Luigi Rizzo [Sun, 29 Jul 2007 10:13:14 +0000 (10:13 +0000)]
build the version of sip_tech with no send_digit_begin
at load time instead of duplicating the initializer.
This should remove the risk of forgetting fields in the
initializer.
Luigi Rizzo [Sun, 29 Jul 2007 09:27:30 +0000 (09:27 +0000)]
remove bit position from description of SIP_* flags.
use AST_FORMAT_AUDIO_MASK instead of playing with AST_FORMAT_MAX_AUDIO
to determine audio formats.
There is a dubious use of AST_FORMAT_MAX_AUDIO in sip_request_call()
which surely needs fixing, namely:
/* mask request with some set of allowed formats.
* XXX this needs to be fixed.
* The original code uses AST_FORMAT_AUDIO_MASK, but it is
* unclear what to use here. We have global_capabilities, which is
* configured from sip.conf, and sip_tech.capabilities, which is
* hardwired to all audio formats.
*/
The latter is possibly something to backport when fixed.
Luigi Rizzo [Sun, 29 Jul 2007 08:58:10 +0000 (08:58 +0000)]
back on cleaning up the usage of flags.
Move together flags used in the same way (e.g. dialog only,
dialog-peer, ...) so it will become easier to deal with them
in a more systematic way.
This is being done in stages so it will be easier to detect
breakage, if any should occur.
Luigi Rizzo [Sat, 28 Jul 2007 23:43:35 +0000 (23:43 +0000)]
add some documentation to auto_congest(), and some
dialog_ref/unref (they are a no-op at the moment).
Also clean a pointer after freeing memory to avoid
dangling references, and write a for() loop in canonical form.
In practice, everything in this commit is a no-op.
Luigi Rizzo [Sat, 28 Jul 2007 16:25:25 +0000 (16:25 +0000)]
start introducing hooks for reference counts on dialog descriptors.
This commit is, for all practical purposes, a no-op, as it only
introduces the dialog_ref() and dialog_unref() methods, and uses them
in a few places (not all the places where they would be needed).
The goal is to start annotating the code with these calls, so the transition
to a proper container will be easier.
Luigi Rizzo [Sat, 28 Jul 2007 07:44:16 +0000 (07:44 +0000)]
make use of received= and rport= fields in sip replies.
In a nutshell, these fields are used to tell a sip entity
the address and port its request came from, and are extremely
useful in the presence of NATs, especially with symmetric NATs
where STUN is totally ineffective.
This patch stores the address and port in the 'ourip' field of
the dialog descriptor, so they can be reused in subsequent transactions.
As it is, it works well for things like REGISTER requiring authentication,
because the second REGISTER request (with auth credentials) will carry
the correct address. Maybe it can also be useful, in case of an address
change, to do one or both of the following:
+ propagate the new address to the parent user/peer descriptor so that new
dialogs will use the correct address from the beginning.
This is trivial to implement, I am just waiting for feedback on this.
+ re-issue a request in case of an address change. This a lot less trivial,
maybe unnecessary, and probably covered by the previous item.
I would seriously consider this patch for addition to 1.4 and 1.2.
The code is very little intrusive, and it would solve in a correct
way the nat traversal problems for which externip/externaddr/stunaddr
are only a partial and expensive workaround.
(closes issue #10310)
Reported by: prashant_jois
Patches:
cdr_pgsql.patch uploaded by prashant (license 114)
Finish the Postgresql connection after the log messages are printed so we don't access invalid memory.
(closes issue #10323)
Reported by: julianjm
Patches:
chan_sip_device_state_hold_fix.v1.diff.txt uploaded by julianjm (license 99)
Clear ONHOLD flag when decrementing the onHold peer count. If we did not do this the count may keep decreasing.
Steve Murphy [Fri, 27 Jul 2007 15:46:20 +0000 (15:46 +0000)]
These fixes take care of two problems: a complaint in asterisk-dev that goto's aren't working in trunk, a side effect of the move to commas as arg seps in apps and funcs; and a problem I spotted myself with dial's 'e' option, where gotos were off by one, because I forgot to set the AUTOLOOP flag in the peer channel.
(closes issue #10302)
Reported by: litnialex
If a DTMF end frame comes from a channel without a begin and it is going to a technology that only accepts end frames (aka INFO) then use the minimum DTMF duration if one is not in the frame already.
AST_DEVMODE was defined in trunk, but not in 1.4. When Asterisk is compiled
under dev mode, AST_DEVMODE will get defined in buildopts.h. Change 1.4 to
define it in the same way that trunk does. Also, revert the change that added
this define in the Makefile
The advantage to doing it this way is that buildopts.h gets installed when
you install Asterisk. Then, when building any out of tree modules, or
building asterisk-addons, these modules know which options the rest of Asterisk
was built with.
