August 2010 - Freeradius-Devel - lists.freeradius.org

Missing TLS Change Cipher Spec and TLS Finished in EAP-TLS exchanges
by arno＠natisbad.org 09 Jul '11

09 Jul '11

Hi, I upgraded an internal freeradius server (Linux, Debian unstable, libssl 0.9.8) used for EAP-TLS (802.11 WPA2 access) from 2.0.5 to 2.1.3. With 2.1.3, client authentication fails in the following way (wpasupplicant on Debian unstable, working with 2.0.5): - the TLS exchange is completed from freeradius point of view. It even sends the Access-Accept message to the NAS. - from the clients point of view, the TLS exchange is not over, it expects some more data from the server which are never sent. I took two pcap traces (sent to you privately, Alan), one with 2.0.5, one with 2.1.3. It looks like the TLS Change Cipher and TLS Tinished elements of the TLS exchange are not in the EAP encapsulated TLS packets sent by the server. Compare packet #23 of 2.1.3 trace with packet #27 of 2.0.5 trace. If you use wireshark, you will notice that the SSL layer is dissected; this is because #27 has a small brother: #29. It is not sent by 2.1.3. Alan, if you have an idea or a patch, I can test it on monday. Cheers, a+

5 15

eap2 in freeradius
by Chanthearith HUON 04 Apr '11

04 Apr '11

Hi everyone, I am testing eap2 in freeradius. I run "radiusd -X", I got following error message " radiusd: symbol lookup error: /usr/lib/rlm_eap2-2.1.0.so: undefined symbol: eap_server_register_methods " Anyone, have any idea what might be wrong with the configuration? Thank in advance, Thierry

5 10

Hanging Problem in 2.1.10
by David Mitchell 10 Sep '10

10 Sep '10

Alan, I had a hang pretty quickly this time. As last time, the last output was that it the 'entering group authenticate' section. I got the bt output. I'm not sure if it's helpful or not. I just re-read doc/bugs and will try it again with gdb logging enabled. I'm not sure if that will help or not. Thu Aug 26 08:33:10 2010 : Info: # Executing section authorize from file /usr/local/etc/raddb/sites-enabled/default Thu Aug 26 08:33:10 2010 : Info: +- entering group authorize {...} Thu Aug 26 08:33:10 2010 : Info: ++[preprocess] returns ok Thu Aug 26 08:33:10 2010 : Info: [suffix] No '@' in User-Name = "mark", looking up realm NULL Thu Aug 26 08:33:10 2010 : Info: [suffix] Found realm "~.+$" Thu Aug 26 08:33:10 2010 : Info: [suffix] Adding Stripped-User-Name = "mark" Thu Aug 26 08:33:10 2010 : Info: [suffix] Adding Realm = "(null)" Thu Aug 26 08:33:10 2010 : Info: [suffix] Authentication realm is LOCAL. Thu Aug 26 08:33:10 2010 : Info: ++[suffix] returns ok Thu Aug 26 08:33:10 2010 : Info: [eap] EAP packet type response id 2 length 136 Thu Aug 26 08:33:10 2010 : Info: [eap] No EAP Start, assuming it's an on-going EAP conversation Thu Aug 26 08:33:10 2010 : Info: ++[eap] returns updated Thu Aug 26 08:33:10 2010 : Info: ++[files] returns noop Thu Aug 26 08:33:10 2010 : Info: Found Auth-Type = EAP Thu Aug 26 08:33:10 2010 : Info: # Executing group from file /usr/local/etc/raddb/sites-enabled/default Thu Aug 26 08:33:10 2010 : Info: +- entering group authenticate {...} ^C Program received signal SIGINT, Interrupt. [Switching to Thread 0xb75af8c0 (LWP 14706)] 0xb77b7424 in __kernel_vsyscall () (gdb) bt #0 0xb77b7424 in __kernel_vsyscall () #1 0xb7752c99 in __lll_lock_wait () from /lib/i686/cmov/libpthread.so.0 #2 0xb774e0c4 in _L_lock_89 () from /lib/i686/cmov/libpthread.so.0 #3 0xb774d9f2 in pthread_mutex_lock () from /lib/i686/cmov/libpthread.so.0 #4 0xb75ac26f in eap_handler_free (inst=0x827fd98, handler=0x82af1a8) at mem.c:137 #5 0xb75ac3d5 in eaplist_expire (inst=0x827fd98, timestamp=1282833190) at mem.c:315 #6 0xb75ac856 in eaplist_find (inst=0x827fd98, request=0x82cee78, eap_packet=0x82ccc40) at mem.c:502 #7 0xb75aa76f in eap_handler (inst=0x827fd98, eap_packet_p=0xbfffd118, request=0x82cee78) at eap.c:974 #8 0xb75a9d80 in eap_authenticate (instance=0x827fd98, request=0x82cee78) at rlm_eap.c:298 #9 0x08089bbc in modcall (component=0, c=0x8293f88, request=0x82cee78) at modcall.c:297 #10 0x0808651b in indexed_modcall (comp=0, idx=6, request=0x82cee78) at modules.c:728 #11 0x0808692c in module_authenticate (auth_type=6, request=0x82cee78) at modules.c:1502 #12 0x080753c4 in rad_authenticate (request=0x82cee78) at auth.c:371 #13 0x08095854 in radius_handle_request (request=0x82cee78, fun=0x8074a10 <rad_authenticate>) at event.c:3767 #14 0x0808c9bf in thread_pool_addrequest (request=0x82cee78, fun=0x8074a10 <rad_authenticate>) at threads.c:874 #15 0x08091f7a in event_socket_handler (xel=0x82980a8, fd=7, ctx=0x829bb40) at event.c:3412 #16 0xb77a808c in fr_event_loop (el=0x82980a8) at event.c:411 #17 0x08092df6 in radius_event_process () at event.c:3753 #18 0x0808a782 in main (argc=2, argv=0xbfffe5a4) at radiusd.c:406 (gdb) -- ----------------------------------------------------------------- | David Mitchell (mitchell(a)ucar.edu) Network Engineer IV | | Tel: (303) 497-1845 National Center for | | FAX: (303) 497-1818 Atmospheric Research | -----------------------------------------------------------------

