In Part I, I have shown what I did to get the build and installation going. In Part II, I will show what steps I took to get a simplest test like the following done: A EAP-MD5 test that involves an OpenDiameter server (aaad), an OpenDiameter client (nasd), and a EAP-MD5 client (pacd) talking to nasd using PANA. All three parties reside on one single host.
Simple as the test is, a lot of work is needed in OpenDiameter's case, as we will see soon.
Some Background Information
The Diameter Protocol
Diameter is intended to be the next generation AAA protocol that replaces RADIUS. Wheather and when it will replace RADIUS is not for me to discuss here, but it has already found uses in a lot of applications, notably in 3GPP and WiMAX infrastructures. I will assume that you are familiar with AAA or RADIUS and I will just try to point out a few concepts that is unique in Diameter and we will encounter in our little test.
Peer Diameter nodes like server and NAS talk connection-oriented protocols of TCP or SCTP, as opposed to the connectionless protocol of UDP used by RADIUS. The connections are also maintained and monitored by using watch-dog messages. As a result, in Diameter, "peer" is used to refer to the ends of a connection, and please don't confuse it with the "peer" used in EAP which is really the supplicant. One Diameter node could have multiple peers and a peer table usually exists in its configurations.
Applications Diameter introduces the concept of application which is a major mechanism for extensibility. Each application is identified by a unique 32 bit application ID. EAP is implemented as an application too in Diameter and it is assigned an ID of value 5.
AVPs and dictionary Just like RADIUS, Diameter's function largely involves carrying attributes around. However, the attribute space is much bigger in Diameter and the structure to carry them is a new design that is called Attribute-Value Pair, or AVP. Naturally, when there are attributes, there will be dictionaries for the nodes to look them up.
NAI, realm routing Network Access identifier (NAI) is the standard form of user Identity. It normally contains a realm part, like firstname.lastname@example.org. The realm part can be used by Diameter nodes to decicde which peer/domain it should be sent to, hence the concept of "realm routing". A routing table can be found in a Diameter node that contains information about for what realm what action should be taken and which peer should be used. It is most commonly used by Diameter agents that does proxy or relay, and since our little test doesn't involve an agent, it shouldn't matter. However, in Opendiameter's nasd implementation, it also uses the realm to decide which Diameter server should it forward messages to, and that is why I mention it here.
PANA PANA stands for Protocol for carrying Authentication for Network Access. The idea is like this. In a 3-party EAP model, the authenticator acts as an pass-through device that relays authentication requests from a supplicant to an authentication server. A EAP packet doesn't travel itself and needs to be carried in some other transport protocol. RADIUS and Diameter can both be used to carry EAP from the authenticator to the authentication server. However, on the other side of authenticator that faces the supplicant (the Access Side) a different carrier protocol is needed. Often an authenticator that directly faces a layer 2 network uses a layer 2 carrier protocol like 802.1x(EAPOL) and PPP. However, if an authenticator sits deeper inside a network and does not face the edge, it might make sense to use an uppler layer transport to do the job. PANA is such a carrier protocol that is IP based and operates in layer 3. OpenDiameter's nasd uses PANA, and in PANA terminology nasd is a PAA (PANA authentication agent) which accepts EAP requests from supplicant. The other end of PANA connection is called PaC (PANA Client). In OpenDiameter, the pacd program acts as a combined PaC and supplicant.
Binaries from OpenDiameter
As already mentioned, we will be using the server binary (aaad), authenticator/ NAS binary (nasd) and the PANA client + Supplicant binary (pacd). Opendiameter also generates a series of test clients and test servers, but we will not use them here.
OpenDiameter Config files
OpenDiameter's configuration files are mostly .xml files that are in XML format and each .xml file has an companion .dtd file for Document Type Definition (DTD).
Sorry, more compilation please
I must admit that in Part I, which should take care of all compilations, I omitted one inconconvenient fact that requires a quick re-complilation. The reason is that it only affects nasd which may not be essential for a lot of uses where people are only interested in the server.
In the code that came with 1.0.7-i, nasd somehow has commented out lines related to Diameter EAP application in its initialization part. That not only makes nasd not usable for Diameter EAP, but also pretty much makes it unsable for anything. The reason is that nasd defines its peer table inside of the nasd_diameter_eap.xml (which is weird by itself since EAP is just one application) and the file won't even be read with those lines commented out.
so, we need to apply the following simple diff and redo "make; make install" in the opendiameter root directory.
