aland [Wed, 10 Aug 2005 01:11:02 +0000 (01:11 +0000)]
More attributes, as found on the net
aland [Tue, 9 Aug 2005 23:00:11 +0000 (23:00 +0000)]
More utilities to hack dictionaries
aland [Tue, 9 Aug 2005 22:57:18 +0000 (22:57 +0000)]
Even more
aland [Tue, 9 Aug 2005 22:44:49 +0000 (22:44 +0000)]
More attributes, as found on the net
aland [Tue, 9 Aug 2005 21:31:32 +0000 (21:31 +0000)]
New dictionaries, as found on the net
aland [Tue, 9 Aug 2005 21:29:58 +0000 (21:29 +0000)]
In CVS, but not used until the server can handle 16/16 bit VSA's.
nbk [Tue, 9 Aug 2005 11:20:06 +0000 (11:20 +0000)]
Fix compilation warnings.
nbk [Mon, 8 Aug 2005 23:06:30 +0000 (23:06 +0000)]
Delete post_proxy_authorize. The modules dealing with proxy
replies should be put in the 'post-proxy' section only.
aland [Mon, 8 Aug 2005 22:30:34 +0000 (22:30 +0000)]
change passwd_rad_attr to PW_USER_PASSWORD
aland [Mon, 8 Aug 2005 22:27:43 +0000 (22:27 +0000)]
More reformatting
aland [Mon, 8 Aug 2005 22:27:31 +0000 (22:27 +0000)]
More fixes
aland [Mon, 8 Aug 2005 22:23:34 +0000 (22:23 +0000)]
Run the dictionaries through the new formatting program,
so they look all nice and pretty
aland [Mon, 8 Aug 2005 22:22:52 +0000 (22:22 +0000)]
Be a little nicer
aland [Mon, 8 Aug 2005 20:22:14 +0000 (20:22 +0000)]
Script to format dictionaries in a consistent manner, to avoid
hand-formatting issues
aland [Sat, 6 Aug 2005 00:36:03 +0000 (00:36 +0000)]
Reformatted & added attributes
aland [Fri, 5 Aug 2005 17:18:36 +0000 (17:18 +0000)]
From ethereal
aland [Fri, 5 Aug 2005 17:09:10 +0000 (17:09 +0000)]
From ethereal
aland [Fri, 5 Aug 2005 00:59:21 +0000 (00:59 +0000)]
More attributes, as found on the net
aland [Thu, 4 Aug 2005 23:49:51 +0000 (23:49 +0000)]
More attributes, as found on the net
aland [Thu, 4 Aug 2005 23:37:55 +0000 (23:37 +0000)]
Added more attributes, as found on the net
aland [Thu, 4 Aug 2005 23:35:14 +0000 (23:35 +0000)]
Include new Lucent dictionary for Taos16
aland [Thu, 4 Aug 2005 23:34:58 +0000 (23:34 +0000)]
Added 4-octet VSA's, but commented out, because they don't
really make sense...
aland [Thu, 4 Aug 2005 23:19:56 +0000 (23:19 +0000)]
More attributes, as found on the net
aland [Thu, 4 Aug 2005 23:12:14 +0000 (23:12 +0000)]
More attributes
aland [Thu, 4 Aug 2005 18:19:32 +0000 (18:19 +0000)]
Lucent 16-bit VSA's. No VALUE's yet.
Note that we call them "Lucent-*", rather than "Ascend-*", to
avoid stomping on the old Ascend attributes. People who read
the Lucent docs & try to use the same names will be confused,
sorry.
aland [Thu, 4 Aug 2005 17:47:54 +0000 (17:47 +0000)]
More attributes, as found on the net
aland [Thu, 4 Aug 2005 17:40:48 +0000 (17:40 +0000)]
More attributes, as found on the net
aland [Thu, 4 Aug 2005 17:00:44 +0000 (17:00 +0000)]
Corrected typo
aland [Wed, 3 Aug 2005 17:19:39 +0000 (17:19 +0000)]
Don't set auth-type. It's almost always wrong
phampson [Wed, 3 Aug 2005 15:16:48 +0000 (15:16 +0000)]
dpatch-2 patches
phampson [Wed, 3 Aug 2005 15:16:05 +0000 (15:16 +0000)]
Forward port all the interesting things from FreeRADIUS 1.0.x in the Debian
archive to CVS HEAD.
