asterisk-tools:lbts-asterisk-with-uk-clid.git
8 years agoMerge in revision 269335. Updated ChangeLog and .version files. v1.4.33-rc2
Leif Madsen [Wed, 9 Jun 2010 17:55:11 +0000 (17:55 +0000)]
Merge in revision 269335. Updated ChangeLog and .version files.

git-svn-id: http://svn.asterisk.org/svn/asterisk/tags/1.4.33-rc2@269350 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoCreate 1.4.33-rc2 from 1.4.33-rc1.
Leif Madsen [Wed, 9 Jun 2010 17:48:42 +0000 (17:48 +0000)]
Create 1.4.33-rc2 from 1.4.33-rc1.
Shoutouts to pabelanger and the Maniacal Maniac.

git-svn-id: http://svn.asterisk.org/svn/asterisk/tags/1.4.33-rc2@269348 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoUse autotagged externals v1.4.33-rc1
Leif Madsen [Tue, 1 Jun 2010 15:59:25 +0000 (15:59 +0000)]
Use autotagged externals

git-svn-id: http://svn.asterisk.org/svn/asterisk/tags/1.4.33-rc1@266648 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoImporting files for 1.4.33-rc1 release.
Leif Madsen [Tue, 1 Jun 2010 15:59:14 +0000 (15:59 +0000)]
Importing files for 1.4.33-rc1 release.

git-svn-id: http://svn.asterisk.org/svn/asterisk/tags/1.4.33-rc1@266647 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoCreating tag for the release of asterisk-1.4.33-rc1
Leif Madsen [Tue, 1 Jun 2010 15:57:23 +0000 (15:57 +0000)]
Creating tag for the release of asterisk-1.4.33-rc1

git-svn-id: http://svn.asterisk.org/svn/asterisk/tags/1.4.33-rc1@266646 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoPrevent CLI prompt from distorting output of lines shorter than the prompt.
Tilghman Lesher [Tue, 1 Jun 2010 15:17:46 +0000 (15:17 +0000)]
Prevent CLI prompt from distorting output of lines shorter than the prompt.

Uses the VT100 method of clearing the line from the cursor position to the
end of the line:  Esc-0K

(closes issue #17160)
 Reported by: coolmig
 Patches:
       20100531__issue17160.diff.txt uploaded by tilghman (license 14)
 Tested by: coolmig

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@266585 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoFix formatting issue with previous patch.
Paul Belanger [Tue, 1 Jun 2010 14:57:49 +0000 (14:57 +0000)]
Fix formatting issue with previous patch.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@266580 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoMissing fallback to audio fax feature when T.38 re-INVITE failed
Paul Belanger [Tue, 1 Jun 2010 14:54:05 +0000 (14:54 +0000)]
Missing fallback to audio fax feature when T.38 re-INVITE failed

When a T.38 re-INVITE failed with an 488 or 606 answer, we should
fallback to audio fax by send a re-re-INVITE without T.38. The
function is backported from 1.6 asterisk.

(closes issue #16795)
Reported by: vrban

(closes issue #16692)
Reported by: vrban
Patches:
      t38_fallback_to_audio_v3.patch uploaded by vrban (license 756)
Tested by: lmadsen, vrban, haggard

https://reviewboard.asterisk.org/r/514/

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@266579 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoReverting patch and reopening issue #16784, as patch breaks color display.
Tilghman Lesher [Sun, 30 May 2010 04:43:28 +0000 (04:43 +0000)]
Reverting patch and reopening issue #16784, as patch breaks color display.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@266437 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoUse sigaction for signals which should persist past the initial trigger, not signal.
Tilghman Lesher [Wed, 26 May 2010 21:11:44 +0000 (21:11 +0000)]
Use sigaction for signals which should persist past the initial trigger, not signal.

If you call signal() in a Solaris signal handler, instead of just resetting
the signal handler, it causes the signal to refire, because the signal is not
marked as handled prior to the signal handler being called.  This effectively
causes Solaris to immediately exceed the threadstack in recursive signal
handlers and crash.

(closes issue #17000)
 Reported by: rmcgilvr
 Patches:
       20100526__issue17000.diff.txt uploaded by tilghman (license 14)
 Tested by: rmcgilvr

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@266142 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoadd dahdi_func_write to zap_tech structure
David Vossel [Wed, 26 May 2010 20:33:00 +0000 (20:33 +0000)]
add dahdi_func_write to zap_tech structure

This was supposed to be committed with r263292, the back-port
of teh DAHDI buffer policy dial string option

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@266140 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoMake AgentComplete message more consistent.
Mark Michelson [Wed, 26 May 2010 18:21:10 +0000 (18:21 +0000)]
Make AgentComplete message more consistent.

At times, the "Member" field was not specified during the event.
It's there now.

(closes issue #15638)
Reported by: elbriga
Patches:
      patchAppQueueAgentComplete.diff uploaded by elbriga (license 482)

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@266004 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoNot finding rows in the DB does not rise to the level of a warning.
Tilghman Lesher [Wed, 26 May 2010 16:21:00 +0000 (16:21 +0000)]
Not finding rows in the DB does not rise to the level of a warning.