+++ applications/nas/src/nasd_main.cxx 2008-04-05 06:52:44.000000000 -0700 @@ -51,17 +51,17 @@ int main(int argc, char **argv) /// Node writers MUST add thier initializer /// instance here NASD_PanaInitializer apPanaInit; - //NASD_DiameterEapInitializer aaaDiameterEapInit; + NASD_DiameterEapInitializer aaaDiameterEapInit; NASD_EapBackendInitializer aaaEapBackendInit; NASD_PolicyScriptInitializer plcyScriptInit; std::string strApPanaName("pana"); - //std::string strAaaDiameterEapName("diameter_eap"); + std::string strAaaDiameterEapName("diameter_eap"); std::string strAaaEapBackendName("local_eap_auth"); std::string strPlcyScriptName("script"); NASD_CnInitializer_I->Register(strApPanaName, apPanaInit); - //NASD_CnInitializer_I->Register(strAaaDiameterEapName, aaaDiameterEapInit); + NASD_CnInitializer_I->Register(strAaaDiameterEapName, aaaDiameterEapInit); NASD_CnInitializer_I->Register(strAaaEapBackendName, aaaEapBackendInit);
Crypto related Library linking problems
On some systems I have also seen problems during build phase when linking to crypto libraries, and I suspect it has something to do with system's openssl libcrypto.a. Since it is not universal, I omitted them, but now I feel I probably should point them out too. You can omit this if you didn't have any problem following Part I.
- complaints about dlopen() when compiling ACE You could work-around it by this quick hack on ACE_wrappers/configure file:
--- configure.orig 2008-04-04 12:49:15.000000000 -0700 +++ configure 2008-04-05 05:09:15.000000000 -0700 @@ -6865,7 +6865,7 @@ fi - ace_TLS_LIBS="-lssl -lcrypto" + ace_TLS_LIBS="-lssl -lcrypto -ldl"
- linking errors about MD5 when building OpenDiameter I also have a quick-n-dirty patch for this. Since it involves a few files, I have listed it in the appendix part of this article. Again, don't bother if you didn't have problem in Part I.
Update Shared Library Information
When use your binaries after first time building opendiameter, it might fail complaining about libACE. That is caused by an un-updated shared-library database on the system after new libraries are generated. I fixed it by this:
echo "/usr/local/lib/" > /etc/ld.so.conf.d/local_lib.conf /sbin/ldconfig -v
Fix the configurations for aaad and nasd
We did do "make install" before, but unfortunately there are mutliple things that are missing in the process and we need a do-over.
The binaries of aaad and nasd are installed correctly at /usr/local/bin. There are configuration files copied at /usr/local/etc/opendiameter, but I discovered that a few issues exist:
- missing .dtd files for certain .xml files
- nasd has a file that includes a wrong .dtd file name.
- the include path to a dictionary file are all set to relative path which makes the binaries only able to be invoked from the parent directory of the "config" directory, otherwise the dictionary files could not be found by them.
- there are inconsistencies between dictionary files used by peers.
- nasd's eap dictionary file is completely outdated.
I suggest to follow the following steps to fix those issues:
- Step 1 copy over aaa and nasd config files manually from your OpenDiameter source tree to /etc/opendiameter. Even though the default installation goes to /usr/local/etc/opendiameter, I found there seems to be assumptions in the code about the path /etc/opendiameter. So to avoid issues, I suggest directly go to /etc/opendiameter instead.
mkdir -p /etc/opendiameter/aaa/ mkdir -p /etc/opendiameter/nas/ cp -r ./applications/aaa/config/ /etc/opendiameter/aaa/ cp -r ./applications/nas/config/ /etc/opendiameter/nas/ cp -r ./applications/nas/scripts /etc/opendiameter/nas/
- Step 2 change paths to dictionary files to be absolute so can be invoked from anywhere.
change the following files for aaad and nasd respectively.
Note: I have moved the Diff to the end in the Appendix section because there are some characters in the diff that will make the posting system unhappy and ruin the rest of the formating.