This brings dpatch-2 support, even though there aren't any dpatches needed
yet.
(Changelog entry for 1.1.0~pre0~cvs has become long and unwieldly, may need
to refactor based on the 1.0.x debian-release changelogs.)
aland [Wed, 3 Aug 2005 00:10:51 +0000 (00:10 +0000)]
Don't have a 2-byte over-run read
aland [Tue, 2 Aug 2005 23:11:23 +0000 (23:11 +0000)]
Latest set of changes
aland [Tue, 2 Aug 2005 23:05:13 +0000 (23:05 +0000)]
Parse Lucent 16-bit VSA's in rad_decode, too
aland [Tue, 2 Aug 2005 18:55:54 +0000 (18:55 +0000)]
Create Lucent-style VSA's, too
aland [Tue, 2 Aug 2005 18:51:12 +0000 (18:51 +0000)]
More sanity checking on the attribute numbers, so we don't put
too large attributes into packets
aland [Mon, 1 Aug 2005 17:11:09 +0000 (17:11 +0000)]
Removed usercollide functionality. It can be done with other
features of the server (i.e. rlm_policy)
nbk [Sun, 31 Jul 2005 12:44:18 +0000 (12:44 +0000)]
When linking, it's $(LDFLAGS) instead of $(CFLAGS).
(closes: #208)
Problem reported by Simon Dickhoven <simon@tachyon.net>
nbk [Sat, 30 Jul 2005 21:39:55 +0000 (21:39 +0000)]
Don't zap the session if the function rad_check_ts() failed.
nbk [Sat, 30 Jul 2005 21:38:55 +0000 (21:38 +0000)]
Don't zap the session if the function rad_check_ts() failed.
(closes: #235)
Based on a patch from Stephan Jaeger <stephan.jaeger@ewetel.de>
nbk [Sat, 30 Jul 2005 21:34:31 +0000 (21:34 +0000)]
Add a comment about the return values of the function rad_check_ts()
nbk [Thu, 28 Jul 2005 19:47:32 +0000 (19:47 +0000)]
Check the value of realm->striprealm before overwriting the
User-Name with the Stripped-User-Name.
aland [Tue, 26 Jul 2005 22:27:23 +0000 (22:27 +0000)]
Ports are unsigned int's
aland [Mon, 25 Jul 2005 20:41:06 +0000 (20:41 +0000)]
Use new realm disable & stuff
aland [Mon, 25 Jul 2005 18:55:44 +0000 (18:55 +0000)]
Make realm_disable() work for all kinds of packets.
nbk [Sun, 24 Jul 2005 19:15:17 +0000 (19:15 +0000)]
Additional Cisco VoIP attributes, as posted to the bugzilla
(closes: #249)
Patch from Ilia Chipitsine <ilia@rediska.ru>
nbk [Sat, 23 Jul 2005 15:41:10 +0000 (15:41 +0000)]
Replace '%{NAS-Port}' with %{NAS-Port:-NULL} because it was
reported on the bugzilla that some NASes (for example AS5350
IOS 12.3.T) don't send the attribute in Accounting-Request
packets. (closes: #223)
Bug found by Ilia Chipitsine <ilia@rediska.ru>
aland [Fri, 22 Jul 2005 23:52:00 +0000 (23:52 +0000)]
Allow Attribute-Name[stuff] to work
aland [Fri, 22 Jul 2005 23:40:44 +0000 (23:40 +0000)]
Check return code for rad_send()
aland [Fri, 22 Jul 2005 23:31:02 +0000 (23:31 +0000)]
Fix test for port over-ride
nbk [Fri, 22 Jul 2005 11:01:55 +0000 (11:01 +0000)]
As iODBC defines success codes other than SQL_SUCCESS, we now
use their macro SQL_SUCCEEDED(). (closes: #138)
Bug found by Jeff Brown <jeff.brn@gmail.com>
aland [Thu, 21 Jul 2005 22:44:39 +0000 (22:44 +0000)]
prefer shortname, and set it if unset
aland [Thu, 21 Jul 2005 22:34:28 +0000 (22:34 +0000)]
Allow multiple "listen" sections to use the same clients section,
by putting the RADCLIENT_LIST ptr into the clients section, via
cf_data_add()
nbk [Thu, 21 Jul 2005 13:10:42 +0000 (13:10 +0000)]
s/mainconfig.client_trees/mainconfig.clients/
nbk [Thu, 21 Jul 2005 08:07:48 +0000 (08:07 +0000)]
Install radsqlrelay, too. (the new Perl version)
aland [Thu, 21 Jul 2005 03:18:16 +0000 (03:18 +0000)]
Document "ipv6addr" in listen{}
Document "clients" in listen{}
aland [Thu, 21 Jul 2005 03:07:28 +0000 (03:07 +0000)]
Added support for per-socket client lists.
