]> git.ipfire.org Git - thirdparty/freeradius-server.git/log
thirdparty/freeradius-server.git
5 weeks agonote that people should use github
Alan T. DeKok [Thu, 19 Mar 2026 17:18:19 +0000 (13:18 -0400)] 
note that people should use github

5 weeks agoformat and note github access tokens
Alan T. DeKok [Thu, 19 Mar 2026 17:12:05 +0000 (13:12 -0400)] 
format and note github access tokens

5 weeks agoAdd test of association_time to redis_ippool tests
Nick Porter [Thu, 19 Mar 2026 17:54:35 +0000 (17:54 +0000)] 
Add test of association_time to redis_ippool tests

5 weeks agoRenumber lua script lines
Nick Porter [Thu, 19 Mar 2026 17:53:53 +0000 (17:53 +0000)] 
Renumber lua script lines

5 weeks agoExpiry status is not relevant to sticky IP allocation
Nick Porter [Thu, 19 Mar 2026 17:51:14 +0000 (17:51 +0000)] 
Expiry status is not relevant to sticky IP allocation

If the device -> IP association persists, and the IP has not been
allocated to another device then re-allocate it to the requesting
device.

5 weeks agoAdd association_time option to rlm_redis_ippool
Nick Porter [Thu, 19 Mar 2026 16:45:02 +0000 (16:45 +0000)] 
Add association_time option to rlm_redis_ippool

To allow for "sticky" addressing.

This allows the owner -> IP association to last longer than the address
lease, so that allocation requests can return the address previously
used by a device if it has not been allocated to another owner since the
lease expiry.

5 weeks agoAdd fr_coord_pair_int_request_start()
Nick Porter [Thu, 19 Mar 2026 12:07:08 +0000 (12:07 +0000)] 
Add fr_coord_pair_int_request_start()

To start a request within a coordinator rather than in response to
worker -> coordinator data.

5 weeks agoBreak apart coord_pair_request_bootstrap
Nick Porter [Thu, 19 Mar 2026 11:51:14 +0000 (11:51 +0000)] 
Break apart coord_pair_request_bootstrap

To allow for requests being run through a coordinator process module to
come from sources other than worker -> coordinator data.

5 weeks agoAdd fr_coord_pair_request_coord_pair
Nick Porter [Wed, 18 Mar 2026 15:26:10 +0000 (15:26 +0000)] 
Add fr_coord_pair_request_coord_pair

To retrieve the coord_pair associated with a given request
being run through a process module.

5 weeks agoAdd fr_coord_worker_reply_broadcast
Nick Porter [Wed, 18 Mar 2026 15:20:30 +0000 (15:20 +0000)] 
Add fr_coord_worker_reply_broadcast

So replies from coordinator process modules can be sent to all workers.

5 weeks agoCorrect naming
Nick Porter [Wed, 18 Mar 2026 15:19:14 +0000 (15:19 +0000)] 
Correct naming

5 weeks agoset variables to NULL on free
Alan T. DeKok [Thu, 19 Mar 2026 15:16:18 +0000 (11:16 -0400)] 
set variables to NULL on free

and goto correct cleanup path

5 weeks agore-add request=child, as it's used to set pairs in the child
Alan T. DeKok [Thu, 19 Mar 2026 15:15:04 +0000 (11:15 -0400)] 
re-add request=child, as it's used to set pairs in the child

5 weeks agoalways add tls client hello version
Alan T. DeKok [Thu, 19 Mar 2026 15:03:09 +0000 (11:03 -0400)] 
always add tls client hello version

5 weeks agofree extensions in more error cases
Alan T. DeKok [Thu, 19 Mar 2026 13:45:16 +0000 (09:45 -0400)] 
free extensions in more error cases

5 weeks agoonly set CRL if we're using certificates
Alan T. DeKok [Thu, 19 Mar 2026 09:22:09 +0000 (05:22 -0400)] 
only set CRL if we're using certificates

and complain if check_crl=yes, but OpenSSL doesn't have the
relevant flags

5 weeks agoset ASYNC for non-certificate case, too
Alan T. DeKok [Thu, 19 Mar 2026 08:47:20 +0000 (04:47 -0400)] 
set ASYNC for non-certificate case, too