2 4

Additional EAP-TLS Logging Option
by Ross, Michael 08 Sep '10

08 Sep '10

Hello, I customized a freeradius installation to include an option to log information from the client certificates (subject, issuer, serial, expiration) when doing eap-tls authentication. This is done through a new option, log_certificates, in the tls section of the eap.conf file. Is there any interest in adding this capability into the baseline freeradius? I'm happy to make any changes that would be required to do that. Below are the updated rlm_eap_tls.h and rlm_eap_tls.c. Thanks, Mike Ross /* * rlm_eap_tls.h * * Version: $Id: rlm_eap_tls.h,v 1.13 2007/04/25 14:19:26 aland Exp $ * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA * * Copyright 2001 hereUare Communications, Inc. <raghud(a)hereuare.com> * Copyright 2003 Alan DeKok <aland(a)freeradius.org> * Copyright 2006 The FreeRADIUS server project * *----------------------------------------------------------------------------- * Modified by The Boeing Company, August 26, 2008 * * Added int log_certificates to the eap_tls_conf struct * * Copyright 2008 The Boeing Company <Michael.Ross2(a)boeing.com> *----------------------------------------------------------------------------- */ #ifndef _RLM_EAP_TLS_H #define _RLM_EAP_TLS_H #include <freeradius-devel/ident.h> RCSIDH(rlm_eap_tls_h, "$Id: rlm_eap_tls.h,v 1.13 2007/04/25 14:19:26 aland Exp $") #include "eap_tls.h" #include <freeradius-devel/radiusd.h> #include <freeradius-devel/modules.h> /* configured values goes right here */ typedef struct eap_tls_conf { char *private_key_password; char *private_key_file; char *certificate_file; char *random_file; char *ca_path; char *ca_file; char *dh_file; char *rsa_file; char *make_cert_command; int rsa_key; int dh_key; int rsa_key_length; int dh_key_length; int verify_depth; int file_type; int include_length; /* * Always < 4096 (due to radius limit), 0 by default = 2048 */ int fragment_size; int check_crl; char *check_cert_cn; char *cipher_list; char *check_cert_issuer; int log_certificates; /* log_certificates added by Boeing, 9/26/2008 */ } EAP_TLS_CONF; /* This structure gets stored in arg */ typedef struct _eap_tls_t { EAP_TLS_CONF *conf; SSL_CTX *ctx; } eap_tls_t; #endif /* _RLM_EAP_TLS_H */ /* * rlm_eap_tls.c contains the interfaces that are called from eap * * Version: $Id: rlm_eap_tls.c,v 1.54 2008/03/22 08:31:03 aland Exp $ * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA * * Copyright 2001 hereUare Communications, Inc. <raghud(a)hereuare.com> * Copyright 2003 Alan DeKok <aland(a)freeradius.org> * Copyright 2006 The FreeRADIUS server project * *----------------------------------------------------------------------------- * Modified by The Boeing Company, August 26, 2008 * * Added log_certificates to module_config CONF_PARSER * Added additional authentication logging into cbtls_verify * Added log_full_cert function to write certificate into log file * * Copyright 2008 The Boeing Company <Michael.Ross2(a)boeing.com> *----------------------------------------------------------------------------- */ #include <freeradius-devel/ident.h> RCSID("$Id: rlm_eap_tls.c,v 1.54 2008/03/22 08:31:03 aland Exp $") #include <freeradius-devel/autoconf.h> #ifdef HAVE_OPENSSL_RAND_H #include <openssl/rand.h> #endif #include "rlm_eap_tls.h" #ifdef HAVE_SYS_STAT_H #include <sys/stat.h> #endif static CONF_PARSER module_config[] = { { "rsa_key_exchange", PW_TYPE_BOOLEAN, offsetof(EAP_TLS_CONF, rsa_key), NULL, "no" }, { "dh_key_exchange", PW_TYPE_BOOLEAN, offsetof(EAP_TLS_CONF, dh_key), NULL, "yes" }, { "rsa_key_length", PW_TYPE_INTEGER, offsetof(EAP_TLS_CONF, rsa_key_length), NULL, "512" }, { "dh_key_length", PW_TYPE_INTEGER, offsetof(EAP_TLS_CONF, dh_key_length), NULL, "512" }, { "verify_depth", PW_TYPE_INTEGER, offsetof(EAP_TLS_CONF, verify_depth), NULL, "0" }, { "CA_path", PW_TYPE_FILENAME, offsetof(EAP_TLS_CONF, ca_path), NULL, NULL }, { "pem_file_type", PW_TYPE_BOOLEAN, offsetof(EAP_TLS_CONF, file_type), NULL, "yes" }, { "private_key_file", PW_TYPE_FILENAME, offsetof(EAP_TLS_CONF, private_key_file), NULL, NULL }, { "certificate_file", PW_TYPE_FILENAME, offsetof(EAP_TLS_CONF, certificate_file), NULL, NULL }, { "CA_file", PW_TYPE_FILENAME, offsetof(EAP_TLS_CONF, ca_file), NULL, NULL }, { "private_key_password", PW_TYPE_STRING_PTR, offsetof(EAP_TLS_CONF, private_key_password), NULL, NULL }, { "dh_file", PW_TYPE_STRING_PTR, offsetof(EAP_TLS_CONF, dh_file), NULL, NULL }, { "random_file", PW_TYPE_STRING_PTR, offsetof(EAP_TLS_CONF, random_file), NULL, NULL }, { "fragment_size", PW_TYPE_INTEGER, offsetof(EAP_TLS_CONF, fragment_size), NULL, "1024" }, { "include_length", PW_TYPE_BOOLEAN, offsetof(EAP_TLS_CONF, include_length), NULL, "yes" }, { "check_crl", PW_TYPE_BOOLEAN, offsetof(EAP_TLS_CONF, check_crl), NULL, "no"}, { "check_cert_cn", PW_TYPE_STRING_PTR, offsetof(EAP_TLS_CONF, check_cert_cn), NULL, NULL}, { "cipher_list", PW_TYPE_STRING_PTR, offsetof(EAP_TLS_CONF, cipher_list), NULL, NULL}, { "check_cert_issuer", PW_TYPE_STRING_PTR, offsetof(EAP_TLS_CONF, check_cert_issuer), NULL, NULL}, { "make_cert_command", PW_TYPE_STRING_PTR, offsetof(EAP_TLS_CONF, make_cert_command), NULL, NULL}, /* log_certificates added by Boeing, 9/26/2008 */ { "log_certificates", PW_TYPE_BOOLEAN, offsetof(EAP_TLS_CONF, log_certificates), NULL, "no" }, { NULL, -1, 0, NULL, NULL } /* end the list */ }; /* * log_full_cert added by Boeing, 9/26/2008 * * Write the certificate serial number, expiration * date, issuer, and subject to the log file */ static int log_full_cert(int ok, X509 *cert) { const int ub_serial = 20; const int ub_time = 15; char serialNumber[2*ub_serial+1]; char expiration[ub_time+1]; char issuer[ub_name+1]; char subject[ub_name+1]; char *pTemp = NULL; int local_ok = ok; int i = 0; ASN1_INTEGER *pSn = NULL; ASN1_TIME *pTime = NULL; /* * Get the Serial Number and format for display */ serialNumber[0] = '\0'; pSn = X509_get_serialNumber(cert); pTemp = serialNumber; if ( pSn == NULL || pSn->length > ub_serial ) { radlog(L_AUTH, "rlm_eap_tls: Malformed Certificate Serial Number"); local_ok = 0; } else { for (i = 0; i < pSn->length; i++) { sprintf(pTemp, "%02x", (int)pSn->data[i]); pTemp += 2; } } /* * Get the Expiration Date and format for display */ expiration[0] = '\0'; pTime = X509_get_notAfter(cert); if ( pTime == NULL || pTime->length > ub_time ) { radlog(L_AUTH, "rlm_eap_tls: Malformed Certificate Expiration"); local_ok = 0; } else { strncpy(expiration, pTime->data, pTime->length); expiration[pTime->length] = '\0'; } /* * Get the Issuer */ issuer[0] = '\0'; X509_NAME_oneline(X509_get_issuer_name(cert), issuer, ub_name); issuer[0] = ' '; issuer[ub_name] = '\0'; for (i = 0; i < ub_name; i++) { if (issuer[i] == '/') { issuer[i] = ','; } } /* * Get the Subject */ subject[0] = '\0'; X509_NAME_oneline(X509_get_subject_name(cert), subject, ub_name); subject[0] = ' '; subject[ub_name] = '\0'; for (i = 0; i < ub_name; i++) { if (subject[i] == '/') { subject[i] = ','; } } /* * Log required attributes */ radlog(L_AUTH, "Certificate: Serial=%s; Expiration=%s; Issuer:%s; Subject:%s", serialNumber, expiration, issuer, subject); return local_ok; } /* * TODO: Check for the type of key exchange * like conf->dh_key */ static int load_dh_params(SSL_CTX *ctx, char *file) { DH *dh = NULL; BIO *bio; if ((bio = BIO_new_file(file, "r")) == NULL) { radlog(L_ERR, "rlm_eap_tls: Unable to open DH file - %s", file); return -1; } dh = PEM_read_bio_DHparams(bio, NULL, NULL, NULL); BIO_free(bio); if (!dh) { DEBUG2("WARNING: rlm_eap_tls: Unable to set DH parameters. DH cipher suites may not work!"); DEBUG2("WARNING: Fix this by running the OpenSSL command listed in eap.conf"); return 0; } if (SSL_CTX_set_tmp_dh(ctx, dh) < 0) { radlog(L_ERR, "rlm_eap_tls: Unable to set DH parameters"); DH_free(dh); return -1; } DH_free(dh); return 0; } /* * Before trusting a certificate, you must make sure that the * certificate is 'valid'. There are several steps that your * application can take in determining if a certificate is * valid. Commonly used steps are: * * 1.Verifying the certificate's signature, and verifying that * the certificate has been issued by a trusted Certificate * Authority. * * 2.Verifying that the certificate is valid for the present date * (i.e. it is being presented within its validity dates). * * 3.Verifying that the certificate has not been revoked by its * issuing Certificate Authority, by checking with respect to a * Certificate Revocation List (CRL). * * 4.Verifying that the credentials presented by the certificate * fulfill additional requirements specific to the application, * such as with respect