Alan T. DeKok [Tue, 1 Apr 2014 15:32:51 +0000 (11:32 -0400)]
Document how the data is passed to/from Perl
Alan T. DeKok [Tue, 1 Apr 2014 15:26:37 +0000 (11:26 -0400)]
Move DEBUG messages to RDEBUG
and clean them up
Alan T. DeKok [Tue, 1 Apr 2014 15:19:46 +0000 (11:19 -0400)]
Don't escape strings to / from Perl
Arran Cudbard-Bell [Tue, 1 Apr 2014 15:59:36 +0000 (16:59 +0100)]
Fix formatting in rlm_perl
nchaigne [Tue, 1 Apr 2014 15:42:02 +0000 (16:42 +0100)]
Update to rlm_perl to also handle sub-sections in "config" section
Arran Cudbard-Bell [Tue, 1 Apr 2014 15:00:51 +0000 (16:00 +0100)]
Add missing comment symbols in mysql.conf
Alan T. DeKok [Tue, 1 Apr 2014 12:33:32 +0000 (08:33 -0400)]
\, --> , for consistency with everything else
Alan T. DeKok [Mon, 31 Mar 2014 21:05:42 +0000 (17:05 -0400)]
Pre-parse static RHS of update section.
Because we can now do it, and it will save us time.
Arran Cudbard-Bell [Mon, 31 Mar 2014 21:17:47 +0000 (22:17 +0100)]
Add regression text for 0x prefix
Arran Cudbard-Bell [Mon, 31 Mar 2014 21:17:31 +0000 (22:17 +0100)]
Remember which hex test failed
Arran Cudbard-Bell [Mon, 31 Mar 2014 21:09:45 +0000 (22:09 +0100)]
Doxygen
Arran Cudbard-Bell [Mon, 31 Mar 2014 21:04:52 +0000 (22:04 +0100)]
Fix memory leaked at exit time
Alan T. DeKok [Mon, 31 Mar 2014 20:07:55 +0000 (16:07 -0400)]
Note recent changes
Alan T. DeKok [Mon, 31 Mar 2014 20:04:50 +0000 (16:04 -0400)]
add const to cast
Alan T. DeKok [Mon, 31 Mar 2014 19:58:29 +0000 (15:58 -0400)]
Allow integer64 to be converted to ethernet
Alan T. DeKok [Mon, 31 Mar 2014 19:53:23 +0000 (15:53 -0400)]
Checks for casts to differing sized attributes
Alan T. DeKok [Mon, 31 Mar 2014 19:42:34 +0000 (15:42 -0400)]
Check for unknown attr on the RHS
Alan T. DeKok [Mon, 31 Mar 2014 19:02:56 +0000 (15:02 -0400)]
Remove leading "0x" when expanding octet types
Which is a simple patch now that the evaluation of casts avoids
printing && parsing of attributes
Alan T. DeKok [Mon, 31 Mar 2014 19:00:56 +0000 (15:00 -0400)]
When evaluating a cast, avoid vp_aprint() if possible
This means doing more work on the raw data, but that will avoid
a lot of unnecessary string print and parsing
Alan T. DeKok [Mon, 31 Mar 2014 19:00:34 +0000 (15:00 -0400)]
better debug messages for casts
Alan T. DeKok [Mon, 31 Mar 2014 18:56:39 +0000 (14:56 -0400)]
Remove unnecessary casts in switch / case evaluation
Alan T. DeKok [Mon, 31 Mar 2014 18:44:29 +0000 (14:44 -0400)]
Remember which tests failed
so we can have "fail 1" and "fail 4", instead of "fail" and "fail"
Alan T. DeKok [Mon, 31 Mar 2014 18:29:47 +0000 (14:29 -0400)]
Tests for cast on LHS and attr on RHS
Alan T. DeKok [Mon, 31 Mar 2014 18:27:58 +0000 (14:27 -0400)]
Rewrite unnecessary casts
Alan T. DeKok [Mon, 31 Mar 2014 17:47:49 +0000 (13:47 -0400)]
Certain data types are forbidden in casts
Arran Cudbard-Bell [Mon, 31 Mar 2014 18:15:27 +0000 (19:15 +0100)]
We can actually remove valuepair2str entirely
*pvp could never of been NULL as this was checked for
Arran Cudbard-Bell [Mon, 31 Mar 2014 18:08:06 +0000 (19:08 +0100)]
Should still indicate truncation
Arran Cudbard-Bell [Mon, 31 Mar 2014 18:05:49 +0000 (19:05 +0100)]
Escape tests depend on xlat-attr-number
Arran Cudbard-Bell [Mon, 31 Mar 2014 18:04:54 +0000 (19:04 +0100)]
Change order of hash functions in rlm_expr.c
Arran Cudbard-Bell [Mon, 31 Mar 2014 17:24:07 +0000 (18:24 +0100)]
Don't write a terminating 0 to a zero length buffer!
