res == PJ_SUCCESS) {
ast_debug(3, "Set %zu credentials in auth session\n", cred_count);
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
} else {
ast_log(LOG_ERROR, "Failed to set %zu credentials in auth session\n", cred_count);
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
}
cleanup:
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
AST_VECTOR_FREE(&auth_creds);
return res;
}
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
/*!
* \internal
* \brief Create new tdata with auth based on original tdata
* \param auth_ids_vector Vector of auth IDs retrieved from endpoint
* \param challenge rdata of the response from the UAS with challenge
* \param old_request tdata from the original request
* \param new_request tdata of the new request with the auth
*
* This function is what's registered with ast_sip_register_outbound_authenticator()
*
* \retval 0 success
* \retval -1 failure
*/
static int digest_create_request_with_auth(const struct ast_sip_auth_vector *auth_ids_vector,
pjsip_rx_data *challenge, pjsip_tx_data *old_request, pjsip_tx_data **new_request)
{
pjsip_auth_clt_sess auth_sess;
pjsip_cseq_hdr *cseq;
pj_status_t status;
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
struct ast_sip_auth_objects_vector auth_objects_vector;
size_t auth_object_count = 0;
struct ast_sip_endpoint *endpoint;
char *id = NULL;
const char *id_type;
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
struct ast_str *realms = NULL;
pjsip_dialog *dlg;
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
int res = -1;
/*
* Some older compilers have an issue with initializing structures with
* pjsip_auth_clt_sess auth_sess = { 0, };
* so we'll just do it the old fashioned way.
*/
memset(&auth_sess, 0, sizeof(auth_sess));
dlg = pjsip_rdata_get_dlg(challenge);
if (dlg) {
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
/* The only thing we use endpoint for is to get an id for error/debug messages */
endpoint = ast_sip_dialog_get_endpoint(dlg);
id = endpoint ? ast_strdupa(ast_sorcery_object_get_id(endpoint)) : NULL;
ao2_cleanup(endpoint);
id_type = "Endpoint";
}
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
/* If there was no dialog, then this is probably a REGISTER so no endpoint */
if (!id) {
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
/* The only thing we use the address for is to get an id for error/debug messages */
id = ast_alloca(AST_SOCKADDR_BUFLEN);
pj_sockaddr_print(&challenge->pkt_info.src_addr, id, AST_SOCKADDR_BUFLEN, 3);
id_type = "Host";
}
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
if (!auth_ids_vector || AST_VECTOR_SIZE(auth_ids_vector) == 0) {
ast_log(LOG_ERROR, "%s: '%s': There were no auth ids available\n", id_type, id);
return -1;
}
if (AST_VECTOR_INIT(&auth_objects_vector, AST_VECTOR_SIZE(auth_ids_vector)) != 0) {
ast_log(LOG_ERROR, "%s: '%s': Couldn't initialize auth object vector\n", id_type, id);
return -1;
}
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
/*
* We don't really care about ast_sip_retrieve_auths_vector()'s return code
* because we're checking the count of objects in the vector.
*
* Don't forget to call
* ast_sip_cleanup_auth_objects_vector(&auth_objects_vector);
* AST_VECTOR_FREE(&auth_objects_vector);
* when you're done with the vector
*/
ast_sip_retrieve_auths_vector(auth_ids_vector, &auth_objects_vector);
auth_object_count = AST_VECTOR_SIZE(&auth_objects_vector);
if (auth_object_count == 0) {
/*
* If none of the auth ids were found, we can't continue.
* We're OK if there's at least one left.
* ast_sip_retrieve_auths_vector() will print a warning for every
* id that wasn't found.
*/
res = -1;
goto cleanup;
}
if (pjsip_auth_clt_init(&auth_sess, ast_sip_get_pjsip_endpoint(),
old_request->pool, 0) != PJ_SUCCESS) {
ast_log(LOG_ERROR, "%s: '%s': Failed to initialize client authentication session\n",
id_type, id);
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
res = -1;
goto cleanup;
}
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
/*
* realms is used only for displaying good error messages.
*/
realms = ast_str_create(32);
if (!realms) {
res = -1;
goto cleanup;
}
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
/*
* Load pjproject with the valid credentials for the Authentication headers
* received on the 401 or 407 response.
*/
status = set_outbound_authentication_credentials(&auth_sess, &auth_objects_vector, challenge, &realms);
switch (status) {
case PJ_SUCCESS:
break;
case PJSIP_ENOCREDENTIAL:
ast_log(LOG_WARNING,
"%s: '%s': No auth objects matching realm(s) '%s' from challenge found.\n", id_type, id,
realms ? ast_str_buffer(realms) : "<none>");
res = -1;
goto cleanup;
default:
ast_log(LOG_WARNING, "%s: '%s': Failed to set authentication credentials\n", id_type, id);
res = -1;
goto cleanup;
}
/*
* reinit_req actually creates the Authorization headers to send on
* the next request. If reinit_req already has a cached credential
* from an earlier successful authorization, it'll use it. Otherwise
* it'll create a new authorization and cache it.