5 weeks agoadd server libraries to json fuzzer
Alan T. DeKok [Thu, 19 Mar 2026 07:00:41 +0000 (03:00 -0400)] 
add server libraries to json fuzzer

5 weeks agoadd FR_TRUNK_LIST_FUNC(draining, trunk_connection_t)
Catalina Pineros [Wed, 18 Mar 2026 18:15:46 +0000 (14:15 -0400)] 
add FR_TRUNK_LIST_FUNC(draining, trunk_connection_t)

5 weeks agoadd FR_TRUNK_LIST_FUNC(inactive_draining, trunk_connection_t)
Catalina Pineros [Wed, 18 Mar 2026 17:44:00 +0000 (13:44 -0400)] 
add FR_TRUNK_LIST_FUNC(inactive_draining, trunk_connection_t)

5 weeks agoadd FR_TRUNK_LIST_FUNC(inactive, trunk_connection_t)
Catalina Pineros [Wed, 18 Mar 2026 15:22:52 +0000 (11:22 -0400)] 
add FR_TRUNK_LIST_FUNC(inactive, trunk_connection_t)

5 weeks agouse FR_TRUNK_LIST_FUNC for 'full' list
Catalina Pineros [Thu, 19 Mar 2026 06:53:34 +0000 (02:53 -0400)] 
use FR_TRUNK_LIST_FUNC for 'full' list

5 weeks agoupdate FR_TRUNC_LIST macro to specify a target data type
Catalina Pineros [Thu, 19 Mar 2026 06:37:09 +0000 (02:37 -0400)] 
update FR_TRUNC_LIST macro to specify a target data type

5 weeks agofix typos in license
Alan T. DeKok [Thu, 19 Mar 2026 06:17:55 +0000 (02:17 -0400)] 
fix typos in license

5 weeks agofix typos
Catalina Pineros [Wed, 18 Mar 2026 20:16:54 +0000 (16:16 -0400)] 
fix typos

5 weeks agofix typos
Catalina Pineros [Tue, 17 Mar 2026 15:47:14 +0000 (11:47 -0400)] 
fix typos

5 weeks agofix typo in comment
Catalina Pineros [Tue, 17 Mar 2026 15:35:48 +0000 (11:35 -0400)] 
fix typo in comment
(rlm_redis_ippool_tool: fix wrong doc comment on ippool_tool_stats_t.free 354)

5 weeks agomulti-server: use fixed subnet and remove env-setup.sh
Arran Cudbard-Bell [Thu, 19 Mar 2026 02:57:33 +0000 (20:57 -0600)] 
multi-server: use fixed subnet and remove env-setup.sh

Set TEST_SUBNET via compose environment with a fixed 172.28.0.0/16
subnet (overridable via test_subnet param). Remove env-setup.sh
which ran apt-get install on every container startup.

5 weeks agoPass the scheduler as a context for thread instantiate
Arran Cudbard-Bell [Thu, 19 Mar 2026 02:04:15 +0000 (20:04 -0600)] 
Pass the scheduler as a context for thread instantiate

5 weeks agoCreate thread instance data for the main thread too (when not running in single threa...
Arran Cudbard-Bell [Wed, 18 Mar 2026 23:32:15 +0000 (17:32 -0600)] 
Create thread instance data for the main thread too (when not running in single threaded mode)

This is needed for triggers to be able to call xlats

5 weeks agomulti-server: refactor test framework for parameterised tests
Arran Cudbard-Bell [Wed, 18 Mar 2026 17:22:33 +0000 (11:22 -0600)] 
multi-server: refactor test framework for parameterised tests

Restructure the multi-server test framework so that each test is
defined by a single YAML params file.  All templates (compose,
radiusd configs, test steps) are rendered from shared Jinja2
templates using those params.