Arran Cudbard-Bell [Mon, 31 Mar 2014 16:25:51 +0000 (17:25 +0100)]
Foreach should return an escaped copy of the string
Arran Cudbard-Bell [Mon, 31 Mar 2014 15:57:47 +0000 (16:57 +0100)]
Fix vp_aprint to always print escaped versions of strings, this lets the rest of the code deal with embedded NULLs correctly.
Add tests for escaping and unescaping special chars.
Arran Cudbard-Bell [Mon, 31 Mar 2014 15:54:55 +0000 (16:54 +0100)]
Minor formatting
Arran Cudbard-Bell [Mon, 31 Mar 2014 15:54:45 +0000 (16:54 +0100)]
Fix fr_print_string's behaviour to work with buffers of exactly the right length
Else we need to add more hacks in fr_print_string_len if result < 4 pad result etc... and that's just nasty.
Alan T. DeKok [Mon, 31 Mar 2014 15:00:09 +0000 (11:00 -0400)]
Point to correct precursor test
Alan T. DeKok [Mon, 31 Mar 2014 14:55:44 +0000 (10:55 -0400)]
Note recent changes
Alan T. DeKok [Mon, 31 Mar 2014 14:53:20 +0000 (10:53 -0400)]
Check for request->proxy, not request->proxy_reply
Which may be NULL if there was no reply
Alan T. DeKok [Mon, 31 Mar 2014 13:02:21 +0000 (09:02 -0400)]
Enable pre-compiled regex_t
And add regression tests for pre-parsed regex
Alan T. DeKok [Mon, 31 Mar 2014 13:01:42 +0000 (09:01 -0400)]
Add regex_t to value_pair_tmpl_t
And move includes from radiusd.h, so that map.h has regex_t
defined.
Alan T. DeKok [Mon, 31 Mar 2014 12:38:15 +0000 (08:38 -0400)]
Syntax check regular expressions at load time
And re-arrange related code to make the code paths clearer
Alan T. DeKok [Sat, 29 Mar 2014 16:29:02 +0000 (12:29 -0400)]
Update comments, and point to "man users"
Alan T. DeKok [Sat, 29 Mar 2014 16:28:38 +0000 (12:28 -0400)]
Update documentation
Arran Cudbard-Bell [Sat, 29 Mar 2014 19:12:08 +0000 (19:12 +0000)]
Minor resource leak in radclient CID 1194946
Arran Cudbard-Bell [Sat, 29 Mar 2014 18:18:57 +0000 (18:18 +0000)]
Don't segv printing the Pre-Proxy type
Arran Cudbard-Bell [Sat, 29 Mar 2014 17:42:20 +0000 (17:42 +0000)]
Formatting in SHA1
Alan T. DeKok [Sat, 29 Mar 2014 16:08:21 +0000 (12:08 -0400)]
Formatting and naming
to make it consistent with the rest of the server
Alan T. DeKok [Sat, 29 Mar 2014 15:59:09 +0000 (11:59 -0400)]
Config files are in ${modconfdir}
So that we don't put compile-time paths into the modules
Arran Cudbard-Bell [Sat, 29 Mar 2014 10:12:12 +0000 (10:12 +0000)]
Minor reformatting and tweaks
Arran Cudbard-Bell [Sat, 29 Mar 2014 09:49:30 +0000 (09:49 +0000)]
Add libunbound to travis
skids [Thu, 8 Aug 2013 15:17:57 +0000 (11:17 -0400)]
Add rlm_unbound a wrapper around libunbound
skids [Wed, 31 Jul 2013 00:24:48 +0000 (20:24 -0400)]
Expose the main event loop so modules can use it.
Initialize, but do not start, event loop earlier.
Do it through a new API function that leaves room for enhancements
such as a second thread/event loop for modules that can do their
events in a threadsafe manner. But don't actually do that yet.