*/
status = pjsip_auth_clt_reinit_req(&auth_sess, challenge, old_request, new_request);
switch (status) {
case PJ_SUCCESS:
/* PJSIP creates a new transaction for new_request (meaning it creates a new
* branch). However, it recycles the Call-ID, from-tag, and CSeq from the
* original request. Some SIP implementations will not process the new request
* since the CSeq is the same as the original request. Incrementing it here
* fixes the interop issue
*/
cseq = pjsip_msg_find_hdr((*new_request)->msg, PJSIP_H_CSEQ, NULL);
ast_assert(cseq != NULL);
++cseq->cseq;
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
res = 0;
goto cleanup;
case PJSIP_ENOCREDENTIAL:
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
/*
* This should be rare since set_outbound_authentication_credentials()
* did the matching but you never know.
*/
ast_log(LOG_WARNING,
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
"%s: '%s': No auth objects matching realm(s) '%s' from challenge found.\n", id_type, id,
realms ? ast_str_buffer(realms) : "<none>");
break;
case PJSIP_EAUTHSTALECOUNT:
ast_log(LOG_WARNING,
"%s: '%s': Unable to create request with auth. Number of stale retries exceeded.\n",
id_type, id);
break;
case PJSIP_EFAILEDCREDENTIAL:
ast_log(LOG_WARNING, "%s: '%s': Authentication credentials not accepted by server.\n",
id_type, id);
break;
default:
ast_log(LOG_WARNING, "%s: '%s': Unable to create request with auth. Unknown failure.\n",
id_type, id);
break;
}
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
res = -1;
cleanup:
#if defined(HAVE_PJSIP_AUTH_CLT_DEINIT)
/* If we initialized the auth_sess, clean it up */
if (auth_sess.endpt) {
pjsip_auth_clt_deinit(&auth_sess);
}
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
#endif
ast_sip_cleanup_auth_objects_vector(&auth_objects_vector);
AST_VECTOR_FREE(&auth_objects_vector);
ast_free(realms);
res_pjsip_outbound_authenticator_digest: Be tolerant of RFC8760 UASs RFC7616 and RFC8760 allow more than one WWW-Authenticate or Proxy-Authenticate header per realm, each with different digest algorithms (including new ones like SHA-256 and SHA-512-256). Thankfully however a UAS can NOT send back multiple Authenticate headers for the same realm with the same digest algorithm. The UAS is also supposed to send the headers in order of preference with the first one being the most preferred. We're supposed to send an Authorization header for the first one we encounter for a realm that we can support. The UAS can also send multiple realms, especially when it's a proxy that has forked the request in which case the proxy will aggregate all of the Authenticate headers and then send them all back to the UAC. It doesn't stop there though... Each realm can require a different username from the others. There's also nothing preventing each digest algorithm from having a unique password although I'm not sure if that adds any benefit. So now... For each Authenticate header we encounter, we have to determine if we support the digest algorithm and, if not, just skip the header. We then have to find an auth object that matches the realm AND the digest algorithm or find a wildcard object that matches the digest algorithm. If we find one, we add it to the results vector and read the next Authenticate header. If the next header is for the same realm AND we already added an auth object for that realm, we skip the header. Otherwise we repeat the process for the next header. In the end, we'll have accumulated a list of credentials we can pass to pjproject that it can use to add Authentication headers to a request. NOTE: Neither we nor pjproject can currently handle digest algorithms other than MD5. We don't even have a place for it in the ast_sip_auth object. For this reason, we just skip processing any Authenticate header that's not MD5. When we support the others, we'll move the check into the loop that searches the objects. Changes: * Added a new API ast_sip_retrieve_auths_vector() that takes in a vector of auth ids (usually supplied on a call to ast_sip_create_request_with_auth()) and populates another vector with the actual objects. * Refactored res_pjsip_outbound_authenticator_digest to handle multiple Authenticate headers and set the stage for handling additional digest algorithms. * Added a pjproject patch that allows them to ignore digest algorithms they don't support. This patch has already been merged upstream. * Updated documentation for auth objects in the XML and in pjsip.conf.sample. * Although res_pjsip_authenticator_digest isn't affected by this change, some debugging and a testsuite AMI event was added to facilitate testing. Discovered during OpenSIPit 2021. ASTERISK-29397 Change-Id: I3aef5ce4fe1d27e48d61268520f284d15d650281
2021-04-15 15:43:48 +00:00
return res;
}
static struct ast_sip_outbound_authenticator digest_authenticator = {
.create_request_with_auth = digest_create_request_with_auth,
};
static int load_module(void)
{
if (ast_sip_register_outbound_authenticator(&digest_authenticator)) {
return AST_MODULE_LOAD_DECLINE;
}
return AST_MODULE_LOAD_SUCCESS;
}
static int unload_module(void)
{
ast_sip_unregister_outbound_authenticator(&digest_authenticator);
return 0;
}
AST_MODULE_INFO(ASTERISK_GPL_KEY, AST_MODFLAG_LOAD_ORDER, "PJSIP authentication resource",
.support_level = AST_MODULE_SUPPORT_CORE,
.load = load_module,
.unload = unload_module,
.load_pri = AST_MODPRI_CHANNEL_DEPEND,
.requires = "res_pjsip",
);