]> git.ipfire.org Git - thirdparty/freeradius-server.git/log
thirdparty/freeradius-server.git
6 years agodoc: Don't append extras '/'
Jorge Pereira [Tue, 13 Aug 2019 18:54:23 +0000 (15:54 -0300)] 
doc: Don't append extras '/'

6 years agoDon't need to process *.md in doc/ anymore (#2882)
Jorge Pereira [Tue, 13 Aug 2019 18:27:59 +0000 (15:27 -0300)] 
Don't need to process *.md in doc/ anymore (#2882)

We do not have any *.md in doc/, so don't need it.

6 years ago doc: Convert doc/raddb/mods-available/README.adoc (#2881)
Jorge Pereira [Tue, 13 Aug 2019 18:20:16 +0000 (15:20 -0300)] 
 doc: Convert doc/raddb/mods-available/README.adoc (#2881)

* doc: Rename doc/raddb/mods-available/README.{md,adoc}

* doc: Convert doc/raddb/mods-available/README.adoc from markdown to asciidoc

6 years ago doc: Convert doc/raddb/mods-config/README.adoc from markdown to asciidoc (#2880)
Jorge Pereira [Tue, 13 Aug 2019 18:10:30 +0000 (15:10 -0300)] 
 doc: Convert doc/raddb/mods-config/README.adoc from markdown to asciidoc (#2880)

* doc: Rename doc/raddb/mods-config/README.{md,adoc}

* doc: Convert doc/raddb/mods-config/README.adoc from markdown to asciidoc

6 years ago doc: Convert doc/raddb/README.adoc from markdown to asciidoc (#2879)
Jorge Pereira [Tue, 13 Aug 2019 17:52:27 +0000 (14:52 -0300)] 
 doc: Convert doc/raddb/README.adoc from markdown to asciidoc (#2879)

* doc: Rename doc/raddb/README.{md,adoc}

* doc: Convert doc/raddb/README.adoc from markdown to asciidoc

6 years ago doc: Convert doc/directories.adoc from markdown to asciidoc (#2878)
Jorge Pereira [Tue, 13 Aug 2019 17:51:52 +0000 (14:51 -0300)] 
 doc: Convert doc/directories.adoc from markdown to asciidoc (#2878)

* doc: Rename doc/directories.{md,adoc}

* doc: Convert doc/directories.adoc from markdown to asciidoc

6 years agodoc: Convert doc/README.md to doc/README.adoc (#2832)
Jorge Pereira [Tue, 13 Aug 2019 16:56:32 +0000 (13:56 -0300)] 
doc: Convert doc/README.md to doc/README.adoc (#2832)

6 years agoremove hack
Alan T. DeKok [Tue, 13 Aug 2019 15:37:42 +0000 (11:37 -0400)] 
remove hack

6 years agoloop detect on calling instruction, which is simpler
Alan T. DeKok [Tue, 13 Aug 2019 15:32:26 +0000 (11:32 -0400)] 
loop detect on calling instruction, which is simpler

as it gets us virtual server, packet types, etc.

6 years agoRemove previous hacks to cf_data_find() and use CF_IDENT_ANY
Alan T. DeKok [Tue, 13 Aug 2019 15:24:28 +0000 (11:24 -0400)] 
Remove previous hacks to cf_data_find() and use CF_IDENT_ANY

with the caveat that the caller needs to use _cf_data_find(),
so that we are not looking for talloc type "CF_IDENT_ANY"

6 years agoformatting
Alan T. DeKok [Tue, 13 Aug 2019 14:16:14 +0000 (10:16 -0400)] 
formatting

6 years agoformat it
Alan T. DeKok [Tue, 13 Aug 2019 14:15:55 +0000 (10:15 -0400)] 
format it

6 years agoformat STRUCT, too
Alan T. DeKok [Tue, 13 Aug 2019 14:15:15 +0000 (10:15 -0400)] 
format STRUCT, too

6 years agoparse sub-structs
Alan T. DeKok [Tue, 13 Aug 2019 13:49:04 +0000 (09:49 -0400)] 
parse sub-structs

6 years agomore sanity check on STRUCT
Alan T. DeKok [Tue, 13 Aug 2019 13:48:29 +0000 (09:48 -0400)] 
more sanity check on STRUCT

that the parent MUST be of type 'struct'

that the previous MEMBER MUST be of fixed size

6 years agodon't check data types when looking up children
Alan T. DeKok [Tue, 13 Aug 2019 13:35:34 +0000 (09:35 -0400)] 
don't check data types when looking up children

We already check for da->children.  If there are no children,
then we can return, independent of data types.  The rest of the
code (i.e. adding attributes) ensures that children are added
only when necessary

6 years agoclean up and rearrange
Alan T. DeKok [Tue, 13 Aug 2019 12:24:10 +0000 (08:24 -0400)] 
clean up and rearrange

6 years agodon't have NULL entries.
Alan T. DeKok [Tue, 13 Aug 2019 03:19:40 +0000 (23:19 -0400)] 
don't have NULL entries.