Arran Cudbard-Bell [Sat, 29 Mar 2014 09:28:11 +0000 (09:28 +0000)]
Update ChangeLog
Arran Cudbard-Bell [Sat, 29 Mar 2014 09:21:48 +0000 (09:21 +0000)]
Fix test/instance number separator
Arran Cudbard-Bell [Fri, 28 Mar 2014 17:44:38 +0000 (17:44 +0000)]
Add basic filter support to radclient
Add backport multiple fixes from master branch
Alan T. DeKok [Fri, 28 Mar 2014 19:18:35 +0000 (15:18 -0400)]
Document changes to switch / case
Alan T. DeKok [Fri, 28 Mar 2014 18:15:11 +0000 (14:15 -0400)]
Doxygen typo
Alan T. DeKok [Fri, 28 Mar 2014 18:10:12 +0000 (14:10 -0400)]
Remove unused argument
Alan T. DeKok [Fri, 28 Mar 2014 18:03:31 +0000 (14:03 -0400)]
Allow a wider range of attribute comparisons in switch / case
via casting the attribute in the 'case' to the one in the 'switch'
Alan T. DeKok [Fri, 28 Mar 2014 18:00:44 +0000 (14:00 -0400)]
Fix typo
Alan T. DeKok [Fri, 28 Mar 2014 17:57:30 +0000 (13:57 -0400)]
switching over a non-existent attribute jumps to the default
So we don't do any expansions, hostname lookups, etc.
Otherwise if we're switching over an IP address attribute, it
might end up doing a DNS lookup for every 'case' statement
Alan T. DeKok [Fri, 28 Mar 2014 17:42:12 +0000 (13:42 -0400)]
Parse values for switch statements
If we're switch over an attribute, check that the values in the
case statements are appropriate for the attribute.
Alan T. DeKok [Fri, 28 Mar 2014 17:33:06 +0000 (13:33 -0400)]
Expose radius_tmpl_cast() for others to use
Arran Cudbard-Bell [Fri, 28 Mar 2014 17:56:03 +0000 (17:56 +0000)]
Fix cosmetic issue in EAP-SIM error messages
Arran Cudbard-Bell [Fri, 28 Mar 2014 17:42:18 +0000 (17:42 +0000)]
Merge pull request #570 from nchaigne/v3.0.x
V3.0.x - rlm_perl patch to manage "config" perl configuration items.
nchaigne [Fri, 28 Mar 2014 08:10:11 +0000 (09:10 +0100)]
Update to rlm_perl to handle "config" section
Arran Cudbard-Bell [Fri, 28 Mar 2014 15:57:31 +0000 (15:57 +0000)]
Make "can't find attribute" errors clearer in EAP-Sim
Arran Cudbard-Bell [Fri, 28 Mar 2014 15:55:51 +0000 (15:55 +0000)]
Add hack for backwards compatibility (look in reply lists if we can't fine the triplets in the control list)
Arran Cudbard-Bell [Tue, 24 Sep 2013 16:28:09 +0000 (17:28 +0100)]
Look for triplets, and ki in control: not reply:
Arran Cudbard-Bell [Tue, 24 Sep 2013 11:49:23 +0000 (12:49 +0100)]
Add support for computing Kc and SRES from Ki and RAND for Comp128-1 (and possibly Comp128-2 and Comp128-3)
Alan T. DeKok [Fri, 28 Mar 2014 15:46:54 +0000 (11:46 -0400)]
Precompile 'switch' and 'case' arguments.
and add regression tests.
Previously "switch blah" with a static string would always hit
the "case {..." block. It will now match a "case blah {" block,
if it exists.
Alan T. DeKok [Fri, 28 Mar 2014 15:18:38 +0000 (11:18 -0400)]
Added cf_section_name2_type
For switch and case statements
Alan T. DeKok [Fri, 28 Mar 2014 13:20:01 +0000 (09:20 -0400)]
set fr_strerror() if the reference isn't an attribute
Alan T. DeKok [Fri, 28 Mar 2014 13:19:38 +0000 (09:19 -0400)]
If there's no dictionary entry, it's a literal.
And use the same literal code as the other cases
Alan T. DeKok [Fri, 28 Mar 2014 13:05:45 +0000 (09:05 -0400)]
use FR_TOKEN, not int. Better error message for parse errors
case "foo {
is a syntax error. It shouldn't print a complaint about an
invalid token.