- Replace monolithic all.mk with macro-based framework that
  auto-discovers test suites and param files
- Move configs to configs/freeradius/, environments to environments/
- Rename test suites to proxy-accept and proxy-multihop-accept
- Use *.test.yml for param files, *.ci.test.yml for CI-only tests
- Add test.multi-server.ci target for short CI tests
- Render all build products into build/ directory
- Support parallel execution with unique Docker project names
- Add healthchecks to compose templates
- Remove hardcoded port mappings (containers use internal networking)
- Add CI workflow for GitHub Actions with Docker-in-Docker

5 weeks agodocker: build libkqueue from source on non-amd64 architectures
Arran Cudbard-Bell [Wed, 18 Mar 2026 02:14:45 +0000 (20:14 -0600)] 
docker: build libkqueue from source on non-amd64 architectures

The NetworkRADIUS extras repository only provides amd64 packages
for libkqueue. On other architectures (e.g. arm64), build libkqueue
from source and install the resulting .deb packages in both the
build and runtime stages.

5 weeks agodebian: remove --without-rlm_sql_unixodbc from rules
Arran Cudbard-Bell [Wed, 18 Mar 2026 01:44:23 +0000 (19:44 -0600)] 
debian: remove --without-rlm_sql_unixodbc from rules

The unixodbc module was explicitly disabled in configure but the
freeradius-unixodbc package was still defined in control.in with
its .install file expecting the .so to exist, causing dh_install
to fail with missing files.

5 weeks agoclean up variable on exit
Alan T. DeKok [Wed, 18 Mar 2026 10:51:52 +0000 (06:51 -0400)] 
clean up variable on exit

5 weeks agouse goto inside "if" instead of "break"
Alan T. DeKok [Wed, 18 Mar 2026 10:50:58 +0000 (06:50 -0400)] 
use goto inside "if" instead of "break"

5 weeks agoCorrect comments
Nick Porter [Wed, 18 Mar 2026 11:29:37 +0000 (11:29 +0000)] 
Correct comments

5 weeks agoRemove incorrect frees
Nick Porter [Mon, 16 Mar 2026 12:55:35 +0000 (12:55 +0000)] 
Remove incorrect frees

5 weeks agoprint relevant variable in error message
Catalina Pineros [Tue, 17 Mar 2026 14:25:17 +0000 (10:25 -0400)] 
print relevant variable in error message

5 weeks agofix typos in comments
Catalina Pineros [Mon, 16 Mar 2026 18:19:18 +0000 (14:19 -0400)] 
fix typos in comments

5 weeks agocatch corner case of allocation failure
Alan T. DeKok [Wed, 18 Mar 2026 09:55:18 +0000 (05:55 -0400)] 
catch corner case of allocation failure

5 weeks agomove bnctx to thread instance data
Alan T. DeKok [Wed, 18 Mar 2026 09:51:46 +0000 (05:51 -0400)] 
move bnctx to thread instance data

5 weeks agoadd DIAG_OFF to quiet C compiler warnings
Alan T. DeKok [Wed, 18 Mar 2026 09:51:20 +0000 (05:51 -0400)] 
add DIAG_OFF to quiet C compiler warnings

6 weeks agofix typo
Alan T. DeKok [Tue, 17 Mar 2026 20:51:58 +0000 (16:51 -0400)] 
fix typo

6 weeks agoformat to follow coding style / standards
Alan T. DeKok [Tue, 17 Mar 2026 12:56:51 +0000 (08:56 -0400)] 
format to follow coding style / standards

6 weeks agoOSS-Fuzz: Add fuzzer for json
Arthur Chan [Thu, 12 Mar 2026 22:10:03 +0000 (22:10 +0000)] 
OSS-Fuzz: Add fuzzer for json

Signed-off-by: Arthur Chan <arthur.chan@adalogics.com>
6 weeks agovarious corner cases in PEAP
Alan T. DeKok [Tue, 17 Mar 2026 05:45:40 +0000 (01:45 -0400)] 
various corner cases in PEAP

when TLS API calls fail, return unlang fail.