- Step 3
correct nasd_pana_dictionary.xml file which points to wrong .dtd file:
--- nasd_pana_dictionary.xml.orig 2008-04-05 15:36:40.000000000 -0700 +++ nasd_pana_dictionary.xml 2008-04-05 15:39:02.000000000 -0700 @@ -1,5 +1,5 @@ <?xml version="1.0" encoding="UTF-8"?> -<!DOCTYPE dictionary SYSTEM "dictionary.dtd"> +<!DOCTYPE dictionary SYSTEM "nasd_pana_dictionary.dtd"> <dictionary>
- Step 4
The nasd_diameter_eap_dictionary.xml file is outdated and copy a correct one from source tree:
cp /usr/local/src/opendiameter-1.0.7-i/libdiameter/config/dictionary.xml /etc/opendiameter/nas/config/nasd_diameter_eap_dictionary.xml
then update its 2nd line from this
<!DOCTYPE dictionary SYSTEM "dictionary.dtd">
to be like this:
<!DOCTYPE dictionary SYSTEM "nasd_diameter_eap_dictionary.dtd">
fake host names
In our test, we will be using a single local host to simulate a 3-party network environment. To get things working, I am faking different domain names for nasd and aaad.
I am giving aaad the name localaaa.localdomain1.net, and nasd the name localnas.localdomain2.net and I added the following two entries in the local /etc/hosts file:
127.0.0.1 localaaa localaaa.localdomain1.net 127.0.0.1 localnas localnas.localdomain2.net
Get aaad and nasd talk to each first
Now, with all those setups, aaad and nasd should be able to talk to each other as peers without EAP in picture. All we need to do is to tell them who themselves are and who their peer is.
in aaad, configure it its identity of localaaa.localdomain1.net, and its peer localnas.localdomain2.net. Apply the following change to file aaad_diameter_server.xml.
--- aaad_diameter_server.xml.orig 2008-04-05 15:14:33.000000000 -0700 +++ aaad_diameter_server.xml 2008-04-05 15:29:07.000000000 -0700 @@ -24,11 +24,11 @@ </vendor_specific_application_id> </general> <parser> - <dictionary>config/aaad_diameter_dictionary.xml</dictionary> + <dictionary>/etc/opendiameter/aaa/config/aaad_diameter_dictionary.xml</dictionary> </parser> <transport_mngt> - <identity>server.isp.net</identity> - <realm>isp.net</realm> + <identity>localaaa.localdomain1.net</identity> + <realm>localdomain1.net</realm> <tcp_listen_port>1812</tcp_listen_port> <sctp_listen_port>1813</sctp_listen_port> <use_ipv6>0</use_ipv6> @@ -38,15 +38,16 @@ <request_retransmission_interval>10</request_retransmission_interval> <max_request_retransmission_count>3</max_request_retransmission_count> <receive_buffer_size>2048</receive_buffer_size> - <advertised_hostname>server1.isp.net</advertised_hostname> + <advertised_hostname>localaaa.localdomain1.net</advertised_hostname> <peer_table> <expiration_time>1</expiration_time> <peer> - <hostname>nas.access1.net</hostname> + <hostname>localnas.localdomain2.net</hostname> <port>1811</port> <use_sctp>0</use_sctp> <tls_enabled>0</tls_enabled> </peer> + </peer_table> <route_table> <expire_time>0</expire_time>
Note that I also commented out some original sample peers for simplicity.
For nasd, update the file nasd_diameter_eap.xml similiarly: use localnas.localdomain2.net as identity and use localaaa.localdomain1.net as peer.
@@ -24,11 +25,11 @@ </vendor_specific_application_id> </general> <parser> - <dictionary>config/nasd_diameter_eap_dictionary.xml</dictionary> + <dictionary>/etc/opendiameter/nas/config/nasd_diameter_eap_dictionary.xml</dictionary> </parser> <transport_mngt> - <identity>nas.access1.net</identity> - <realm>access1.net</realm> + <identity>localnas.localdomain2.net</identity> + <realm>localdomain2.net</realm> <tcp_listen_port>1810</tcp_listen_port> <sctp_listen_port>1811</sctp_listen_port> <use_ipv6>0</use_ipv6> @@ -41,7 +42,7 @@ <peer_table> <expiration_time>1</expiration_time> <peer> - <hostname>server.isp.net</hostname> + <hostname>localaaa.localdomain1.net</hostname> <port>1812</port> <use_sctp>0</use_sctp> <tls_enabled>0</tls_enabled>
- with those changes, now start aaad and nasd by simply invoking them (as root) without any parameters (They don't have much command line options anyway). You should see constant messages about watch-dog if the connections are successful, like this:
(7867|3038682000) Watchdog msg from [localaaa.localdomain1.net.localdomain1.net], state=1207434903, time=1207435199
I suggest that you start aaad before nasd. Based on my experience, on different systems, if doing the other way around, it either takes longer or nasd crashes. :)
Fix up pacd
The files installed at /usr/local/etc/opendiameter/pana don't seem to be usable by pacd either, so I decided not to use them at all. Unfortunately, I didn't find any proper candiate from the source tree. The best I came up with was an ugly solution (appologies!) by downloading from the latest opendiameter repository with some modifications.