Fixes to get ast_backtrace working properly. The AST_DEVMODE macro was never defined so the majority of ast_backtrace never
attempted compilation. The makefile now defines AST_DEVMODE if configure was run with --enable-dev-mode. Also, changes were
made to acccomodate 64 bit systems in ast_backtrace.
Thanks to qwell, kpfleming, and Corydon76 for their roles in allowing me to get this committed
Two consecutive calls to PQfinish could occur, meaning free gets called on the same variable twice.
This patch sets the connection to NULL after calls to PQfinish so that the problem does not occur.
Also in this patch, prashant_jois informed me that it is safe to pass a null pointer to PQfinish, so
I have removed the check for conn's existence from my_unload_module.
(closes issue 10295, reported by junky, patched by me with input from prashant_jois)
Russell Bryant [Thu, 26 Jul 2007 14:49:51 +0000 (14:49 +0000)]
Merge a big batch of documentation fixes for escaping, marking URLs, places
where verbatim text went off the end of the page on the PDF, and various
other improvements
(closes issue #10307, IgorG)
Russell Bryant [Thu, 26 Jul 2007 13:26:44 +0000 (13:26 +0000)]
Revert some changes to call abs() on the result of ast_random().
* random() is defined to return a positive result, and now ast_random()
will always do so as well
Steve Murphy [Thu, 26 Jul 2007 01:33:55 +0000 (01:33 +0000)]
The upgrade of application argument separators to comma has an effect on AEL; I commented out the code that substitutes commas with vertbars, so we can get apps to parse their args correctly.
This fix solves problem with intense squelch noise when someone joins conf in bug 9430; We repro'd the problem with meetme opts of 'CciMo'; Josh Colp supplied this patch, and I'm applying it. It looks like playing the recorded username will louse up the next thing played into the channel. Josh rearranged the code so as to start things over before playing data directly into the conference.
........
(closes issue #10303)
Reported by: jtodd
Add SPEECH_DTMF_TERMINATOR variable so the user can specify the digit to terminate a DTMF string with. If none is specified then no terminator will be used.
Luigi Rizzo [Wed, 25 Jul 2007 21:58:13 +0000 (21:58 +0000)]
silence a warning in ast-devmode on a potentially uninitialized var.
At first sight (but the function is very large so i am not 100% sure)
the code seems correct, so maybe my compiler is just not smart
enough to figure that out at the optimization level it has.
Added a membercount variable to call_queue struct which keeps track of the number of logged in members in a particular queue.
This makes it so that the 'n' option for Queue() can act properly depending on which strategy is used. If the strategy is
roundrobin, rrmemory, or ringall, we want to ring each phone once before moving on in the dialplan. However, if any other strategy is
used, we will only ring one phone since it cannot be guaranteed that a different phone will ring on subsequent attempts to ring a phone.
As a side effect of this, the QUEUE_MEMBER_COUNT dialplan function now just reads the membercount variable instead of traversing through
the member list to figure out how many members there are.
Special thanks to blitzrage for helping to test this out.
(closes issue #10127, reported by bcnit, patched by me, tested by blitzrage)
Jason Parker [Tue, 24 Jul 2007 15:35:58 +0000 (15:35 +0000)]
The chan_skinny Dial() syntax was funky. You had to do Dial(Skinny/line@device)
This allows you to just Dial(Skinny/line), as long as line isn't ambiguous.
Note that this does not remove or deprecate the "old" syntax, as it's still
quite useful - even moreso if shared lines get implemented.
Initial patch by me, with some changes and suggestions from wedhorn.
(closes issue #10263)
Luigi Rizzo [Tue, 24 Jul 2007 14:49:49 +0000 (14:49 +0000)]
two small fixes when using stun (reported by Marta Carbone):
+ externexpire was not initialized properly;
+ stunaddr was not handled properly on a sip reload
You need to put static in front of a static RWLIST declaration to make it really static... and don't call AST_RWLIST_HEAD_DESTROY on a statically declared list.
It was our stated intention for 1.4 that files created in app_voicemail should
depend upon the umask. Unfortunately, mkstemp() creates files with mode 0600,
regardless of the umask. This corrects that deficiency.
Enhance AGI with several fixes:
- Makes the structures handling external AGI commands a bit more thread-safe
- Makes AGI transparently work with both live and hungup channels
- DeadAGI is hence no longer necessary and is deprecated
- CLI bug fixes
- Commands will refuse to run if the channel is dead and the command is nonsensical
for dead channels.
