]>
git.ipfire.org Git - thirdparty/freeradius-server.git/log
Alan T. DeKok [Mon, 24 Apr 2017 19:54:30 +0000 (15:54 -0400)]
fix build issues
Alan T. DeKok [Mon, 24 Apr 2017 19:39:25 +0000 (15:39 -0400)]
use PRIu64
Alan T. DeKok [Mon, 24 Apr 2017 19:17:33 +0000 (15:17 -0400)]
add more debugging, and set count=1 when we've found one
Arran Cudbard-Bell [Mon, 24 Apr 2017 18:55:21 +0000 (14:55 -0400)]
Clarify what "server" is in common_socket_print
Arran Cudbard-Bell [Mon, 24 Apr 2017 18:54:58 +0000 (14:54 -0400)]
Fixup how memory is freed in rbtrees
Alan T. DeKok [Mon, 24 Apr 2017 15:02:09 +0000 (11:02 -0400)]
renamed receiver to network.
It's not perfect, but it's clearer and makes more sense
Arran Cudbard-Bell [Sat, 22 Apr 2017 19:33:35 +0000 (15:33 -0400)]
Fix binary unsafe access of request->password, and compiler signed complains in rlm_pap
Arran Cudbard-Bell [Sat, 22 Apr 2017 19:27:52 +0000 (15:27 -0400)]
Avoid infinite loop of fr_strerror_pops
Arran Cudbard-Bell [Sat, 22 Apr 2017 18:57:47 +0000 (14:57 -0400)]
Mark more BFD structures as packet
Arran Cudbard-Bell [Sat, 22 Apr 2017 14:23:09 +0000 (10:23 -0400)]
Fix build errors
Arran Cudbard-Bell [Thu, 20 Apr 2017 22:16:22 +0000 (18:16 -0400)]
Split fr_ldap_reuslt and use new perror functions to print errors
Arran Cudbard-Bell [Thu, 20 Apr 2017 22:15:51 +0000 (18:15 -0400)]
Fixup const order in talloc.h
Arran Cudbard-Bell [Thu, 20 Apr 2017 22:15:33 +0000 (18:15 -0400)]
Add perror functions to drain fr_strerror buffers
Arran Cudbard-Bell [Thu, 20 Apr 2017 22:14:37 +0000 (18:14 -0400)]
Fix issue where PATCH value is bad
Alan T. DeKok [Thu, 20 Apr 2017 14:58:41 +0000 (10:58 -0400)]
remember lowest_active_callers, too
Alan T. DeKok [Thu, 20 Apr 2017 14:56:09 +0000 (10:56 -0400)]
initialize lowest number to ~0, so we pick a real lowest number
Alan T. DeKok [Thu, 20 Apr 2017 01:20:27 +0000 (21:20 -0400)]
we've renamed the file
Alan T. DeKok [Wed, 19 Apr 2017 19:40:47 +0000 (15:40 -0400)]
remove restrictions on parallel
the contents can now be anything :(
Alan T. DeKok [Wed, 19 Apr 2017 19:35:27 +0000 (15:35 -0400)]
notes on updated load-balance sections
Alan T. DeKok [Wed, 19 Apr 2017 19:31:53 +0000 (15:31 -0400)]
markdown fixes
Alan T. DeKok [Wed, 19 Apr 2017 19:29:18 +0000 (15:29 -0400)]
renamed and moved to markdown
which is what we're standardizing on
Alan T. DeKok [Wed, 19 Apr 2017 19:08:36 +0000 (15:08 -0400)]
recursively collect active callers.
So load-balancing works for policies, too
Alan T. DeKok [Wed, 19 Apr 2017 17:02:51 +0000 (13:02 -0400)]
remove unused variable
Alan T. DeKok [Wed, 19 Apr 2017 16:47:53 +0000 (12:47 -0400)]
notes on selection of redundant node after failure / resumption
Alan T. DeKok [Wed, 19 Apr 2017 16:44:21 +0000 (12:44 -0400)]
use module thread-specific data to do thread-specific load-balancing
By counting the number of yielded modules, and choosing the one
with the lowest number of outstanding requests.