to access control lists or with respect * to OCSP (Online Certificate Status Processing). * * NOTE: This callback will be called multiple times based on the * depth of the root certificate chain */ static int cbtls_verify(int ok, X509_STORE_CTX *ctx) { char subject[1024]; /* Used for the subject name */ char issuer[1024]; /* Used for the issuer name */ char common_name[1024]; char cn_str[1024]; EAP_HANDLER *handler = NULL; X509 *client_cert; SSL *ssl; int err, depth; EAP_TLS_CONF *conf; int my_ok = ok; client_cert = X509_STORE_CTX_get_current_cert(ctx); err = X509_STORE_CTX_get_error(ctx); depth = X509_STORE_CTX_get_error_depth(ctx); /* * Moved below the conf check by Boeing, 9/26/2008 */ /* *if (!my_ok) { * radlog(L_ERR,"--> verify error:num=%d:%s\n",err, * X509_verify_cert_error_string(err)); * return my_ok; *} */ /* * Retrieve the pointer to the SSL of the connection currently treated * and the application specific data stored into the SSL object. */ ssl = X509_STORE_CTX_get_ex_data(ctx, SSL_get_ex_data_X509_STORE_CTX_idx()); handler = (EAP_HANDLER *)SSL_get_ex_data(ssl, 0); conf = (EAP_TLS_CONF *)SSL_get_ex_data(ssl, 1); /* * Copied from above by Boeing, 9/26/2008 */ if (!my_ok) { radlog(L_ERR,"--> verify error:num=%d:%s\n",err, X509_verify_cert_error_string(err)); /* * Added by Boeing, 9/26/2008 * If the config file is set to log certificates * log the certificate if there has been an authentication * error */ if (conf->log_certificates) { log_full_cert(my_ok, client_cert); } return my_ok; } /* * Get the Subject & Issuer */ subject[0] = issuer[0] = '\0'; X509_NAME_oneline(X509_get_subject_name(client_cert), subject, sizeof(subject)); X509_NAME_oneline(X509_get_issuer_name(ctx->current_cert), issuer, sizeof(issuer)); subject[sizeof(subject) - 1] = '\0'; issuer[sizeof(issuer) - 1] = '\0'; /* * Get the Common Name */ X509_NAME_get_text_by_NID(X509_get_subject_name(client_cert), NID_commonName, common_name, sizeof(common_name)); common_name[sizeof(common_name) - 1] = '\0'; switch (ctx->error) { case X509_V_ERR_UNABLE_TO_GET_ISSUER_CERT: radlog(L_ERR, "issuer= %s\n", issuer); break; case X509_V_ERR_CERT_NOT_YET_VALID: case X509_V_ERR_ERROR_IN_CERT_NOT_BEFORE_FIELD: radlog(L_ERR, "notBefore="); #if 0 ASN1_TIME_print(bio_err, X509_get_notBefore(ctx->current_cert)); #endif break; case X509_V_ERR_CERT_HAS_EXPIRED: case X509_V_ERR_ERROR_IN_CERT_NOT_AFTER_FIELD: radlog(L_ERR, "notAfter="); #if 0 ASN1_TIME_print(bio_err, X509_get_notAfter(ctx->current_cert)); #endif break; } /* * If we're at the actual client cert, apply additional * checks. */ if (depth == 0) { /* * If the conf tells us to, check cert issuer * against the specified value and fail * verification if they don't match. */ if (conf->check_cert_issuer && (strcmp(issuer, conf->check_cert_issuer) != 0)) { radlog(L_AUTH, "rlm_eap_tls: Certificate issuer (%s) does not match specified value (%s)!", issuer, conf->check_cert_issuer); my_ok = 0; } /* * If the conf tells us to, check the CN in the * cert against xlat'ed value, but only if the * previous checks passed. */ if (my_ok && conf->check_cert_cn) { if (!radius_xlat(cn_str, sizeof(cn_str), conf->check_cert_cn, handler->request, NULL)) { radlog(L_ERR, "rlm_eap_tls (%s): xlat failed.", conf->check_cert_cn); /* if this fails, fail the verification */ my_ok = 0; } else { DEBUG2(" rlm_eap_tls: checking certificate CN (%s) with xlat'ed value (%s)", common_name, cn_str); if (strcmp(cn_str, common_name) != 0) { radlog(L_AUTH, "rlm_eap_tls: Certificate CN (%s) does not match specified value (%s)!", common_name, cn_str); my_ok = 0; } } } /* check_cert_cn */ /* * Added by Boeing, 9/26/2008 * If the config file is set to log certificates * log the certificate */ if (conf->log_certificates) { my_ok = log_full_cert(my_ok, client_cert); } } /* depth == 0 */ if (debug_flag > 0) { DEBUG2("chain-depth=%d, ", depth); DEBUG2("error=%d", err); DEBUG2("--> User-Name = %s", handler->identity); DEBUG2("--> BUF-Name = %s", common_name); DEBUG2("--> subject = %s", subject); DEBUG2("--> issuer = %s", issuer); DEBUG2("--> verify return:%d", my_ok); } return my_ok; } /* * Create Global context SSL and use it in every new session * * - Load the trusted CAs * - Load the Private key & the certificate * - Set the Context options & Verify options */ static SSL_CTX *init_tls_ctx(EAP_TLS_CONF *conf) { SSL_METHOD *meth; SSL_CTX *ctx; X509_STORE *certstore; int verify_mode = SSL_VERIFY_NONE; int ctx_options = 0; int type; /* * Add all the default ciphers and message digests * Create our context. */ SSL_library_init(); SSL_load_error_strings(); meth = TLSv1_method(); ctx = SSL_CTX_new(meth); /* * Identify the type of certificates that needs to be loaded */ if (conf->file_type) { type = SSL_FILETYPE_PEM; } else { type = SSL_FILETYPE_ASN1; } /* * Set the password to load private key */ if (conf->private_key_password) { SSL_CTX_set_default_passwd_cb_userdata(ctx, conf->private_key_password); SSL_CTX_set_default_passwd_cb(ctx, cbtls_password); } /* * Load our keys and certificates * * If certificates are of type PEM then we can make use * of cert chain authentication using openssl api call * SSL_CTX_use_certificate_chain_file. Please see how * the cert chain needs to be given in PEM from * openSSL.org */ if (type == SSL_FILETYPE_PEM) { if (!(SSL_CTX_use_certificate_chain_file(ctx, conf->certificate_file))) { radlog(L_ERR, "rlm_eap: SSL error %s", ERR_error_string(ERR_get_error(), NULL)); radlog(L_ERR, "rlm_eap_tls: Error reading certificate file %s", conf->certificate_file); return NULL; } } else if (!(SSL_CTX_use_certificate_file(ctx, conf->certificate_file, type))) { radlog(L_ERR, "rlm_eap: SSL error %s", ERR_error_string(ERR_get_error(), NULL)); radlog(L_ERR, "rlm_eap_tls: Error reading certificate file %s", conf->certificate_file); return NULL; } /* Load the CAs we trust */ if (!SSL_CTX_load_verify_locations(ctx, conf->ca_file, conf->ca_path)) { radlog(L_ERR, "rlm_eap: SSL error %s", ERR_error_string(ERR_get_error(), NULL)); radlog(L_ERR, "rlm_eap_tls: Error reading Trusted root CA list %s",conf->ca_file ); return NULL; } if (conf->ca_file && *conf->ca_file) SSL_CTX_set_client_CA_list(ctx, SSL_load_client_CA_file(conf->ca_file)); if (!(SSL_CTX_use_PrivateKey_file(ctx, conf->private_key_file, type))) { radlog(L_ERR, "rlm_eap: SSL error %s", ERR_error_string(ERR_get_error(), NULL)); radlog(L_ERR, "rlm_eap_tls: Error reading private key file %s", conf->private_key_file); return NULL; } /* * Check if the loaded private key is the right one */ if (!SSL_CTX_check_private_key(ctx)) { radlog(L_ERR, "rlm_eap_tls: Private key does not match the certificate public key"); return NULL; } /* * Set ctx_options */ ctx_options |= SSL_OP_NO_SSLv2; ctx_options |= SSL_OP_NO_SSLv3; /* * SSL_OP_SINGLE_DH_USE must be used in order to prevent * small subgroup attacks and forward secrecy. Always * using * * SSL_OP_SINGLE_DH_USE has an impact on the computer * time needed during negotiation, but it is not very * large. */ ctx_options |= SSL_OP_SINGLE_DH_USE; /* * SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS to work around issues * in Windows Vista client. * http://www.openssl.org/~bodo/tls-cbc.txt * http://www.nabble.com/(RADIATOR)-Radiator-Version-3.16-released-t2600070.ht… */ ctx_options |= SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS; SSL_CTX_set_options(ctx, ctx_options); /* * TODO: Set the RSA & DH * SSL_CTX_set_tmp_rsa_callback(ctx, cbtls_rsa); * SSL_CTX_set_tmp_dh_callback(ctx, cbtls_dh); */ /* * set the message callback to identify the type of * message. For every new session, there can be a * different callback argument. * * SSL_CTX_set_msg_callback(ctx, cbtls_msg); */ /* Set Info callback */ SSL_CTX_set_info_callback(ctx, cbtls_info); /* * Check the certificates for revocation. */ #ifdef X509_V_FLAG_CRL_CHECK if (conf->check_crl) { certstore = SSL_CTX_get_cert_store(ctx); if (certstore == NULL) { radlog(L_ERR, "rlm_eap: SSL error %s", ERR_error_string(ERR_get_error(), NULL)); radlog(L_ERR, "rlm_eap_tls: Error reading Certificate Store"); return NULL; } X509_STORE_set_flags(certstore, X509_V_FLAG_CRL_CHECK); } #endif /* * Set verify modes * Always verify the peer certificate */ verify_mode |= SSL_VERIFY_PEER; verify_mode |= SSL_VERIFY_FAIL_IF_NO_PEER_CERT; verify_mode |= SSL_VERIFY_CLIENT_ONCE; SSL_CTX_set_verify(ctx, verify_mode, cbtls_verify); if (conf->verify_depth) { SSL_CTX_set_verify_depth(ctx, conf->verify_depth); } /* Load randomness */ if (!