- go to your source tree's applications/pana directory, and this is where the pacd binary is at.
- use svn to download just the pana config directory from the opendiameter repository:
svn co https://diameter.svn.sourceforge.net/svnroot/diameter/cplusplus/applications/pana/config
(if you don't have svn installed, it is easy. For example, in ubuntu, do sudo apt-get install subversion)
- The newer dictionary file is not exactly consistent with the 1.0.7-i's version used by nasd, and that will cause problem. In theory, the newer ones should be latest, but to avoid trouble, I reverted the newer one to older one instead. Just copy the one used by nasd, and update its correponding .dtd file line.
cp /etc/opendiameter/nas/config/nasd_pana_dictionary.xml config/pana_dictionary.xml
and update its 2nd line to:
<!DOCTYPE dictionary SYSTEM "pana_dictionary.dtd">
Configure aaad, nasd and pacd for the final EAP-MD5 test
Finally, after all the hassle, we are ready to do some "true" configurations.
- enable eap application in nasd and aaad. As I mentioned, EAP is implemented as an application in Diameter. And during initial connection establishment, peers will exchange capability information (in CER/CEA messages) and see if both support some common applications. For our test, we must enable both aaad and nasd to support EAP.s EAP applicaiton's assigned ID is 5. Just add the line
in both aaad_diameter_server.xml and nasd_diameter_eap.xml after the line of
- User information
I will use email@example.com for our EAP-MD5 test. Note that I assigned it to be in the same domain as our Diameter server.
- aaad part side
Apply this change:
+++ aaad_user_db.xml 2008-04-05 17:57:43.000000000 -0700 @@ -41,12 +41,20 @@ file. --> + <user_entry> + <name_match>firstname.lastname@example.org</name_match> + <eap_method>md5</eap_method> + <md5> + <password_type>flat</password_type> + <secret>12345</secret> + </md5> + </user_entry> <user_entry> - <name_match>user</name_match> + <name_match>dont_match_anyone</name_match> - <username>user1</username> + <username>email@example.com</username>
Password Both sides already default to use password "12345", so that doesn't need to be changed.
- Add a route entry for nasd
As I mentioned before, nasd uses the realm part of NAI to determine peer, and for our test, I added the following route in the route table of nasd's nasd_diameter_eap.xml file:
<route> <realm>localdomain1.net</realm> <role>1</role> <redirect_usage>0</redirect_usage> <application> <application_id>5</application_id> <vendor_id>0</vendor_id> <peer_entry> <server>localaaa.localdomain1.net</server> <metric>1</metric> </peer_entry> </application> </route>
That tells nasd that, for requests of the realm localdomain1.net, forward it to peer localaaa.localdomain1.net.
- Make the requirement of User-Name avp consistent between aaa and nasd:
The dictionary files used by peers are supposed to be consistent. There is one little inconsistency between them that will cause problem. In nasd's nasd_diameter_eap_dictionary.xml file, the AVP User-Name is defined as mandotory but it is not the case in aaad. Remove the mandotory part from it and make them consistent.
line 734 of nasd_diameter_eap_dictionary.xml:
<avp name="User-Name" code="1" mandatory="must"> <type type-name="UTF8String"/>
line 740 of aaad_diameter_dictionary.xml:
<avp name="User-Name" code="1"> <type type-name="UTF8String"/> </avp>
- The Final Go!
Finally, we are ready.
Restart your aaad and nasd like before, and make sure they have established connections.
Now start pacd from the applications/pana directory like this: (as root)
./pacd -f config/pana_setup.xml
And if you see messages from pacd window like this, then you are good.
Peer: Success received. Peer: Success. Authentication success at peer Welcome to the world, firstname.lastname@example.org !!!
