]>
git.ipfire.org Git - thirdparty/freeradius-server.git/log
Arran Cudbard-Bell [Sun, 1 Nov 2020 00:03:31 +0000 (19:03 -0500)]
s/vp_map_t/map_t/g
Arran Cudbard-Bell [Sat, 31 Oct 2020 23:57:58 +0000 (18:57 -0500)]
Simplify debugging
Arran Cudbard-Bell [Sat, 31 Oct 2020 18:10:48 +0000 (13:10 -0500)]
Weaken the loop tests
Arran Cudbard-Bell [Sat, 31 Oct 2020 17:39:40 +0000 (12:39 -0500)]
Only push the section if we have something to execute
Arran Cudbard-Bell [Sat, 31 Oct 2020 17:24:47 +0000 (12:24 -0500)]
Allow call to run with no children
Arran Cudbard-Bell [Sat, 31 Oct 2020 17:24:18 +0000 (12:24 -0500)]
Formatting
Arran Cudbard-Bell [Sat, 31 Oct 2020 16:41:21 +0000 (11:41 -0500)]
Formatting
Arran Cudbard-Bell [Sat, 31 Oct 2020 16:19:31 +0000 (11:19 -0500)]
subrequest instruction needs to be talloced
James Jones [Sat, 31 Oct 2020 13:26:11 +0000 (08:26 -0500)]
Correct test_dbuff_no_advance() (#3718)
fr_dbuff_end() just returns the end pointer; actually moving to
the end requires fr_dbuff_set_to_end().
Arran Cudbard-Bell [Sat, 31 Oct 2020 04:32:18 +0000 (23:32 -0500)]
unlang: Use unlang keyword specialisation structs instead of separate allocations
James Jones [Fri, 30 Oct 2020 21:23:52 +0000 (16:23 -0500)]
Test that operations on a dbuff created with FR_DBUFF_NO_ADVANCE() don't affect its parent's position. (#3714)
Jorge Pereira [Fri, 30 Oct 2020 21:23:20 +0000 (18:23 -0300)]
rlm_python: Add comments for -frecord-gcc-switches #3693 (#3702)
Alan T. DeKok [Fri, 30 Oct 2020 15:09:28 +0000 (11:09 -0400)]
add descriptive error messages
Alan T. DeKok [Fri, 30 Oct 2020 12:52:32 +0000 (08:52 -0400)]
remove filter from value_box
Alan T. DeKok [Fri, 30 Oct 2020 12:49:14 +0000 (08:49 -0400)]
remove 'abinary' data type.
Document the changes. Update the RADIUS encode / decode tests.
Remove FR_TYPE_ABINARY from everything.
Alan T. DeKok [Fri, 30 Oct 2020 12:35:18 +0000 (08:35 -0400)]
remove / fixup abinary tests
in preparation for moving "abinary" data type to "string",
and making it radius-specific
Alan T. DeKok [Fri, 30 Oct 2020 12:30:54 +0000 (08:30 -0400)]
update comments to match functionality
left over from changed behavior in commit
323f9c88a3
Alan T. DeKok [Fri, 30 Oct 2020 11:04:27 +0000 (07:04 -0400)]
remove debug stuff :(
Arran Cudbard-Bell [Fri, 30 Oct 2020 02:26:34 +0000 (21:26 -0500)]
Fix issues with errors when a length is passed to pton functions
Alan T. DeKok [Thu, 29 Oct 2020 18:44:27 +0000 (14:44 -0400)]
set the async flag as appropriate
Alan T. DeKok [Thu, 29 Oct 2020 18:29:06 +0000 (14:29 -0400)]
fix for async. left over from changed behavior in commit
323f9c88a3
Alan T. DeKok [Thu, 29 Oct 2020 18:04:02 +0000 (14:04 -0400)]
print the correct quotes around FR_TYPE_TMPL, too
Alan T. DeKok [Thu, 29 Oct 2020 18:00:05 +0000 (14:00 -0400)]
no need to use public functions for internal values
we're already using internal values everywhere else in the parser
Alan T. DeKok [Thu, 29 Oct 2020 17:54:01 +0000 (13:54 -0400)]
convert "header" to using tmpl
which now lets the header be taken from an attribute, if the
user so desires
Alan T. DeKok [Thu, 29 Oct 2020 17:51:34 +0000 (13:51 -0400)]
add FR_TYPE_NON_BLOCKING
for use with tmpl_eval(). We can mark configuration items as
allowing tmpls / xlats, BUT also mark them up with a requirement
for non-blocking operation.