Alan T. DeKok [Wed, 19 Apr 2017 16:31:03 +0000 (12:31 -0400)]
track total / active calls to modules
Alan T. DeKok [Wed, 19 Apr 2017 15:52:51 +0000 (11:52 -0400)]
make code in resume the same as in module_call
The priority wasn't being set correctly, and the code was just
different.
Alan T. DeKok [Wed, 19 Apr 2017 15:49:21 +0000 (11:49 -0400)]
instantiate thread-specific data in single-threaded mode, too
Alan T. DeKok [Wed, 19 Apr 2017 15:25:59 +0000 (11:25 -0400)]
return module_thread_instance_t, so we can access counters
Alan T. DeKok [Wed, 19 Apr 2017 15:22:57 +0000 (11:22 -0400)]
always create thread-local module data
Alan T. DeKok [Wed, 19 Apr 2017 14:56:21 +0000 (10:56 -0400)]
minor cleanups
Alan T. DeKok [Wed, 19 Apr 2017 14:32:10 +0000 (10:32 -0400)]
remove redundant install rule
Alan T. DeKok [Wed, 19 Apr 2017 14:31:16 +0000 (10:31 -0400)]
remove last vestiges of pass2
Alan T. DeKok [Wed, 19 Apr 2017 14:16:01 +0000 (10:16 -0400)]
separate compilation of redundant and load-balance
the functions are just renamed for now. Later commits will
change the functionality.
Alan T. DeKok [Wed, 19 Apr 2017 13:49:03 +0000 (09:49 -0400)]
remove redundant declarations
Alan T. DeKok [Wed, 19 Apr 2017 13:18:56 +0000 (09:18 -0400)]
finalize fix for previous commit
Arran Cudbard-Bell [Wed, 19 Apr 2017 01:29:27 +0000 (21:29 -0400)]
Fix overzealous search and replace at some point…
Alan T. DeKok [Tue, 18 Apr 2017 21:05:34 +0000 (17:05 -0400)]
include stdint.h too
Alan T. DeKok [Tue, 18 Apr 2017 21:03:28 +0000 (17:03 -0400)]
doxygen
Alan T. DeKok [Tue, 18 Apr 2017 20:32:26 +0000 (16:32 -0400)]
typo
Alan T. DeKok [Tue, 18 Apr 2017 20:30:55 +0000 (16:30 -0400)]
Move the configuration into place, too
Alan T. DeKok [Tue, 18 Apr 2017 20:27:48 +0000 (16:27 -0400)]
move rlm_dynamic_clients to rlm_client
Alan T. DeKok [Tue, 18 Apr 2017 20:25:43 +0000 (16:25 -0400)]
this is no longer necessary
The code is in the dynamic_clients module
Alan T. DeKok [Tue, 18 Apr 2017 20:25:03 +0000 (16:25 -0400)]
copy client xlat and map from rlm_client to here
Arran Cudbard-Bell [Tue, 18 Apr 2017 18:29:24 +0000 (14:29 -0400)]
Move talloc functions to their own src file
Arran Cudbard-Bell [Thu, 13 Apr 2017 17:27:35 +0000 (13:27 -0400)]
Hack to get our extended log fmt specified working for DEBUG() and friends
Arran Cudbard-Bell [Tue, 18 Apr 2017 00:44:53 +0000 (20:44 -0400)]
…and a few more
Arran Cudbard-Bell [Tue, 18 Apr 2017 00:08:09 +0000 (20:08 -0400)]
Hmm might not need to wrap talloc_get_type_abort with TALLOC_GET_TYPE_ABORT_NOOP
Arran Cudbard-Bell [Mon, 17 Apr 2017 23:36:09 +0000 (19:36 -0400)]
Retrieve all entries by default
Arran Cudbard-Bell [Mon, 17 Apr 2017 22:50:24 +0000 (18:50 -0400)]
Remove #ifndef NDEBUG talloc_get_type_abort (where appropriate)
We use TALLOC_GET_TYPE_ABORT_NOOP now, which means in many cases we don’t need to explicitly ifdef out talloc_get_type_abort.
Where we have (void) talloc_get_type_abort() this should be ifdef’d out with #ifndef TALLOC_GET_TYPE_ABORT_NOOP so we don’t get self assignment errors from static analysis tools.