I don't know about you, but I feel tired by now after all this, and that is just a EAP-MD5 test that can't be made simpler. As I stated before, I am new to OpenDiameter myself and my way of doing things might not be the right way, so be cautioned there! I am also sorry that at the point I don't have much more to contribute, but I do hope this little two-part article can help other OpenDiameter beginners. If it does help you to get started, I encourage you to contribute back by sharing your tricks too.
In my opinion, OpenDiameter does not seems to be in a mature stage yet. It suits to serve as a starting point to develop your own applications or to conduct your Diameter tests in the labs, but I am not sure about more serious deployments. Again, I could be totally off, but if that is the case, I hope the true experts could step up and provide more documentations.
I am also very much looking forward to the next release (1.0.8) which is delayed. And I hope all those issues I have met, if they are real, will be fixed and more importantly, more documents please!
 OpenDiameter README file from 1.0.7-i source package
 IETF PANA Draft
Appendix for MD5 link problem
Please note that, since this (ugly) patch affects the Makefile.in files, you will need to redo ./configure for the OpenDiameter after applying the changes.
--- ./libdiameternasreq/Makefile.in.orig 2008-04-05 13:08:04.000000000 -0700 +++ ./libdiameternasreq/Makefile.in 2008-04-05 13:26:32.000000000 -0700 @@ -381,12 +381,12 @@ test/client_test.$(OBJEXT): test/$(am__d test/$(DEPDIR)/$(am__dirstamp) client_test$(EXEEXT): $(client_test_OBJECTS) $(client_test_DEPENDENCIES) @rm -f client_test$(EXEEXT) - $(CXXLINK) $(client_test_LDFLAGS) $(client_test_OBJECTS) $(client_test_LDADD) $(LIBS) + $(CXXLINK) $(client_test_LDFLAGS) $(client_test_OBJECTS) $(client_test_LDADD) $(LIBS) -lcrypto test/server_test.$(OBJEXT): test/$(am__dirstamp) \ test/$(DEPDIR)/$(am__dirstamp) server_test$(EXEEXT): $(server_test_OBJECTS) $(server_test_DEPENDENCIES) @rm -f server_test$(EXEEXT) - $(CXXLINK) $(server_test_LDFLAGS) $(server_test_OBJECTS) $(server_test_LDADD) $(LIBS) + $(CXXLINK) $(server_test_LDFLAGS) $(server_test_OBJECTS) $(server_test_LDADD) $(LIBS) -lcrypto mostlyclean-compile: -rm -f *.$(OBJEXT) --- ./applications/nas/Makefile.in.orig 2008-04-05 13:08:11.000000000 -0700 +++ ./applications/nas/Makefile.in 2008-04-05 13:26:40.000000000 -0700 @@ -331,7 +331,7 @@ src/nasd_diameter_eap.$(OBJEXT): src/$(a src/$(DEPDIR)/$(am__dirstamp) nasd$(EXEEXT): $(nasd_OBJECTS) $(nasd_DEPENDENCIES) @rm -f nasd$(EXEEXT) - $(CXXLINK) $(nasd_LDFLAGS) $(nasd_OBJECTS) $(nasd_LDADD) $(LIBS) + $(CXXLINK) $(nasd_LDFLAGS) $(nasd_OBJECTS) $(nasd_LDADD) $(LIBS) -lcrypto mostlyclean-compile: -rm -f *.$(OBJEXT) --- ./applications/aaa/Makefile.in.orig 2008-04-05 13:08:20.000000000 -0700 +++ ./applications/aaa/Makefile.in 2008-04-05 13:26:49.000000000 -0700 @@ -315,7 +315,7 @@ src/aaad_diameter_eap.