This lets us convert much of the code to using tmpl_eval() instead
of xlat_eval(). At the same time, the admin is prevented from
breaking the server by using asychronous operations in places
which do not support them.
Once all of the callers of foo_eval() have been converted to
async, we can remove the FR_TYPE_NON_BLOCKING flag.
Alan T. DeKok [Thu, 29 Oct 2020 17:50:21 +0000 (13:50 -0400)]
correct documentation and behavior
to match the name of the variable, and how all of the code
uses the variable
Alan T. DeKok [Thu, 29 Oct 2020 14:55:31 +0000 (10:55 -0400)]
now that abinary is a pointer, treat it as such
Alan T. DeKok [Thu, 29 Oct 2020 14:53:10 +0000 (10:53 -0400)]
comment out abinary tests until they're fixed
Alan T. DeKok [Thu, 29 Oct 2020 14:44:07 +0000 (10:44 -0400)]
print out string values with correct quotes and escaping
so that the output is as close as possible to the input
Alan T. DeKok [Thu, 29 Oct 2020 14:27:38 +0000 (10:27 -0400)]
no need to quote the string
Alan T. DeKok [Thu, 29 Oct 2020 14:16:50 +0000 (10:16 -0400)]
use cp->attr internally, instead of cf_pair_attr(cp)
and clean up debug printing for FR_TYPE_STRING
Alan T. DeKok [Thu, 29 Oct 2020 13:52:56 +0000 (09:52 -0400)]
we cannot call fr_value_box_from_str() on any structural types
Alan T. DeKok [Thu, 29 Oct 2020 12:56:35 +0000 (08:56 -0400)]
We cannot assign hex strings to attributes of type "vsa"
Fix the test so that it has the correct error message
Alan T. DeKok [Wed, 28 Oct 2020 14:13:32 +0000 (10:13 -0400)]
remove unused macros
Arran Cudbard-Bell [Wed, 28 Oct 2020 15:28:58 +0000 (09:28 -0600)]
Fix pair type checks to allow all nested types not just FR_TYPE_GROUP
Arran Cudbard-Bell [Wed, 28 Oct 2020 14:23:42 +0000 (08:23 -0600)]
sbuff: Set const depending on start pointer
Alan T. DeKok [Wed, 28 Oct 2020 14:01:13 +0000 (10:01 -0400)]
pass the value_box directly to print_abinary
Alan T. DeKok [Wed, 28 Oct 2020 13:45:21 +0000 (09:45 -0400)]
move print_abinary() to use sbuff internally
Alan T. DeKok [Wed, 28 Oct 2020 12:22:56 +0000 (08:22 -0400)]
cleanup ascend filters
make it a bit more forgiving with respect to input size, by
requiring only as much data as we need. Instead of mandating it
is 32 octets.
Remove unused definitions. Convert the union to anonymous.
Arran Cudbard-Bell [Tue, 27 Oct 2020 20:10:45 +0000 (14:10 -0600)]
Fix talloc ctx for group attributes
Arran Cudbard-Bell [Tue, 27 Oct 2020 20:07:39 +0000 (14:07 -0600)]
Add nested subrequest tests
James Jones [Tue, 27 Oct 2020 19:07:37 +0000 (14:07 -0500)]
Add functions and macros to read values and copy bytes from dbuffs (#3709)
Alan T. DeKok [Tue, 27 Oct 2020 16:55:36 +0000 (12:55 -0400)]
parse Juniper filters with IPv6 addresses
Alan T. DeKok [Tue, 27 Oct 2020 15:46:19 +0000 (11:46 -0400)]
move ascend filters to pointer, and fix use of vb_length
vb_length _was_ inside of the union for variable-sized types,
but the code also set it for fixed-size types. Only the size
of the ascend filter prevented various over-writes.