Alan T. DeKok [Fri, 28 Mar 2014 12:56:49 +0000 (08:56 -0400)]
Dictionary names begin with uppercase letters
Arran Cudbard-Bell [Fri, 28 Mar 2014 15:20:47 +0000 (15:20 +0000)]
Remove extraneous format marker *sigh*
Arran Cudbard-Bell [Fri, 28 Mar 2014 14:25:46 +0000 (14:25 +0000)]
Extraneous comma
Arran Cudbard-Bell [Fri, 28 Mar 2014 12:59:01 +0000 (12:59 +0000)]
Include version in user agent string
Print the headers were adding
Fix whitespace
Arran Cudbard-Bell [Fri, 28 Mar 2014 12:24:19 +0000 (12:24 +0000)]
Update ChangeLog
Arran Cudbard-Bell [Fri, 28 Mar 2014 10:52:26 +0000 (10:52 +0000)]
Fix whitespace
Arran Cudbard-Bell [Fri, 28 Mar 2014 11:21:57 +0000 (11:21 +0000)]
connect_uri can now be NULL
Since the default was removed
Arran Cudbard-Bell [Thu, 27 Mar 2014 19:14:08 +0000 (19:14 +0000)]
Swap defined(sun) for defined(__sun) which is apparently defined in more cases
Arran Cudbard-Bell [Thu, 27 Mar 2014 17:57:10 +0000 (17:57 +0000)]
Fixup all lingering instances of old style expansion
Arran Cudbard-Bell [Thu, 27 Mar 2014 15:21:06 +0000 (15:21 +0000)]
Update ChangeLog
Arran Cudbard-Bell [Thu, 27 Mar 2014 15:10:16 +0000 (15:10 +0000)]
Display truncated format error marker, instead of no error marker at all
Arran Cudbard-Bell [Thu, 27 Mar 2014 10:09:36 +0000 (10:09 +0000)]
Correct length returned by vp_prints_value_json
Arran Cudbard-Bell [Thu, 27 Mar 2014 09:37:33 +0000 (09:37 +0000)]
Revert "CID 1187991"
Until we know more about the possible side effects
This reverts commit
0d0e29676da0e451845ca59b3a0dab14b0cdb493.
Alan Buxey [Thu, 27 Mar 2014 09:17:00 +0000 (09:17 +0000)]
Update variables.rst
Arran Cudbard-Bell [Thu, 27 Mar 2014 08:55:23 +0000 (08:55 +0000)]
Use newer style alternation in oracle queries
Arran Cudbard-Bell [Thu, 27 Mar 2014 08:48:46 +0000 (08:48 +0000)]
Update queries.conf
Arran Cudbard-Bell [Thu, 27 Mar 2014 07:47:33 +0000 (07:47 +0000)]
Merge pull request #564 from alanbuxey/dictionary
updated TERENA dictionary
Alan Buxey [Thu, 27 Mar 2014 07:38:03 +0000 (07:38 +0000)]
updated TERENA dictionary
update of TERENA dictionary
Alan Buxey [Wed, 26 Mar 2014 20:19:07 +0000 (20:19 +0000)]
expose server version to xlat
expose server version to slat (so we can print it out in e.g. line log
with %v).
I note that a few expansions are present but are in variables.rst -
e.g. %C, %V and %Z
Arran Cudbard-Bell [Wed, 26 Mar 2014 16:52:27 +0000 (16:52 +0000)]
Change asserts to rad_asserts
Arran Cudbard-Bell [Wed, 26 Mar 2014 16:39:15 +0000 (16:39 +0000)]
Add custom body encoder
This allows admin to specify an arbitrary content-type and body data
Arran Cudbard-Bell [Wed, 26 Mar 2014 16:33:45 +0000 (16:33 +0000)]
Always add headers
Arran Cudbard-Bell [Wed, 26 Mar 2014 16:29:22 +0000 (16:29 +0000)]
Simplify freeing
Arran Cudbard-Bell [Wed, 26 Mar 2014 16:28:35 +0000 (16:28 +0000)]
Check 'supported' array and number of types are the same
Arran Cudbard-Bell [Wed, 26 Mar 2014 16:09:12 +0000 (16:09 +0000)]
Whitespace fixes in rlm_expr
Alan T. DeKok [Wed, 26 Mar 2014 15:15:41 +0000 (11:15 -0400)]
Make 'i' and 'len' the same type as 'digestlen'
Because they're being compared to it. Fixes #561