$(OBJEXT): src/$(a src/$(DEPDIR)/$(am__dirstamp) aaad$(EXEEXT): $(aaad_OBJECTS) $(aaad_DEPENDENCIES) @rm -f aaad$(EXEEXT) - $(CXXLINK) $(aaad_LDFLAGS) $(aaad_OBJECTS) $(aaad_LDADD) $(LIBS) + $(CXXLINK) $(aaad_LDFLAGS) $(aaad_OBJECTS) $(aaad_LDADD) $(LIBS) -lcrypto mostlyclean-compile: -rm -f *.$(OBJEXT) --- ./applications/pana/Makefile.in.orig 2008-04-05 13:08:27.000000000 -0700 +++ ./applications/pana/Makefile.in 2008-04-05 13:27:12.000000000 -0700 @@ -273,10 +273,10 @@ clean-noinstPROGRAMS: done client_pana_eaptls$(EXEEXT): $(client_pana_eaptls_OBJECTS) $(client_pana_eaptls_DEPENDENCIES) @rm -f client_pana_eaptls$(EXEEXT) - $(CXXLINK) $(client_pana_eaptls_LDFLAGS) $(client_pana_eaptls_OBJECTS) $(client_pana_eaptls_LDADD) $(LIBS) + $(CXXLINK) $(client_pana_eaptls_LDFLAGS) $(client_pana_eaptls_OBJECTS) $(client_pana_eaptls_LDADD) $(LIBS) -lcrypto pacd$(EXEEXT): $(pacd_OBJECTS) $(pacd_DEPENDENCIES) @rm -f pacd$(EXEEXT) - $(CXXLINK) $(pacd_LDFLAGS) $(pacd_OBJECTS) $(pacd_LDADD) $(LIBS) + $(CXXLINK) $(pacd_LDFLAGS) $(pacd_OBJECTS) $(pacd_LDADD) $(LIBS) -lcrypto mostlyclean-compile: -rm -f *.$(OBJEXT) --- ./libeap/Makefile.in.orig 2008-04-05 13:10:41.000000000 -0700 +++ ./libeap/Makefile.in 2008-04-05 13:25:21.000000000 -0700 @@ -569,12 +569,12 @@ tests/md5_test.$(OBJEXT): tests/$(am__di tests/$(DEPDIR)/$(am__dirstamp) md5_test$(EXEEXT): $(md5_test_OBJECTS) $(md5_test_DEPENDENCIES) @rm -f md5_test$(EXEEXT) - $(CXXLINK) $(md5_test_LDFLAGS) $(md5_test_OBJECTS) $(md5_test_LDADD) $(LIBS) + $(CXXLINK) $(md5_test_LDFLAGS) $(md5_test_OBJECTS) $(md5_test_LDADD) $(LIBS) -lcrypto tests/tls_test.$(OBJEXT): tests/$(am__dirstamp) \ tests/$(DEPDIR)/$(am__dirstamp) tls_test$(EXEEXT): $(tls_test_OBJECTS) $(tls_test_DEPENDENCIES) @rm -f tls_test$(EXEEXT) - $(CXXLINK) $(tls_test_LDFLAGS) $(tls_test_OBJECTS) $(tls_test_LDADD) $(LIBS) + $(CXXLINK) $(tls_test_LDFLAGS) $(tls_test_OBJECTS) $(tls_test_LDADD) $(LIBS) -lcrypto mostlyclean-compile: -rm -f *.$(OBJEXT) --- ./libdiametereap/Makefile.in.orig 2008-04-05 14:05:46.000000000 -0700 +++ ./libdiametereap/Makefile.in 2008-04-05 14:06:11.000000000 -0700 @@ -418,7 +418,7 @@ src/diameter_eap_der_parser.lo: src/$(am src/diameter_eap_dea_parser.lo: src/$(am__dirstamp) \ src/$(DEPDIR)/$(am__dirstamp) libdiametereap.la: $(libdiametereap_la_OBJECTS) $(libdiametereap_la_DEPENDENCIES) - $(CXXLINK) -rpath $(libdir) $(libdiametereap_la_LDFLAGS) $(libdiametereap_la_OBJECTS) $(libdiametereap_la_LIBADD) $(LIBS) + $(CXXLINK) -rpath $(libdir) $(libdiametereap_la_LDFLAGS) $(libdiametereap_la_OBJECTS) $(libdiametereap_la_LIBADD) $(LIBS) -lcrypto clean-noinstPROGRAMS: @list='$(noinst_PROGRAMS)'; for p in $$list; do \ @@ -436,22 +436,22 @@ test/client_test.$(OBJEXT): test/$(am__d test/$(DEPDIR)/$(am__dirstamp) client_test$(EXEEXT): $(client_test_OBJECTS) $(client_test_DEPENDENCIES) @rm -f client_test$(EXEEXT) - $(CXXLINK) $(client_test_LDFLAGS) $(client_test_OBJECTS) $(client_test_LDADD) $(LIBS) + $(CXXLINK) $(client_test_LDFLAGS) $(client_test_OBJECTS) $(client_test_LDADD) $(LIBS) -lcrypto test/client_test_tls.