6 weeks agocatch corner cases in TTLS.
Alan T. DeKok [Tue, 17 Mar 2026 04:14:11 +0000 (00:14 -0400)] 
catch corner cases in TTLS.

6 weeks agocatch various extreme corner cases
Alan T. DeKok [Tue, 17 Mar 2026 03:53:34 +0000 (23:53 -0400)] 
catch various extreme corner cases

6 weeks agofree memory on error, and reference data which is used multiple times
Alan T. DeKok [Tue, 17 Mar 2026 03:32:38 +0000 (23:32 -0400)] 
free memory on error, and reference data which is used multiple times

6 weeks agohandle the case of zero-length fields
Alan T. DeKok [Tue, 17 Mar 2026 03:18:41 +0000 (23:18 -0400)] 
handle the case of zero-length fields

memcmp() can't handle NULL pointers.

6 weeks agojust use CMP() macro
Alan T. DeKok [Tue, 17 Mar 2026 03:14:40 +0000 (23:14 -0400)] 
just use CMP() macro

6 weeks agoremove assert: fr_value_box_cmp() is now a stable comparison
Alan T. DeKok [Tue, 17 Mar 2026 03:00:50 +0000 (23:00 -0400)] 
remove assert: fr_value_box_cmp() is now a stable comparison

6 weeks agodon't print anything for "make clean"
Alan T. DeKok [Tue, 17 Mar 2026 02:55:30 +0000 (22:55 -0400)] 
don't print anything for "make clean"

6 weeks agofixed overflow
Catalina Pineros [Mon, 16 Mar 2026 20:03:02 +0000 (16:03 -0400)] 
fixed overflow

if there is KEY_SLOTS, the valid range is [0, KEY_SLOTS - 1], and key KEY_SLOTS should be outside of redis lot range

6 weeks agofix typos
Catalina Pineros [Mon, 16 Mar 2026 18:32:33 +0000 (14:32 -0400)] 
fix typos

6 weeks agofix same typo in comment, in different files
Catalina Pineros [Mon, 16 Mar 2026 16:19:27 +0000 (12:19 -0400)] 
fix same typo in comment, in different files

6 weeks agofix typo in comment
Catalina Pineros [Mon, 16 Mar 2026 16:14:54 +0000 (12:14 -0400)] 
fix typo in comment

6 weeks agofix typo in comment
Catalina Pineros [Mon, 16 Mar 2026 16:03:07 +0000 (12:03 -0400)] 
fix typo in comment

6 weeks agofix typo in comment
Catalina Pineros [Mon, 16 Mar 2026 15:58:19 +0000 (11:58 -0400)] 
fix typo in comment

6 weeks agofix misplaced parenthesis
Catalina Pineros [Mon, 16 Mar 2026 14:55:13 +0000 (10:55 -0400)] 
fix misplaced parenthesis

tolower(cmd_str[1] != s ) does not actually turn anything into lowercase, since it is evaluating a number and not a char

6 weeks agofr_value_box_cmp() is used in stable sorts
Alan T. DeKok [Mon, 16 Mar 2026 22:18:39 +0000 (18:18 -0400)] 
fr_value_box_cmp() is used in stable sorts

and therefore can't call fr_digest_cmp().

6 weeks agouse MEMCMP_FIELDS() in paircmp() too
Alan T. DeKok [Mon, 16 Mar 2026 22:18:07 +0000 (18:18 -0400)] 
use MEMCMP_FIELDS() in paircmp() too

and add note about the secret flag

6 weeks agodefine MEMCMP_FIELDS() and use it
Alan T. DeKok [Mon, 16 Mar 2026 21:47:10 +0000 (17:47 -0400)] 
define MEMCMP_FIELDS() and use it

which is a little clearer (and less wrappers) than MEMCMP_RETURN

6 weeks agojust call cache_expire() instead of doing it locally
Alan T. DeKok [Mon, 16 Mar 2026 20:54:40 +0000 (16:54 -0400)] 
just call cache_expire() instead of doing it locally

which handles more corner cases.