changing ascend filter to pointer instead of an array highlighted
this issue, and causes us to move the "length" field out of the
union in order to stop the over-write
Alan T. DeKok [Tue, 27 Oct 2020 15:33:06 +0000 (11:33 -0400)]
abstract datum.length to vb_length
Alan T. DeKok [Tue, 27 Oct 2020 15:30:32 +0000 (11:30 -0400)]
abstract datum.length
Alan T. DeKok [Tue, 27 Oct 2020 15:26:36 +0000 (11:26 -0400)]
use vb_length instead of mangling things manually
Alan T. DeKok [Tue, 27 Oct 2020 15:11:28 +0000 (11:11 -0400)]
add more tests for ascend data filters
James Jones [Mon, 26 Oct 2020 21:54:39 +0000 (16:54 -0500)]
Correct the fr_net_to_int*() functions
Alan T. DeKok [Tue, 27 Oct 2020 14:44:21 +0000 (10:44 -0400)]
use fr_cursor_tail() instead of vp->next
Alan T. DeKok [Tue, 27 Oct 2020 13:30:00 +0000 (09:30 -0400)]
move VMPS to using cursors, instead of VALUE_PAIR**
Alan T. DeKok [Sun, 25 Oct 2020 17:07:06 +0000 (13:07 -0400)]
more xlat tests
Arran Cudbard-Bell [Sun, 25 Oct 2020 15:13:54 +0000 (09:13 -0600)]
Fix vendor specific tests
Alan T. DeKok [Sun, 25 Oct 2020 13:36:42 +0000 (09:36 -0400)]
add tests for \%
which aren't correct right now, but the tests pass...
Alan T. DeKok [Sun, 25 Oct 2020 13:36:11 +0000 (09:36 -0400)]
set server_cs, too
Arran Cudbard-Bell [Sat, 24 Oct 2020 17:29:49 +0000 (12:29 -0500)]
Fix programatically generated subrequests
Arran Cudbard-Bell [Sat, 24 Oct 2020 17:07:06 +0000 (12:07 -0500)]
Add missing vendor attribute
Alan T. DeKok [Sat, 24 Oct 2020 13:32:06 +0000 (09:32 -0400)]
update to match new behavior
Arran Cudbard-Bell [Sat, 24 Oct 2020 03:15:45 +0000 (22:15 -0500)]
s/REQUEST */request_t */g
REMOVE ANOTHER LOUD TYPE
Arran Cudbard-Bell [Fri, 23 Oct 2020 21:59:24 +0000 (16:59 -0500)]
s/VALUE_PAIR/fr_pair_t/g
FIX ONE OF THE LAST ALL CAPS TYPES IN THE SERVER
Jorge Pereira [Fri, 23 Oct 2020 20:57:11 +0000 (17:57 -0300)]
Shut up clang scan (#3705)
Arran Cudbard-Bell [Fri, 23 Oct 2020 20:38:40 +0000 (15:38 -0500)]
Quiet clang scan
Arran Cudbard-Bell [Fri, 23 Oct 2020 20:33:48 +0000 (15:33 -0500)]
Add subrequest src/dst evaluation code
This allows us to produce nested attributes in unlang
Arran Cudbard-Bell [Fri, 23 Oct 2020 20:32:06 +0000 (15:32 -0500)]
Don't assume the first ref is unparsed
Arran Cudbard-Bell [Fri, 23 Oct 2020 20:16:21 +0000 (15:16 -0500)]
Revert "Write bytes as hex"
This reverts commit
4b3623661c930200a95cc9c2acbf757763863408 .
Arran Cudbard-Bell [Fri, 23 Oct 2020 20:13:06 +0000 (15:13 -0500)]
Fix filtering functions to work with nested attributes
Alan T. DeKok [Fri, 23 Oct 2020 20:11:57 +0000 (16:11 -0400)]
\n is EOF, too
Arran Cudbard-Bell [Fri, 23 Oct 2020 19:15:39 +0000 (14:15 -0500)]
Write bytes as hex
Arran Cudbard-Bell [Fri, 23 Oct 2020 18:40:28 +0000 (13:40 -0500)]
Fix typo in process_subrequest
Alan T. DeKok [Fri, 23 Oct 2020 16:28:05 +0000 (12:28 -0400)]
allow unlang in subrequests
Alan T. DeKok [Fri, 23 Oct 2020 16:26:29 +0000 (12:26 -0400)]
whoops, let's initialize everything
Alan T. DeKok [Fri, 23 Oct 2020 16:19:01 +0000 (12:19 -0400)]
reword process_subrequest() to take many arguments
Arran Cudbard-Bell [Fri, 23 Oct 2020 15:29:32 +0000 (10:29 -0500)]
Start of process subrequests
Alan T. DeKok [Fri, 23 Oct 2020 15:13:21 +0000 (11:13 -0400)]
shut up clang scan. Fixes #3704
Alan T. DeKok [Fri, 23 Oct 2020 14:11:32 +0000 (10:11 -0400)]
pass original packet, not a crappy offset from a vector
Arran Cudbard-Bell [Fri, 23 Oct 2020 14:04:07 +0000 (09:04 -0500)]
Default to fr_value_box_to_network
Arran Cudbard-Bell [Fri, 23 Oct 2020 13:56:44 +0000 (08:56 -0500)]
We don't need to list all the valid types for RADIUS either,
It likely makes the code slower as well... switches aren't magic, it's up to the compiler to implement them and it may well choose a binary search for this number of elements.