$(OBJEXT): test/$(am__dirstamp) \ test/$(DEPDIR)/$(am__dirstamp) client_test_tls$(EXEEXT): $(client_test_tls_OBJECTS) $(client_test_tls_DEPENDENCIES) @rm -f client_test_tls$(EXEEXT) - $(CXXLINK) $(client_test_tls_LDFLAGS) $(client_test_tls_OBJECTS) $(client_test_tls_LDADD) $(LIBS) + $(CXXLINK) $(client_test_tls_LDFLAGS) $(client_test_tls_OBJECTS) $(client_test_tls_LDADD) $(LIBS) -lcrypto test/server_test.$(OBJEXT): test/$(am__dirstamp) \ test/$(DEPDIR)/$(am__dirstamp) server_test$(EXEEXT): $(server_test_OBJECTS) $(server_test_DEPENDENCIES) @rm -f server_test$(EXEEXT) - $(CXXLINK) $(server_test_LDFLAGS) $(server_test_OBJECTS) $(server_test_LDADD) $(LIBS) + $(CXXLINK) $(server_test_LDFLAGS) $(server_test_OBJECTS) $(server_test_LDADD) $(LIBS) -lcrypto test/server_test_tls.$(OBJEXT): test/$(am__dirstamp) \ test/$(DEPDIR)/$(am__dirstamp) server_test_tls$(EXEEXT): $(server_test_tls_OBJECTS) $(server_test_tls_DEPENDENCIES) @rm -f server_test_tls$(EXEEXT) - $(CXXLINK) $(server_test_tls_LDFLAGS) $(server_test_tls_OBJECTS) $(server_test_tls_LDADD) $(LIBS) + $(CXXLINK) $(server_test_tls_LDFLAGS) $(server_test_tls_OBJECTS) $(server_test_tls_LDADD) $(LIBS) -lcrypto mostlyclean-compile: -rm -f *.$(OBJEXT)
Appendix for the Diff that missed in Step 2 above
aaad_diameter_server.xml for aaad: --- aaad_diameter_server.xml.orig 2008-04-05 15:14:33.000000000 -0700 +++ aaad_diameter_server.xml 2008-04-05 15:15:19.000000000 -0700 @@ -24,7 +24,7 @@ </vendor_specific_application_id> </general> <parser> - <dictionary>config/aaad_diameter_dictionary.xml</dictionary> + <dictionary>/etc/opendiameter/aaa/config/aaad_diameter_dictionary.xml</dictionary> </parser> <transport_mngt> <identity>server.isp.net</identity> nasd_pana_paa.xml and nasd_diameter_eap.xml for nasd: --- nasd_pana_paa.xml.orig 2008-04-05 15:17:08.000000000 -0700 +++ nasd_pana_paa.xml 2008-04-05 15:18:55.000000000 -0700 @@ -12,7 +12,7 @@ <listen_port>3001</listen_port> - <dictionary_filename>config/dictionary.xml</dictionary_filename> + <dictionary_filename>/etc/opendiameter/nas/config/nasd_pana_dictionary.xml</dictionary_filename>
Free Software news
- RT @phoewass: @iptech and thank you for supporting #FreeSoftware :) #JQuarks
- The... http://t.co/xc1hM0H6Gl #biglots #dollarstore #familydollar #freesoftware #freeware #microwavepopcorn #openoffice #savemoney
- @schestowitz Of course #android needs another browser- don't have all our eggs in 1 basket #oneringtorulethemall #FreeSoftware
- RT @ProphecyEnt: Listen 2 Mi (Riot Squad Riddim) @djcollage ❂♢♪ http://t.co/vXhc0cbbn4 ♪♢❂@ProphecyEnt #Dancehall #Promo | #Ad : http://t.c…
- Getting involved in #FreeSoftware | http://t.co/XH9Dleh3p7 via @jospoortvliet
- Writing device drivers in Linux: A brief tutorial
- Now that the ink has dried, how long do you think Microsoft will wait to start suing free software developers and companies?
- Paper is dead - has PDF followed suit?
- Free Software Magazine: articles on free software
- Ubuntu Touch: the (natural) next step in personal computing?
- GRUB tips and tricks
- Managing users in Ubuntu