NULL is used to signal the end of the list.  It shouldn't be used
as the *first* member of the list.

6 years agouse .Example instead of anything else
Alan T. DeKok [Tue, 13 Aug 2019 00:14:35 +0000 (20:14 -0400)] 
use .Example instead of anything else

6 years agoreformat to follow the standard style
Alan T. DeKok [Tue, 13 Aug 2019 00:10:54 +0000 (20:10 -0400)] 
reformat to follow the standard style

6 years agoallow calling the same server if the packet types are different
Alan T. DeKok [Mon, 12 Aug 2019 23:54:40 +0000 (19:54 -0400)] 
allow calling the same server if the packet types are different

so that we can receive an Access-Request, create a subrequest
of Accounting-Request, and then call ourselves to create a fake
"start" packet

6 years agoprevent "call" loops
Alan T. DeKok [Mon, 12 Aug 2019 23:49:59 +0000 (19:49 -0400)] 
prevent "call" loops

6 years agoword smithing
Alan T. DeKok [Mon, 12 Aug 2019 23:37:37 +0000 (19:37 -0400)] 
word smithing

6 years agodefine Q if necessary
Alan T. DeKok [Mon, 12 Aug 2019 23:32:42 +0000 (19:32 -0400)] 
define Q if necessary

6 years agoadd documentation for "call"
Alan T. DeKok [Mon, 12 Aug 2019 23:25:41 +0000 (19:25 -0400)] 
add documentation for "call"

6 years agorun the child request through the "call" subsection
Alan T. DeKok [Mon, 12 Aug 2019 23:15:34 +0000 (19:15 -0400)] 
run the child request through the "call" subsection

* call FOO means
* call virtual server FOO with a subrequest
* then run the subrequest through the subsection of the CALL
* in order to get post-processing of the reply packet

i.e. to copy packet types, etc.

6 years agodocker: The target 'crossbuild' needs the package 'rsync' previously
Jorge Pereira [Mon, 12 Aug 2019 21:15:30 +0000 (18:15 -0300)] 
docker: The target 'crossbuild' needs the package 'rsync' previously
installed into the images

6 years agoAllow 'make crossbuild' to use 'Q='
Jorge Pereira [Mon, 12 Aug 2019 21:06:45 +0000 (18:06 -0300)] 
Allow 'make crossbuild' to use 'Q='

6 years agodoc: Convert doc/schemas/logstash/README.{md,adoc} (#2877)
Jorge Pereira [Mon, 12 Aug 2019 22:32:58 +0000 (19:32 -0300)] 
doc: Convert doc/schemas/logstash/README.{md,adoc} (#2877)

6 years agomaybe "type" can be NULL
Alan T. DeKok [Mon, 12 Aug 2019 20:57:43 +0000 (16:57 -0400)] 
maybe "type" can be NULL

6 years agoREADME.md no longer exists. Don't install it.
Alan T. DeKok [Mon, 12 Aug 2019 20:56:39 +0000 (16:56 -0400)] 
README.md no longer exists.  Don't install it.

6 years agofix up debug output
Alan T. DeKok [Mon, 12 Aug 2019 20:39:38 +0000 (16:39 -0400)] 
fix up debug output

6 years agore-implement "call" based on subrequest.c
Alan T. DeKok [Mon, 12 Aug 2019 20:20:48 +0000 (16:20 -0400)] 
re-implement "call" based on subrequest.c

Hopefully yield, etc. will work.  Not well tested for now.

6 years agoadd special data type "void" for cf_data_find()
Alan T. DeKok [Mon, 12 Aug 2019 20:18:58 +0000 (16:18 -0400)] 
add special data type "void" for cf_data_find()

when we want to find a named instance, BUT where we don't know
the talloc type.

6 years agocheck if types match in cf_find() when ident1 == CF_IDENT_ANY
Alan T. DeKok [Mon, 12 Aug 2019 20:17:52 +0000 (16:17 -0400)] 
check if types match in cf_find() when ident1 == CF_IDENT_ANY

That way we can return the thing we want, not some random similar
thing.