Right now, you can only have a global list, OR a per-socket list,
but multiple sockets can use the same list.
aland [Tue, 19 Jul 2005 23:05:55 +0000 (23:05 +0000)]
No more radrelay. Say so.
aland [Tue, 19 Jul 2005 23:05:46 +0000 (23:05 +0000)]
Updates
aland [Tue, 19 Jul 2005 20:57:09 +0000 (20:57 +0000)]
Removed radrelay & radsqlrelay
aland [Tue, 19 Jul 2005 20:37:24 +0000 (20:37 +0000)]
Lower the work we do to find a client.
aland [Tue, 19 Jul 2005 20:15:30 +0000 (20:15 +0000)]
Enable SNMP support for clients again.
aland [Mon, 18 Jul 2005 20:26:38 +0000 (20:26 +0000)]
Call client_free() if client_add() fails.
Update SQL code to call new routines. Untested, though...
aland [Mon, 18 Jul 2005 20:22:20 +0000 (20:22 +0000)]
Move RADCLIENT* in mainconfig to rbtree's. This means that
lookups will be faster for large numbers of clients.
Also, the "client_*()" functions now take a pointer to the
"master client list" struct, which means it should be easier
to allow per-"listen" socket clients.
Caveats are
- the SNMP code does NOT return the list of the clients
This may be easy to fix, though.
- due to allowing network/mask, we always walk through a list
of 128 prefixes, even if only 1-2 are defined. This can be
fixed easily, though.
The SQL code has been updated, but not tested
aland [Wed, 13 Jul 2005 23:34:32 +0000 (23:34 +0000)]
Add note re: RFC compliance
aland [Wed, 13 Jul 2005 23:24:15 +0000 (23:24 +0000)]
Let's document EAP, too
nbk [Wed, 13 Jul 2005 16:51:37 +0000 (16:51 +0000)]
Missing rad_free()
aland [Mon, 11 Jul 2005 20:22:20 +0000 (20:22 +0000)]
Make name lookups case sensitive.
Patch from bug #194
aland [Mon, 11 Jul 2005 17:29:12 +0000 (17:29 +0000)]
Make '-p' work again, and catch some corner cases
aland [Mon, 11 Jul 2005 17:20:20 +0000 (17:20 +0000)]
Initialize modules AFTER forking, which may help some situations.
On SEGV, call _exit(), rather than exit()
aland [Mon, 11 Jul 2005 17:18:37 +0000 (17:18 +0000)]
Be less verbose
aland [Mon, 11 Jul 2005 17:05:46 +0000 (17:05 +0000)]
Don't copy VSA's to an Access-Reject packet
This closes bug #207
nbk [Mon, 11 Jul 2005 13:48:43 +0000 (13:48 +0000)]
It's 'realm->acct_ipaddr' instead of 'realm->ipaddr'.
nbk [Sun, 10 Jul 2005 16:53:00 +0000 (16:53 +0000)]
Allow to cancel proxy of accounting with Proxy-To-Realm := LOCAL
aland [Fri, 8 Jul 2005 18:17:31 +0000 (18:17 +0000)]
New function lrad_hash_fold(). Not used, but it should be...
aland [Fri, 8 Jul 2005 18:16:46 +0000 (18:16 +0000)]
Pull "verify packet" into separate routine, to mirror rad_sign()
aland [Thu, 7 Jul 2005 22:01:33 +0000 (22:01 +0000)]
A bit better checks for packet length.