If anyone feels like writing a new talloc_get_type_abort_void() (or whatever macro, and replacing all the #ifndef TALLOC_GET_TYPE_ABORT_NOOP talloc_get_type_abort() lines, feel free.
Alan T. DeKok [Mon, 17 Apr 2017 19:03:35 +0000 (15:03 -0400)]
let the dl code clean up after itself.
We don't need to do it here.
Alan T. DeKok [Wed, 5 Apr 2017 21:19:17 +0000 (17:19 -0400)]
set names for workers
Alan T. DeKok [Wed, 5 Apr 2017 21:02:50 +0000 (17:02 -0400)]
colourise default log output, so we don't get the prefixes
Arran Cudbard-Bell [Mon, 17 Apr 2017 22:03:26 +0000 (18:03 -0400)]
Set TALLOC_GET_TYPE_ABORT_NOOP on non-debug builds
Discuss where this should be…
Arran Cudbard-Bell [Thu, 13 Apr 2017 16:13:17 +0000 (12:13 -0400)]
Limit the message types we call ldap_result for
Arran Cudbard-Bell [Thu, 13 Apr 2017 13:54:22 +0000 (09:54 -0400)]
Spell dereference correctly
Arran Cudbard-Bell [Thu, 13 Apr 2017 13:52:09 +0000 (09:52 -0400)]
Pass "all" to ldap_result unmolested (in case there are special values)
Arran Cudbard-Bell [Thu, 13 Apr 2017 13:51:44 +0000 (09:51 -0400)]
Too many flags to have a static type for FR_CONF_SUBSECTION_MULTI
Arran Cudbard-Bell [Thu, 13 Apr 2017 13:51:22 +0000 (09:51 -0400)]
Fix PW_TYPE_REQUIRED for subsections
Arran Cudbard-Bell [Wed, 12 Apr 2017 22:47:32 +0000 (18:47 -0400)]
Typo
Arran Cudbard-Bell [Wed, 12 Apr 2017 22:42:22 +0000 (18:42 -0400)]
Fix PW_TYPE_SUBSECTION | PW_TYPE_MULTI
The conf parser can now produce an array of allocated sturcts representing the subsections.
Arran Cudbard-Bell [Wed, 12 Apr 2017 17:39:38 +0000 (13:39 -0400)]
s/section_sub/subsection/
Arran Cudbard-Bell [Wed, 12 Apr 2017 17:02:00 +0000 (13:02 -0400)]
Fix typo
Arran Cudbard-Bell [Wed, 12 Apr 2017 17:01:53 +0000 (13:01 -0400)]
Add fr_prefix to more ldap library tables
Add async search function (it's not actually async, it just means ldap_result can be called later).
Arran Cudbard-Bell [Mon, 10 Apr 2017 20:54:23 +0000 (16:54 -0400)]
Reorganise some libraries under src/lib
Arran Cudbard-Bell [Fri, 7 Apr 2017 19:44:51 +0000 (15:44 -0400)]
build util.c
Arran Cudbard-Bell [Fri, 7 Apr 2017 15:52:35 +0000 (11:52 -0400)]
Typo
Arran Cudbard-Bell [Fri, 7 Apr 2017 15:13:28 +0000 (11:13 -0400)]
Typo
Arran Cudbard-Bell [Fri, 7 Apr 2017 13:58:51 +0000 (09:58 -0400)]
Pass through conn currently in use for SASL
Arran Cudbard-Bell [Thu, 6 Apr 2017 21:57:54 +0000 (17:57 -0400)]
Fix doxygen and LDAP function renames that were missed
Arran Cudbard-Bell [Thu, 6 Apr 2017 21:40:44 +0000 (17:40 -0400)]
Breakout libldap functions into a utility library
Arran Cudbard-Bell [Thu, 6 Apr 2017 21:21:00 +0000 (17:21 -0400)]
Change Hdb to Mdb in LDAP test scripts
Arran Cudbard-Bell [Thu, 6 Apr 2017 17:10:22 +0000 (13:10 -0400)]
Rework fr_strerror and friends to allow multiple messages to be logged
Arran Cudbard-Bell [Thu, 6 Apr 2017 17:09:12 +0000 (13:09 -0400)]
Adding missing RCSID macros
Arran Cudbard-Bell [Wed, 5 Apr 2017 20:38:59 +0000 (16:38 -0400)]