(RAND_load_file(conf->random_file, 1024*1024))) { radlog(L_ERR, "rlm_eap: SSL error %s", ERR_error_string(ERR_get_error(), NULL)); radlog(L_ERR, "rlm_eap_tls: Error loading randomness"); return NULL; } /* * Set the cipher list if we were told to */ if (conf->cipher_list) { if (!SSL_CTX_set_cipher_list(ctx, conf->cipher_list)) { radlog(L_ERR, "rlm_eap_tls: Error setting cipher list"); return NULL; } } return ctx; } /* * Detach the EAP-TLS module. */ static int eaptls_detach(void *arg) { EAP_TLS_CONF *conf; eap_tls_t *inst; inst = (eap_tls_t *) arg; conf = inst->conf; if (conf) { memset(conf, 0, sizeof(*conf)); free(inst->conf); inst->conf = NULL; } if (inst->ctx) SSL_CTX_free(inst->ctx); inst->ctx = NULL; free(inst); return 0; } /* * Attach the EAP-TLS module. */ static int eaptls_attach(CONF_SECTION *cs, void **instance) { EAP_TLS_CONF *conf; eap_tls_t *inst; /* Store all these values in the data structure for later references */ inst = (eap_tls_t *)malloc(sizeof(*inst)); if (!inst) { radlog(L_ERR, "rlm_eap_tls: out of memory"); return -1; } memset(inst, 0, sizeof(*inst)); /* * Parse the config file & get all the configured values */ conf = (EAP_TLS_CONF *)malloc(sizeof(*conf)); if (conf == NULL) { free(inst); radlog(L_ERR, "rlm_eap_tls: out of memory"); return -1; } memset(conf, 0, sizeof(*conf)); inst->conf = conf; if (cf_section_parse(cs, conf, module_config) < 0) { eaptls_detach(inst); return -1; } /* * The EAP RFC's say 1020, but we're less picky. */ if (conf->fragment_size < 100) { radlog(L_ERR, "rlm_eap_tls: Fragment size is too small."); eaptls_detach(inst); return -1; } /* * The maximum size for a RADIUS packet is 4096, * minus the header (20), Message-Authenticator (18), * and State (18), etc. results in about 4000 bytes of data * that can be devoted *solely* to EAP. */ if (conf->fragment_size > 4000) { radlog(L_ERR, "rlm_eap_tls: Fragment size is too large."); eaptls_detach(inst); return -1; } /* * Account for the EAP header (4), and the EAP-TLS header * (6), as per Section 4.2 of RFC 2716. What's left is * the maximum amount of data we read from a TLS buffer. */ conf->fragment_size -= 10; /* * This magic makes the administrators life HUGELY easier * on initial deployments. * * If the server starts up in debugging mode, AND the * bootstrap command is configured, AND it exists, AND * there is no server certificate */ if (conf->make_cert_command && (debug_flag >= 2)) { struct stat buf; if ((stat(conf->make_cert_command, &buf) == 0) && (stat(conf->certificate_file, &buf) < 0) && (errno == ENOENT) && (radius_exec_program(conf->make_cert_command, NULL, 1, NULL, 0, NULL, NULL, 0) != 0)) { eaptls_detach(inst); return -1; } } /* * Initialize TLS */ inst->ctx = init_tls_ctx(conf); if (inst->ctx == NULL) { eaptls_detach(inst); return -1; } if (load_dh_params(inst->ctx, conf->dh_file) < 0) { eaptls_detach(inst); return -1; } *instance = inst; return 0; } /* * Send an initial eap-tls request to the peer. * * Frame eap reply packet. * len = header + type + tls_typedata * tls_typedata = flags(Start (S) bit set, and no data) * * Once having received the peer's Identity, the EAP server MUST * respond with an EAP-TLS/Start packet, which is an * EAP-Request packet with EAP-Type=EAP-TLS, the Start (S) bit * set, and no data. The EAP-TLS conversation will then begin, * with the peer sending an EAP-Response packet with * EAP-Type = EAP-TLS. The data field of that packet will * be the TLS data. * * Fragment length is Framed-MTU - 4. * * http://mail.frascone.com/pipermail/public/eap/2003-July/001426.html */ static int eaptls_initiate(void *type_arg, EAP_HANDLER *handler) { int status; tls_session_t *ssn; eap_tls_t *inst; VALUE_PAIR *vp; int client_cert = TRUE; int verify_mode = 0; inst = (eap_tls_t *)type_arg; /* * If we're TTLS or PEAP, then do NOT require a client * certificate. * * FIXME: This should be more configurable. */ if (handler->eap_type != PW_EAP_TLS) { vp = pairfind(handler->request->config_items, PW_EAP_TLS_REQUIRE_CLIENT_CERT); if (!vp) { client_cert = FALSE; } else { client_cert = vp->vp_integer; } } /* * Every new session is started only from EAP-TLS-START. * Before Sending EAP-TLS-START, open a new SSL session. * Create all the required data structures & store them * in Opaque. So that we can use these data structures * when we get the response */ ssn = eaptls_new_session(inst->ctx, client_cert); if (!ssn) { return 0; } /* * Verify the peer certificate, if asked. */ if (client_cert) { DEBUG2(" rlm_eap_tls: Requiring client certificate"); verify_mode = SSL_VERIFY_PEER; verify_mode |= SSL_VERIFY_FAIL_IF_NO_PEER_CERT; verify_mode |= SSL_VERIFY_CLIENT_ONCE; } SSL_set_verify(ssn->ssl, verify_mode, cbtls_verify); /* * Create a structure for all the items required to be * verified for each client and set that as opaque data * structure. * * NOTE: If we want to set each item sepearately then * this index should be global. */ SSL_set_ex_data(ssn->ssl, 0, (void *)handler); SSL_set_ex_data(ssn->ssl, 1, (void *)inst->conf); ssn->length_flag = inst->conf->include_length; /* * We use default fragment size, unless the Framed-MTU * tells us it's too big. Note that we do NOT account * for the EAP-TLS headers if conf->fragment_size is * large, because that config item looks to be confusing. * * i.e. it should REALLY be called MTU, and the code here * should figure out what that means for TLS fragment size. * asking the administrator to know the internal details * of EAP-TLS in order to calculate fragment sizes is * just too much. */ ssn->offset = inst->conf->fragment_size; vp = pairfind(handler->request->packet->vps, PW_FRAMED_MTU); if (vp && ((vp->vp_integer - 14) < ssn->offset)) { /* * Discount the Framed-MTU by: * 4 : EAPOL header * 4 : EAP header (code + id + length) * 1 : EAP type == EAP-TLS * 1 : EAP-TLS Flags * 4 : EAP-TLS Message length * (even if conf->include_length == 0, * just to be lazy). * --- * 14 */ ssn->offset = vp->vp_integer - 14; } handler->opaque = ((void *)ssn); handler->free_opaque = session_free; DEBUG2(" rlm_eap_tls: Initiate"); /* * PEAP-specific breakage. */ if (handler->eap_type == PW_EAP_PEAP) { /* * As it is a poorly designed protocol, PEAP uses * bits in the TLS header to indicate PEAP * version numbers. For now, we only support * PEAP version 0, so it doesn't matter too much. * However, if we support later versions of PEAP, * we will need this flag to indicate which * version we're currently dealing with. */ ssn->peap_flag = 0x00; /* * PEAP version 0 requires 'include_length = no', * so rather than hoping the user figures it out, * we force it here. */ ssn->length_flag = 0; } /* * TLS session initialization is over. Now handle TLS * related handshaking or application data. */ status = eaptls_start(handler->eap_ds, ssn->peap_flag); DEBUG2(" rlm_eap_tls: Start returned %d", status); if (status == 0) return 0; /* * The next stage to process the packet. */ handler->stage = AUTHENTICATE; return 1; } /* * Do authentication, by letting EAP-TLS do most of the work. */ static int eaptls_authenticate(void *arg UNUSED, EAP_HANDLER *handler) { eaptls_status_t status; tls_session_t *tls_session = (tls_session_t *) handler->opaque; DEBUG2(" rlm_eap_tls: Authenticate"); status = eaptls_process(handler); DEBUG2(" eaptls_process returned %d\n", status); switch (status) { /* * EAP-TLS handshake was successful, return an * EAP-TLS-Success packet here. */ case EAPTLS_SUCCESS: break; /* * The TLS code is still working on the TLS * exchange, and it's a valid TLS request. * do nothing. */ case EAPTLS_HANDLED: return 1; /* * Handshake is done, proceed with decoding tunneled * data. */ case EAPTLS_OK: DEBUG2(" rlm_eap_tls: Received unexpected tunneled data after successful handshake."); #ifndef NDEBUG if ((debug_flag > 2) && fr_log_fp) { unsigned int i; unsigned int data_len; unsigned char buffer[1024]; data_len = (tls_session->record_minus)(&tls_session->dirty_in, buffer, sizeof(buffer)); log_debug(" Tunneled data (%u bytes)\n", data_len); for (i = 0; i < data_len; i++) { if ((i & 0x0f) == 0x00) fprintf(fr_log_fp, " %x: ", i); if ((i & 0x0f) == 0x0f) fprintf(fr_log_fp, "\n"); fprintf(fr_log_fp, "%02x ", buffer[i]); } fprintf(fr_log_fp, "\n"); } #endif eaptls_fail(handler->eap_ds, 0); return 0; break; /* * Anything else: fail. */ default: return 0; } /* * Success: Return MPPE keys. */ eaptls_success(handler->eap_ds, 0); eaptls_gen_mppe_keys(&handler->request->reply->vps, tls_session->ssl, "client EAP encryption"); return 1; } /* * The module name should be the only globally exported symbol. * That is, everything else should be 'static'. */ EAP_TYPE rlm_eap_tls = { "eap_tls", eaptls_attach, /* attach */ eaptls_initiate, /* Start the initial request */ NULL, /* authorization */ eaptls_authenticate, /* authentication */ eaptls_detach /* detach */ };