also other minor tweaks

6 weeks agotry to clear out 3 cache entries for every 1 lookup
Alan T. DeKok [Mon, 16 Mar 2026 20:51:56 +0000 (16:51 -0400)] 
try to clear out 3 cache entries for every 1 lookup

which lets us clean up old entries more quickly after a spike of
traffic

6 weeks agorestore write errno after ftruncate in partial header write path
ethan-thompson [Mon, 16 Mar 2026 16:21:51 +0000 (12:21 -0400)] 
restore write errno after ftruncate in partial header write path

Signed-off-by: ethan-thompson <ethan.thompson@networkradius.com>
6 weeks agofix int truncation of ssize_t return from fr_concatv
ethan-thompson [Mon, 16 Mar 2026 15:43:22 +0000 (11:43 -0400)] 
fix int truncation of ssize_t return from fr_concatv

Signed-off-by: ethan-thompson <ethan.thompson@networkradius.com>
6 weeks agoskip NULL request pointers when marking batch entries runnable
ethan-thompson [Mon, 16 Mar 2026 15:36:05 +0000 (11:36 -0400)] 
skip NULL request pointers when marking batch entries runnable

Signed-off-by: ethan-thompson <ethan.thompson@networkradius.com>
6 weeks agouse fr_hash_string for file table lookup to match filename_hash
ethan-thompson [Mon, 16 Mar 2026 15:27:47 +0000 (11:27 -0400)] 
use fr_hash_string for file table lookup to match filename_hash

Signed-off-by: ethan-thompson <ethan.thompson@networkradius.com>
6 weeks agoalways return "updated" if we updated something
Alan T. DeKok [Mon, 16 Mar 2026 18:18:40 +0000 (14:18 -0400)] 
always return "updated" if we updated something

6 weeks agocache xlat can succeed if it returns no data.
Alan T. DeKok [Mon, 16 Mar 2026 18:16:39 +0000 (14:16 -0400)] 
cache xlat can succeed if it returns no data.

i.e. "no data in cache" isn't an xlat failure, it's an xlat
success with no data.

6 weeks agoset_ttl with TTL=0 is a noop
Alan T. DeKok [Mon, 16 Mar 2026 18:09:36 +0000 (14:09 -0400)] 
set_ttl with TTL=0 is a noop

6 weeks agofix cache it logic
Alan T. DeKok [Mon, 16 Mar 2026 14:44:30 +0000 (10:44 -0400)] 
fix cache it logic

so that it follows the documentation

6 weeks agopass correct pointer to handle
Alan T. DeKok [Mon, 16 Mar 2026 14:44:09 +0000 (10:44 -0400)] 
pass correct pointer to handle

6 weeks agoreturn 0 on count error, not negative.
Alan T. DeKok [Mon, 16 Mar 2026 14:42:33 +0000 (10:42 -0400)] 
return 0 on count error, not negative.

because the return value is unsigned.

6 weeks agoallow tests without the full re-run of all tests
Alan T. DeKok [Mon, 16 Mar 2026 15:27:51 +0000 (11:27 -0400)] 
allow tests without the full re-run of all tests

6 weeks agocache_foo depends on rlm_cache, too
Alan T. DeKok [Mon, 16 Mar 2026 15:20:08 +0000 (11:20 -0400)] 
cache_foo depends on rlm_cache, too

6 weeks agoCorrect section name
Nick Porter [Mon, 16 Mar 2026 08:32:20 +0000 (08:32 +0000)] 
Correct section name

6 weeks agominor cleanups in thread code
Alan T. DeKok [Sun, 15 Mar 2026 22:57:08 +0000 (18:57 -0400)] 
minor cleanups in thread code

6 weeks agomove fd management to narrower scope
Alan T. DeKok [Sun, 15 Mar 2026 22:47:58 +0000 (18:47 -0400)] 
move fd management to narrower scope

and use fileno(fp), so that we have fewer concerns with managing
the fd

6 weeks agomove common code to fr_thread_start / fr_thread_exit
Alan T. DeKok [Sun, 15 Mar 2026 19:02:28 +0000 (15:02 -0400)] 
move common code to fr_thread_start / fr_thread_exit