Merge the dialplan_aesthetics branch. Most of this patch simply converts applications
using old methods of parsing arguments to using the standard macros. However, the big
change is that the really old way of specifying application and arguments separated by
a comma will no longer work (e.g. NoOp,foo|bar). Instead, the way that has been
recommended since long before 1.0 will become the only method available (e.g. NoOp(foo,bar).
(closes issue #5866)
Reported by: tyler
Do not force channel format changes when a generator is present. The generator may have changed the formats itself and changing them back would cause issues.
Luigi Rizzo [Mon, 23 Jul 2007 14:45:22 +0000 (14:45 +0000)]
add two missing entries in the replica of the sip_tech that
does not use DTMF BEGIN frames.
1.4 seems correct (it does not have the two fields).
However, as this bug shows, the current way of creating the sip_tech
replica is too error-prone, one can easily forget to update one of
the two entries. Perhaps it would be better to create sip_tech_info
expliclty at module load, by doing
Russell Bryant [Mon, 23 Jul 2007 14:21:41 +0000 (14:21 +0000)]
(closes issue #10192)
Reported by: bbryant
Patches:
20070720__core_debug_by_file.patch uploaded by bbryant (license 36)
(with some modifications by me)
Tested by: russell, bbryant
This set of changes introduces the ability to set the core debug or verbose
levels on a per-file basis. Interestingly enough, in 1.4, you have the ability
to set core debug for a single file, but that functionality was accidentally
lost in the conversion of the CLI commands to the new format.
This patch improves upon what was in 1.4 by letting you set it for more than 1
file, and by also supporting verbose.
*** Janitor Project ***
This patch also introduces a new macro, ast_verb(), which is similar
to ast_debug(). Setting the per file verbose value only works for messages that
use this macro. Converting existing uses of ast_verbose() can be done like:
if (option_debug > 2)
ast_verbose(VERBOSE_PREFIX_3 "Something useful\n");
Luigi Rizzo [Mon, 23 Jul 2007 14:18:04 +0000 (14:18 +0000)]
introduce two functions, map_x_s() and map_s_x(), to map
between integers and strings using a single translation table,
and use them in a few places instead of ad-hoc routines
that duplicate the table.
On passing, note that REFER_CONFIRMED is never used, and add a
few comments.
(closes issue #10268)
Reported by: mvanbaak
Patches:
chan_skinny_openbsd.diff uploaded by mvanbaak (license 7)
Add another OS that has to use the Macros for byte ordering.
Use a signed integer for storing the number of bytes in the packet read from
the network. Using an unsigned value here made it impossible to handle an
error returned from recvfrom(). Furthermore, in the case that recvfrom()
did return an error, this would cause a crash due to a heap overflow.
(closes issue #10265, reported by and fix suggested by timrobbins)
Luigi Rizzo [Sun, 22 Jul 2007 19:08:37 +0000 (19:08 +0000)]
comment and slightly restructure handle_request() in the part that handles
responses, so that there is a common exit point.
Mark two places where probably we could return -1 instead of 0 to report
an error to the caller.
(change triggered by investigations on how the 'SIP_PKT_IGNORE' field was used).
Luigi Rizzo [Sun, 22 Jul 2007 18:41:57 +0000 (18:41 +0000)]
Cleanup of flags used in struct sip_request, moving them to
individual variables. Apart from SIP_PKT_IGNORE which was used
a zillion times, the other two are used seldom.
On passing:
- move the arrays to the end of struct sip_request, so a (small)
buffer overflow is less likely to overwrite the other fields;
- note that the 'ignore' argument to handle_invite_replaces() is not
used and should be removed (will be done in a separate commit).
Luigi Rizzo [Sun, 22 Jul 2007 08:42:24 +0000 (08:42 +0000)]
document the way sipdebug works, and implement it through
variables and not flags.
NOTE:
The old behaviour (preserved in this commit) is that if sipdebug
is set in the config file, it can only be disabled by reloading the
config. I am not sure if this is accidental or voluntary, but it
is really unconvenient and I think it should be handled in the same
way as other options i.e. consider requests from the config file
or the cli (or the command line) to be fully equivalent and act on
the same status variable.
Luigi Rizzo [Sat, 21 Jul 2007 17:32:00 +0000 (17:32 +0000)]
Add a note to document how the temporary 'pvt' should be initialized
before using it.
I am unclear on the details right now so i hope someone can comment
more. The obvious (and lazy) approach would be to bzero() all of it
(except for the string pool), but isn't that too much work ?
Feedback wanted here...
Luigi Rizzo [Sat, 21 Jul 2007 03:53:17 +0000 (03:53 +0000)]
use explicit variables for things that don't need to
be stored in ast_flags. First victim is 'SIP_NO_HISTORY'
replaced by a 'do_history' field in the sip_pvt structure.
projects / thirdparty / asterisk.git / log