Alan T. DeKok [Fri, 23 Oct 2020 13:59:12 +0000 (09:59 -0400)]
don't mangle things until AFTER we've freed the children
Alan T. DeKok [Fri, 23 Oct 2020 13:43:12 +0000 (09:43 -0400)]
recurse into all children for structural types
Alan T. DeKok [Fri, 23 Oct 2020 13:24:04 +0000 (09:24 -0400)]
there's no reason to specify the full path of the files
Alan T. DeKok [Fri, 23 Oct 2020 13:18:42 +0000 (09:18 -0400)]
hoist tag code earlier, so we don't need to talloc_steal()
Arran Cudbard-Bell [Fri, 23 Oct 2020 13:41:18 +0000 (08:41 -0500)]
fr_value_box_to_network already has an "encodable" check
There's no reason to restrict protocols to a subset of types when there are vendor specific attributes involved...
Arran Cudbard-Bell [Fri, 23 Oct 2020 13:24:47 +0000 (08:24 -0500)]
Simpler function
Alan T. DeKok [Fri, 23 Oct 2020 12:40:54 +0000 (08:40 -0400)]
make packet_ctx.vector[16] instead of a pointer. Fixes #3703
Matthew Newton [Fri, 23 Oct 2020 11:47:47 +0000 (12:47 +0100)]
Make sure we never merge flags by mistake
Arran Cudbard-Bell [Fri, 23 Oct 2020 03:22:14 +0000 (22:22 -0500)]
Pass all standard types to fr_value_box_from_network
Arran Cudbard-Bell [Fri, 23 Oct 2020 03:15:19 +0000 (22:15 -0500)]
Don't init the boxes twice
Arran Cudbard-Bell [Fri, 23 Oct 2020 03:11:15 +0000 (22:11 -0500)]
fixup option/len macros in DHCPv6
Arran Cudbard-Bell [Fri, 23 Oct 2020 02:55:50 +0000 (21:55 -0500)]
Revert "minor tweaks"
This reverts commit
44cb7e539f7e52199a30318be980fc2752d2f8f4 .
Arran Cudbard-Bell [Fri, 23 Oct 2020 02:39:50 +0000 (21:39 -0500)]
pair: Pairs should free their children when they're freed
Arran Cudbard-Bell [Fri, 23 Oct 2020 02:38:58 +0000 (21:38 -0500)]
Whitespace
Arran Cudbard-Bell [Fri, 23 Oct 2020 02:38:44 +0000 (21:38 -0500)]
Mark certain attribute references as 'resolve_only'
This means when we're attempting to build out structures we know that if we didn't manage to resolve the ar using the existing pair tree, we need to abort.
Arran Cudbard-Bell [Fri, 23 Oct 2020 02:34:30 +0000 (21:34 -0500)]
Remove white space...
Arran Cudbard-Bell [Fri, 23 Oct 2020 02:34:12 +0000 (21:34 -0500)]
Typo
Arran Cudbard-Bell [Fri, 23 Oct 2020 02:32:40 +0000 (21:32 -0500)]
Unused
Matthew Newton [Thu, 22 Oct 2020 16:32:28 +0000 (17:32 +0100)]
comment why the option is removed
Alan T. DeKok [Thu, 22 Oct 2020 14:37:02 +0000 (10:37 -0400)]
use "text" instead of "varchar".
there's no reason to fix these fields at a small size
Matthew Newton [Thu, 22 Oct 2020 09:27:43 +0000 (10:27 +0100)]
rlm_python: Anchor with a space so we don't mangle the middle of flags
See #3693
Stephen Blackwell [Thu, 22 Oct 2020 00:18:09 +0000 (20:18 -0400)]
rlm_smtp with unit test (#3547)
* rlm_smtp with unit test
* Fix typo, remove extraneous info.
Co-authored-by: Arran Cudbard-Bell <a.cudbardb@freeradius.org>
James Jones [Wed, 7 Oct 2020 22:12:26 +0000 (17:12 -0500)]
Make the underlying dbuff version of fr_radius_encode_pair() visible
This makes it possible to switch users of fr_radius_encode_pair()
over to dbuffs individually rather than all at once.