6 years agoadd the "call" syntax and handlers
Alan T. DeKok [Mon, 12 Aug 2019 19:53:54 +0000 (15:53 -0400)] 
add the "call" syntax and handlers

6 years agofix comments and clean up
Alan T. DeKok [Mon, 12 Aug 2019 18:56:08 +0000 (14:56 -0400)] 
fix comments and clean up

unlang_subrequest_push() no longer calls unlang_subrequest()
So the comments and code handling that are not necessary

6 years agoremove unnecessary assert
Alan T. DeKok [Mon, 12 Aug 2019 18:34:40 +0000 (14:34 -0400)] 
remove unnecessary assert

6 years agodocument "detach" parameter
Alan T. DeKok [Mon, 12 Aug 2019 18:34:24 +0000 (14:34 -0400)] 
document "detach" parameter

6 years agoword smithing
Alan T. DeKok [Mon, 12 Aug 2019 17:59:47 +0000 (13:59 -0400)] 
word smithing

and add reference to "call".

6 years agomoved and cleaned up
Alan T. DeKok [Mon, 12 Aug 2019 17:31:49 +0000 (13:31 -0400)] 
moved and cleaned up

6 years agowe don't need this any more
Alan T. DeKok [Mon, 12 Aug 2019 17:24:40 +0000 (13:24 -0400)] 
we don't need this any more

6 years agoclean up loop in unlang_interpret_synchronous()
Alan T. DeKok [Mon, 12 Aug 2019 15:17:00 +0000 (11:17 -0400)] 
clean up loop in unlang_interpret_synchronous()

6 years agodon't use request->password
Alan T. DeKok [Mon, 12 Aug 2019 14:59:31 +0000 (10:59 -0400)] 
don't use request->password

6 years agoallow hex, too
Alan T. DeKok [Mon, 12 Aug 2019 13:41:24 +0000 (09:41 -0400)] 
allow hex, too

6 years agoset state->result, too
Alan T. DeKok [Mon, 12 Aug 2019 13:36:11 +0000 (09:36 -0400)] 
set state->result, too

6 years agobuild fixes
Alan T. DeKok [Mon, 12 Aug 2019 13:34:47 +0000 (09:34 -0400)] 
build fixes

6 years agosort output names so things are easier to find
Alan T. DeKok [Mon, 12 Aug 2019 13:28:10 +0000 (09:28 -0400)] 
sort output names so things are easier to find

6 years agoclean up loop in unlang_interpret_synchronous()
Alan T. DeKok [Mon, 12 Aug 2019 13:13:49 +0000 (09:13 -0400)] 
clean up loop in unlang_interpret_synchronous()

6 years agoadd tmp-float-0, and do floating point comparison in tests
Alan T. DeKok [Mon, 12 Aug 2019 12:30:43 +0000 (08:30 -0400)] 
add tmp-float-0, and do floating point comparison in tests

6 years agohandle child requests in unlang_interpret_synchronous()
Alan T. DeKok [Mon, 12 Aug 2019 11:28:31 +0000 (07:28 -0400)] 
handle child requests in unlang_interpret_synchronous()

if the parent finishes, BUT there are still children, then keep
running the children until they're done, too.

Then, clean up the children.

6 years agoset state_ctx to talloc_ctx, not to VPS
Alan T. DeKok [Mon, 12 Aug 2019 00:21:20 +0000 (20:21 -0400)] 
set state_ctx to talloc_ctx, not to VPS

6 years agofree child on error, too
Alan T. DeKok [Sun, 11 Aug 2019 16:40:47 +0000 (12:40 -0400)] 
free child on error, too

6 years agodocument Request-Lifetime for detach
Alan T. DeKok [Sun, 11 Aug 2019 14:26:25 +0000 (10:26 -0400)] 
document Request-Lifetime for detach

6 years agorun-time attribute issues are debug messages
Alan T. DeKok [Sun, 11 Aug 2019 14:26:05 +0000 (10:26 -0400)] 
run-time attribute issues are debug messages

so that they don't pollute the main log file

6 years agoword smithing
Alan T. DeKok [Sun, 11 Aug 2019 13:51:58 +0000 (09:51 -0400)] 
word smithing

6 years agoadd documentation
Alan T. DeKok [Sun, 11 Aug 2019 13:51:46 +0000 (09:51 -0400)] 
add documentation

6 years agoadd "parallel detach" syntax
Alan T. DeKok [Sun, 11 Aug 2019 12:48:31 +0000 (08:48 -0400)] 
add "parallel detach" syntax

which starts all children in the detached state

6 years agoadd simple test that "parallel empty" parses
Alan T. DeKok [Sun, 11 Aug 2019 12:29:29 +0000 (08:29 -0400)] 
add simple test that "parallel empty" parses