6 weeks agorename fr_thread functions for clarity
Alan T. DeKok [Sun, 15 Mar 2026 18:32:08 +0000 (14:32 -0400)] 
rename fr_thread functions for clarity

6 weeks agomove thread information to a common data structure
Alan T. DeKok [Sun, 15 Mar 2026 17:54:58 +0000 (13:54 -0400)] 
move thread information to a common data structure

6 weeks agomove thread create && wait to common functions
Alan T. DeKok [Sun, 15 Mar 2026 15:30:18 +0000 (11:30 -0400)] 
move thread create && wait to common functions

6 weeks agomove thread setup to common function
Alan T. DeKok [Sun, 15 Mar 2026 15:16:50 +0000 (11:16 -0400)] 
move thread setup to common function

6 weeks agomove thread instantiation to common function
Alan T. DeKok [Sun, 15 Mar 2026 15:13:03 +0000 (11:13 -0400)] 
move thread instantiation to common function

6 weeks agosimplify the names
Alan T. DeKok [Thu, 12 Mar 2026 17:28:55 +0000 (13:28 -0400)] 
simplify the names

less shouting is good.

6 weeks agoInitial baseline multi-server tests for load-generator -> 5 homeserver and load-gener...
Marc-Andre Casavant [Mon, 23 Feb 2026 19:00:48 +0000 (14:00 -0500)] 
Initial baseline multi-server tests for load-generator -> 5 homeserver and load-generator -> 1 proxy -> 2 homeserver scenarios.

6 weeks agotweaks, and use fr_dlist_foreach() in more places
Alan T. DeKok [Sun, 15 Mar 2026 13:15:00 +0000 (09:15 -0400)] 
tweaks, and use fr_dlist_foreach() in more places

6 weeks agowrap multi-line macro in do/while
Alan T. DeKok [Sun, 15 Mar 2026 13:01:54 +0000 (09:01 -0400)] 
wrap multi-line macro in do/while

6 weeks agofree dbuff on encode failure
Alan T. DeKok [Sun, 15 Mar 2026 12:59:12 +0000 (08:59 -0400)] 
free dbuff on encode failure

6 weeks agominimize includes
Alan T. DeKok [Sun, 15 Mar 2026 12:58:28 +0000 (08:58 -0400)] 
minimize includes

6 weeks agoPacify Coverity (CID #1686327)
Nick Porter [Sun, 15 Mar 2026 12:31:10 +0000 (12:31 +0000)] 
Pacify Coverity (CID #1686327)

6 weeks agoCheck return (CID #1686328)
Nick Porter [Sun, 15 Mar 2026 12:28:42 +0000 (12:28 +0000)] 
Check return (CID #1686328)

6 weeks agoCheck return (CID #1686329)
Nick Porter [Sun, 15 Mar 2026 12:28:16 +0000 (12:28 +0000)] 
Check return (CID #1686329)

6 weeks agoda can't be NULL here
Alan T. DeKok [Sun, 15 Mar 2026 02:42:17 +0000 (22:42 -0400)] 
da can't be NULL here

6 weeks agoreturn failure on error
Alan T. DeKok [Sun, 15 Mar 2026 02:40:56 +0000 (22:40 -0400)] 
return failure on error

6 weeks agocheck the return code of adding the compile list
Alan T. DeKok [Sun, 15 Mar 2026 02:40:16 +0000 (22:40 -0400)] 
check the return code of adding the compile list

6 weeks agotypo
Alan T. DeKok [Sun, 15 Mar 2026 02:39:45 +0000 (22:39 -0400)] 
typo

6 weeks agounify code which warns about unused processing sections
Alan T. DeKok [Sun, 15 Mar 2026 02:38:30 +0000 (22:38 -0400)] 
unify code which warns about unused processing sections