We could use FIONREAD to get the number of bytes, but that's already
in the RADIUS header. And if the RADIUS header lies to use, we'll
discard the packet anyways...
nbk [Thu, 7 Jul 2005 16:05:39 +0000 (16:05 +0000)]
Remove length restrictions on strings in generate_sql_clients()
function because RADCLIENT data structure was changed to use
malloc'd strings. (see src/include/radiusd.h 1.175)
Bug found by Velikanov <alex@asu.farlep.net>
aland [Thu, 7 Jul 2005 03:05:28 +0000 (03:05 +0000)]
Use MSG_PEEK to get the first 4 bytes of the packet, and malloc()
memory around that. The previous method of allocating a 4k buffer
on the stack, then calling malloc(), and copying the data was
less than efficient.
aland [Thu, 7 Jul 2005 01:46:00 +0000 (01:46 +0000)]
clarify rad_tunnel_pwdecode
aland [Wed, 6 Jul 2005 22:28:06 +0000 (22:28 +0000)]
Re-write rad_pwencode() and rad_pwdecode() to be clearer, so that
they don't call memcpy as much.
aland [Wed, 6 Jul 2005 21:04:00 +0000 (21:04 +0000)]
Added 'text' to top of the files, so editors will know what
to do with them
aland [Wed, 6 Jul 2005 17:12:40 +0000 (17:12 +0000)]
Use 'void *' rather than 'unsigned char *'
aland [Tue, 5 Jul 2005 23:43:27 +0000 (23:43 +0000)]
Use certificate chain. Patch from bug #112, and closes it.
aland [Tue, 5 Jul 2005 20:44:57 +0000 (20:44 +0000)]
Hack to allow it to build
pnixon [Fri, 1 Jul 2005 15:03:37 +0000 (15:03 +0000)]
Also store h323voicequality in SQL. Thanks to Ilia Chipitsine
pnixon [Fri, 1 Jul 2005 15:03:03 +0000 (15:03 +0000)]
Change default servername
nbk [Fri, 1 Jul 2005 08:49:04 +0000 (08:49 +0000)]
Fix the IPv4 addresses comparison in client_find()
aland [Wed, 29 Jun 2005 20:40:09 +0000 (20:40 +0000)]
NetBSD compatibility
aland [Wed, 29 Jun 2005 18:46:17 +0000 (18:46 +0000)]
Added one more (important) call to reap_children
aland [Tue, 28 Jun 2005 00:54:38 +0000 (00:54 +0000)]
Updates, as posted to the list
aland [Tue, 28 Jun 2005 00:44:29 +0000 (00:44 +0000)]
Aruba vendor dictionary, as posted to the list
fcusack [Mon, 27 Jun 2005 11:10:14 +0000 (11:10 +0000)]
soft PIN
nbk [Sat, 25 Jun 2005 22:44:35 +0000 (22:44 +0000)]
Fix quote removal of the value in sql_userparse (closes: #242)
aland [Tue, 21 Jun 2005 16:41:14 +0000 (16:41 +0000)]
Removed errant ^G
aland [Mon, 20 Jun 2005 19:24:56 +0000 (19:24 +0000)]
Properly initialize radlog_dest
nbk [Sun, 19 Jun 2005 17:05:14 +0000 (17:05 +0000)]
Man page for radsqlrelay (the new Perl version)
aland [Sat, 18 Jun 2005 01:24:11 +0000 (01:24 +0000)]
Make parser more forgiving.
Patch from Wolfgang Hottgenroth, to close bug #228
aland [Thu, 16 Jun 2005 17:13:53 +0000 (17:13 +0000)]
Suppress sending rejects *unless* DELAYED_REJECT is set.
Clear DELAYED_REJECT *after* sending the delayed packet
aland [Wed, 15 Jun 2005 23:02:20 +0000 (23:02 +0000)]
Be a little more forgiving about where the time may be