Fixup doxygen headers in log.c
Alan T. DeKok [Wed, 5 Apr 2017 20:03:05 +0000 (16:03 -0400)]
update worker processing time
Alan T. DeKok [Wed, 5 Apr 2017 20:02:56 +0000 (16:02 -0400)]
update processing time for NAK
Alan T. DeKok [Wed, 5 Apr 2017 19:51:45 +0000 (15:51 -0400)]
update todo's
Alan T. DeKok [Wed, 5 Apr 2017 19:29:28 +0000 (15:29 -0400)]
build issues
Alan T. DeKok [Wed, 5 Apr 2017 18:58:12 +0000 (14:58 -0400)]
compiler warnings
Alan T. DeKok [Wed, 5 Apr 2017 18:30:09 +0000 (14:30 -0400)]
add fr_strerror handling
Alan T. DeKok [Wed, 5 Apr 2017 17:56:58 +0000 (13:56 -0400)]
add fr_strerror handling
Alan T. DeKok [Wed, 5 Apr 2017 17:52:09 +0000 (13:52 -0400)]
add fr_strerror handling
Alan T. DeKok [Wed, 5 Apr 2017 17:39:43 +0000 (13:39 -0400)]
add fr_strerror handling
Alan T. DeKok [Wed, 5 Apr 2017 17:33:22 +0000 (13:33 -0400)]
add fr_strerror handling
Alan T. DeKok [Wed, 5 Apr 2017 17:25:43 +0000 (13:25 -0400)]
add fr_strerror handling
Alan T. DeKok [Wed, 5 Apr 2017 17:10:54 +0000 (13:10 -0400)]
add error printing
Alan T. DeKok [Tue, 4 Apr 2017 22:23:27 +0000 (18:23 -0400)]
doxygen
Alan T. DeKok [Tue, 4 Apr 2017 22:22:51 +0000 (18:22 -0400)]
remove unused variable
Alan T. DeKok [Tue, 4 Apr 2017 21:11:29 +0000 (17:11 -0400)]
add fr_log_t to worker, and use it for messages
Alan T. DeKok [Tue, 4 Apr 2017 20:50:11 +0000 (16:50 -0400)]
add fr_log_t to receiver, and use it for messages
Alan T. DeKok [Tue, 4 Apr 2017 20:16:37 +0000 (16:16 -0400)]
get receiver working.
the idea is to get the receiver state machine working, and then
gradually glue in the transport layers.
We now have transport read/write mostly working. They will
likely get changed to use IO contexts later
Alan T. DeKok [Tue, 4 Apr 2017 16:14:51 +0000 (12:14 -0400)]
compiler warning
Alan T. DeKok [Tue, 4 Apr 2017 16:12:18 +0000 (12:12 -0400)]
Catch 32-bit overflows
Alan T. DeKok [Tue, 4 Apr 2017 15:36:43 +0000 (11:36 -0400)]
start of receiver read functionality
it now gets channel messages, and sends them to the worker
Alan T. DeKok [Tue, 4 Apr 2017 15:35:22 +0000 (11:35 -0400)]
more messages
Alan T. DeKok [Tue, 4 Apr 2017 15:34:59 +0000 (11:34 -0400)]
signal the worker when a channel is created
which is less surprising than asking the caller to do it
Alan T. DeKok [Tue, 4 Apr 2017 14:52:18 +0000 (10:52 -0400)]
set up the fault handlers.
Now that the fault handlers are no longer retarded
Alan T. DeKok [Tue, 4 Apr 2017 14:49:44 +0000 (10:49 -0400)]
SET DEBUG STATE WHEN TOLD TO SET THE DEBUG STATE
Honestly, this one bug has cost me days of time. Why the HELL
does setting the debug state NOT WORK. EVER. The MAGIC CODE to
figure out whether or not we're in a debugger DOES NOT WORK FOR ME.
The environment variable which FORCES the code to believe that
the debugger attached sets a variable LOCAL TO THE FUNCTION,
and FAILS TO SET THE GLOBAL VARIABLE WHICH SETS THE DEBUG STATE.
<sigh>
Alan T. DeKok [Tue, 4 Apr 2017 14:19:58 +0000 (10:19 -0400)]
create network thread(s) before workers