6 years agoas of 02433d85, the parent does *not* have to be "subrequest"
Alan T. DeKok [Sun, 11 Aug 2019 12:27:19 +0000 (08:27 -0400)] 
as of 02433d85, the parent does *not* have to be "subrequest"

6 years agotypo
Alan T. DeKok [Sat, 10 Aug 2019 18:05:35 +0000 (14:05 -0400)] 
typo

6 years agorlm_sql_oracle: We only support Oracle >= 11
Jorge Pereira [Fri, 9 Aug 2019 19:57:42 +0000 (16:57 -0300)] 
rlm_sql_oracle: We only support Oracle >= 11

6 years agorlm_sql_oracle: Inform the supported versions
Jorge Pereira [Fri, 9 Aug 2019 19:55:52 +0000 (16:55 -0300)] 
rlm_sql_oracle: Inform the supported versions

6 years agorlm_redis_ippool: Comment typo in config.
Terry Burton [Sat, 10 Aug 2019 17:32:45 +0000 (18:32 +0100)] 
rlm_redis_ippool: Comment typo in config.

6 years agodocument the subrequest keyword
Alan T. DeKok [Sat, 10 Aug 2019 17:58:01 +0000 (13:58 -0400)] 
document the subrequest keyword

6 years agowe can have "detach" inside of an "if" statement, too
Alan T. DeKok [Sat, 10 Aug 2019 17:42:03 +0000 (13:42 -0400)] 
we can have "detach" inside of an "if" statement, too

6 years agofix for new syntax
Alan T. DeKok [Sat, 10 Aug 2019 17:20:03 +0000 (13:20 -0400)] 
fix for new syntax

6 years agopoint to correct file
Alan T. DeKok [Sat, 10 Aug 2019 11:44:31 +0000 (07:44 -0400)] 
point to correct file

6 years agoremove warnings.
Alan T. DeKok [Fri, 9 Aug 2019 20:58:52 +0000 (16:58 -0400)] 
remove warnings.

the whole "operator" section should be in another page,
in the "unlang" section

6 years agodon't create child ctx for $INCLUDE
Alan T. DeKok [Fri, 9 Aug 2019 20:53:22 +0000 (16:53 -0400)] 
don't create child ctx for $INCLUDE

instead, use the current one.  And then check it post-facto
to see if it mucked up the stack depth.

6 years agonote that child is freed on exit
Alan T. DeKok [Fri, 9 Aug 2019 20:51:53 +0000 (16:51 -0400)] 
note that child is freed on exit

6 years agodoc: Convert doc/howto/*.md to doc/howto/*.adoc
Jorge Pereira [Thu, 8 Aug 2019 17:58:05 +0000 (14:58 -0300)] 
doc: Convert doc/howto/*.md to doc/howto/*.adoc

6 years agodoc: Rename doc/howto/*.md to doc/howto/*.adoc
Jorge Pereira [Wed, 31 Jul 2019 20:19:59 +0000 (17:19 -0300)] 
doc: Rename doc/howto/*.md to doc/howto/*.adoc

6 years agotypo
Alan T. DeKok [Fri, 9 Aug 2019 18:54:18 +0000 (14:54 -0400)] 
typo

6 years agocheck for TTY and debug level early
Alan T. DeKok [Fri, 9 Aug 2019 18:35:01 +0000 (14:35 -0400)] 
check for TTY and debug level early

so that we don't erroneously add log prefixes, and then later
stop using them.

It's more important that debug output look OK, than the much rarer
case of non-debug output sometimes looking off

6 years agoadded documentation for parallel keyword
Alan T. DeKok [Fri, 9 Aug 2019 17:27:58 +0000 (13:27 -0400)] 
added documentation for parallel keyword

6 years agosimplify use of unwind
Alan T. DeKok [Fri, 9 Aug 2019 16:38:15 +0000 (12:38 -0400)] 
simplify use of unwind

now that we have BEGIN/END markers on the stack

6 years agonotes on more sanity checking
Alan T. DeKok [Fri, 9 Aug 2019 16:36:25 +0000 (12:36 -0400)] 
notes on more sanity checking

6 years agoupdate the stack with BEGIN / END markings
Alan T. DeKok [Fri, 9 Aug 2019 15:37:58 +0000 (11:37 -0400)] 
update the stack with BEGIN / END markings

so that we can tell auto-pushed attrs as separate from manual
BEGIN / END commands

and fix "unwind" so that it unwinds only to the parent nesting

6 years agomove struct member number into stack frame
Alan T. DeKok [Fri, 9 Aug 2019 15:24:10 +0000 (11:24 -0400)] 
move struct member number into stack frame

to allow for nested structs

6 years agozero out the stack frame on push
Alan T. DeKok [Fri, 9 Aug 2019 15:21:08 +0000 (11:21 -0400)] 
zero out the stack frame on push