3 16

Sponsored development rlm_ldap and ocsp
by Stephan Jäger 02 Sep '10

02 Sep '10

Hello developers, for a projekt we need to keep the radius clients in LDAP. As this is not yet possible with rlm_ldap we would like to sponsor this development. Our requirements so far are: - created code should go in the official freeradius code base. - official invoice for our purchasing department. - LDAP schema oder schema enhancement for the clients with required attributes, ideally with oids from the freeradius projekt. Should be pretty easy as most of the needed code is already in rlm_sql and rlm_ldap. Please let me know what you think about this, and if someone is willing to do this please mail me your offer stating your timeframe and price. Additionally we might be interested to sponsor OCSP support for EAP-TLS, so i'd be happy to receive offers on this one too, though this one is not yet decided on internally. Kind regards, Stephan Jaeger EWE TEL GmbH

5 20

Can some one explain me about PDFID and SDFID
by aleem 31 Aug '10

31 Aug '10

Hi All, I Am a newbie , dont know much about the WiMAX.Can any one explain me in detail about the packet data flow ID(PDFID) and service data flow ID(SDFID) in WiMAX. One more thing can we send multiple Service-ID attributes in a single request/response.If yes how? Any kind of help would be aprreciated.Thanks in advance. Regards, Aleem. -- View this message in context: http://freeradius.1045715.n5.nabble.com/Can-some-one-explain-me-about-PDFID… Sent from the FreeRadius - Dev mailing list archive at Nabble.com.

1 0

autoconf uses explicit path
by Scott Neugroschl 27 Aug '10

27 Aug '10

Any autoconf wizards out there? The autoconf scripts for freeradius-client 1.1.8 put the absolute path of where it was run from into the makefiles. Example: [N1:~/freeradius]> grep scottn Makefile ACLOCAL = ${SHELL} /users/scottn/freeradius/missing --run aclocal-1.9 AMTAR = ${SHELL} /users/scottn/freeradius/missing --run tar AUTOCONF = ${SHELL} /users/scottn/freeradius/missing --run autoconf AUTOHEADER = ${SHELL} /users/scottn/freeradius/missing --run autoheader AUTOMAKE = ${SHELL} /users/scottn/freeradius/missing --run automake-1.9 MAKEINFO = ${SHELL} /users/scottn/freeradius/missing --run makeinfo install_sh = /users/scottn/freeradius/install-sh If you wish to move the freeradius-client directory tree elsewhere, you have to rerun autoconf. I am afraid that I'm not familiar enough with autoconf to correct this. Alan, if you want, I can submit this as a bug. ScottN

2 1

Problems with compiling rlm_sql_oracle module
by Dan Lark 18 Aug '10

18 Aug '10

I am desperately trying to get FR compile from the 2.1.9 source tree. In looking through the configure.in file for the rlm_sql_oracle driver from src/modules/rlm_sql/drivers/rlm_sql_oracle, it would appear that it is either looking for the 8i/9i full blown client install, or a 10g/11g full install or a specific version of the InstantClient, 10.1.0.3. (This particular InstantClient release is no longer available, btw.) So now my questions: 1. What would the best way for me to make the configure script best recognize the release of the client software that I have installed? 2. The configure.in (and therefore the configure script) are hardcoded to a specific release and architecture. In my case, I am trying to do an x86_64 build. Is there a way to make this more generic? 3. I have tried doing some stuff with the configure.in file to make it work properly with my release of the InstantClient, but running autoconf against the modified configure.in file creates a configure script with syntax errors. It would appear that the original configure script is based on autoconf 2.59, whereas the installed version that I have 2.61. Would that make a difference? 4. Based on my observations, re. the InstantClient version information, it would appear that this code is probably stale. Is this currently even being worked on? Any suggestions? -dan ______________________ Dan Lark, RHCE Systems Engineer, Hargray Communications dan.lark(a)htc.hargray.com Office: 843-706-1819 | Cell: 843-227-2263 7 Arley Way, Suite 200 (2nd Floor) P.O. Box 3380 Bluffton, SC 29910

3 7

Need to Develope Radius client with All EAP Method support on windows
by Suresh kumar 16 Aug '10

16 Aug '10

Hi, All I have to develope redius client to full fill the follwing requrment. Plz which radius client lib/stack i should use for windows. simulate Radius and dot1x type of client flows The radius tool should allow us to specify authentication type and it should support all EAP methods such as MSCHAP/V2 etc. Tool should allow us to add custom attributes to authentication request and accounting request. Tool should allow end users to specify port information. It should be able to specify retries and time out. It should be able to simulate concurrent radius clients. T

1 0

Database disconnection caused server exiting abnormal
by yangj 13 Aug '10

13 Aug '10

To:freeradius developers Hello! We met a problem as the title said. The entironment is: *Solaris 10+Oracle 10g *freeradius threads number is :100 *Database connection number is :120 We use LoadRunner to deal with the transction, With 300 transactions(authorize and accounting) per secound. the server was running normal. And then we disconnected the Freeradius Server and the Database server . LoadRunner was still working. After 20 minutes later, the Freeradius Server exit . We searched with GDB.and found that: (1)the source where caused the server exit: src/lib/event.c:fr_event_loop() ---- rcode = select(el->maxfd + 1, &read_fds, NULL, NULL, wake); if ((rcode < 0) && (errno != EINTR)) { <--rcode = -1 .errno = 22 el->dispatch = 0; return 0; } ---- (2)At the same function:fr_event_loop() Variable 'when.tv_usec' value is abnormal(ex:when.tv_usec= ). And may cause the above select() exit abnormal. But I don't know why and when the when.tv_usec is set with the abnormal value. Please help me to analyse the reason and How to fix the bug? The attachment is the radius.log and the GDB log. Thanks ---------- yangj(a)ffcs.cn

2 1