6 years agomake the stack a structure
Alan T. DeKok [Fri, 9 Aug 2019 15:18:03 +0000 (11:18 -0400)] 
make the stack a structure

in preparation for nested 'struct', and for better handling
of manual vs automatic BEGIN / END things

6 years agobetter debug messages so that we see when / why it's not replying
Alan T. DeKok [Fri, 9 Aug 2019 14:41:44 +0000 (10:41 -0400)] 
better debug messages so that we see when / why it's not replying

6 years agodefault is NOOP, not NOTFOUND
Alan T. DeKok [Fri, 9 Aug 2019 14:41:31 +0000 (10:41 -0400)] 
default is NOOP, not NOTFOUND

6 years agoword smithing
Alan T. DeKok [Fri, 9 Aug 2019 01:20:41 +0000 (21:20 -0400)] 
word smithing

6 years agoremove test attributes
Alan T. DeKok [Fri, 9 Aug 2019 00:39:28 +0000 (20:39 -0400)] 
remove test attributes

6 years agoclean up and finalize use of ctx->stack
Alan T. DeKok [Thu, 8 Aug 2019 23:17:26 +0000 (19:17 -0400)] 
clean up and finalize use of ctx->stack

so that "previous_attr" becomes "relative_attr".

We also add a dict_ctx_push() function which pushes attributes
onto ctx->stack.

And dict_ctx_unwind() for unwinding the stack multiple frames
when necessary.

There are likely edge cases which aren't handled well.  The code
sometimes does an *implicit* push of struct/tlv.  And other times
an *explicit* push / pop.  The differences are not tracked, which
makes it more difficult to automatically determine what to do.

6 years agoremove erroneous debug message
Alan T. DeKok [Thu, 8 Aug 2019 20:59:35 +0000 (16:59 -0400)] 
remove erroneous debug message

6 years agoclean up process_attribute()
Alan T. DeKok [Thu, 8 Aug 2019 18:40:35 +0000 (14:40 -0400)] 
clean up process_attribute()

to better handle structs and the new stack

6 years agoget rid of "parent" and block_tlv stuff
Alan T. DeKok [Thu, 8 Aug 2019 18:00:43 +0000 (14:00 -0400)] 
get rid of "parent" and block_tlv stuff

And just use a global stack instead

6 years agoremove "block_vendor"
Alan T. DeKok [Thu, 8 Aug 2019 17:44:34 +0000 (13:44 -0400)] 
remove "block_vendor"

and replace with a stack of attributes

6 years agoremove unused "vendor" field
Alan T. DeKok [Thu, 8 Aug 2019 17:43:30 +0000 (13:43 -0400)] 
remove unused "vendor" field

6 years agouse macros to simplify / standardize logic
Alan T. DeKok [Thu, 8 Aug 2019 17:26:46 +0000 (13:26 -0400)] 
use macros to simplify / standardize logic

6 years agouse "." instead of BEGIN/END-TLV
Alan T. DeKok [Thu, 8 Aug 2019 16:56:01 +0000 (12:56 -0400)] 
use "." instead of BEGIN/END-TLV

6 years agomove data type examples to the "data type" page
Alan T. DeKok [Thu, 8 Aug 2019 13:46:31 +0000 (09:46 -0400)] 
move data type examples to the "data type" page

6 years agoadd parsing for STRUCT as sub-structures
Alan T. DeKok [Thu, 8 Aug 2019 13:43:33 +0000 (09:43 -0400)] 
add parsing for STRUCT as sub-structures

which uses the key field.  Note that this does not yet work
because we don't have parent/child relationships in the
dict_from_file_ctx_t structure.

6 years agodisallow uint64 key fields, and update sanity checks
Alan T. DeKok [Thu, 8 Aug 2019 13:41:40 +0000 (09:41 -0400)] 
disallow uint64 key fields, and update sanity checks

6 years agouse macro for common code
Alan T. DeKok [Thu, 8 Aug 2019 13:04:48 +0000 (09:04 -0400)] 
use macro for common code

6 years agodoc: some tweaks
Jorge Pereira [Wed, 7 Aug 2019 21:54:06 +0000 (18:54 -0300)] 
doc: some tweaks

6 years agobuild fixes
Alan T. DeKok [Wed, 7 Aug 2019 19:57:18 +0000 (15:57 -0400)] 
build fixes