(closes issue #17062)
 Reported by: drookie
 Patches:
       20100525__issue17062.diff.txt uploaded by tilghman (license 14)

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@265910 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agofixes build issue with zaptel
David Vossel [Tue, 25 May 2010 17:11:40 +0000 (17:11 +0000)]
fixes build issue with zaptel

(closes issue #17394)
Reported by: aragon
Patches:
      half_buffer_fix.diff uploaded by dvossel (license 671)
Tested by: aragon

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@265613 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoDon't mark the cdr records of unanswered queue calls with "NOANSWER". This restores...
Matthew Nicholson [Tue, 25 May 2010 16:48:19 +0000 (16:48 +0000)]
Don't mark the cdr records of unanswered queue calls with "NOANSWER".  This restores the behavior prior to r258670.

(closes issue #17334)
Reported by: jvandal
Patches:
      queue-cdr-fixes1.diff uploaded by mnicholson (license 96)
Tested by: aragon, jvandal

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@265610 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoMerged revisions 265320,265467 via svnmerge from
Terry Wilson [Tue, 25 May 2010 13:33:21 +0000 (13:33 +0000)]
Merged revisions 265320,265467 via svnmerge from
https://origsvn.digium.com/svn/asterisk/trunk

........
  r265320 | twilson | 2010-05-24 14:06:40 -0500 (Mon, 24 May 2010) | 14 lines

  Add the FullyBooted AMI event

  It is possible to connect to the manager interface before all Asterisk modules
  are loaded. To ensure that an application does not send AMI actions that might
  require a module that has not yet loaded, the application can listen for the
  FullyBooted manager event. It will be sent upon connection if all modules have
  been loaded, or as soon as loading is complete. The event:

     Event: FullyBooted
     Privilege: system,all
     Status: Fully Booted

  Review: https://reviewboard.asterisk.org/r/639/
........
  r265467 | twilson | 2010-05-24 17:21:58 -0500 (Mon, 24 May 2010) | 1 line

  Merge the rest of the FullyBooted patch
........

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@265570 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agofixes segfault when using generic plc
David Vossel [Mon, 24 May 2010 19:37:55 +0000 (19:37 +0000)]
fixes segfault when using generic plc

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@265365 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoDon't hang up on a queue caller if the file we attempt to play does not exist.
Mark Michelson [Fri, 21 May 2010 20:59:14 +0000 (20:59 +0000)]
Don't hang up on a queue caller if the file we attempt to play does not exist.

This also fixes a documentation mistake in file.h that made my original attempt
to correct this problem not work correctly.

(closes issue #17061)
Reported by: RoadKill

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@265089 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoFix grammatical error in comment.
Mark Michelson [Fri, 21 May 2010 16:53:53 +0000 (16:53 +0000)]
Fix grammatical error in comment.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@264999 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoAllow ast_safe_sleep to defer specific frames until after the sleep has concluded.
Mark Michelson [Fri, 21 May 2010 16:28:34 +0000 (16:28 +0000)]
Allow ast_safe_sleep to defer specific frames until after the sleep has concluded.

From reviewboard

Background:
A Digium customer discovered a somewhat odd bug. The setup is that parties A
and B are bridged, and party A places party B on hold. While party B is
listening to hold music, he mashes a bunch of DTMF. Party A takes party
B off hold while this is happening, but party B continues to hear hold
music. I could reproduce this about 1 in 5 times.

The issue:
When DTMF features are enabled and a user presses keys, the channel that
the DTMF is streamed to is placed in an ast_safe_sleep for 100 ms, the
duration of the emulated tone. If an AST_CONTROL_UNHOLD frame is read
from the channel during the sleep, the frame is dropped. Thus the
unhold indication is never made to the channel that was originally placed
on hold.

The fix:
Originally, I discussed with Kevin possible ways of fixing the specific
problem reported. However, we determined that the same type of problem
could happen in other situations where ast_safe_sleep() is used. Using
autoservice as a model, I modified ast_safe_sleep_conditional() to
defer specific frame types so they can be re-queued once the sleep has
finished. I made a common function for determining if a frame should
be deferred so that there are not two identical switch blocks to
maintain.

Review: https://reviewboard.asterisk.org/r/674/

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@264996 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoast_callerid_parse() had a path that left name uninitialized.
Richard Mudgett [Thu, 20 May 2010 23:23:21 +0000 (23:23 +0000)]
ast_callerid_parse() had a path that left name uninitialized.

Several callers of ast_callerid_parse() do not initialize the name
parameter before calling thus there is the potential to use an
uninitialized pointer.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@264820 f38db490-d61c-443f-a65b-d21fe96a405b

8 years ago1.4 version of PLC fix.
Mark Michelson [Thu, 20 May 2010 15:59:44 +0000 (15:59 +0000)]
1.4 version of PLC fix.

Analogous to trunk revision 264452, but without the change
to chan_sip since it is not necessary in this branch.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@264541 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoSet quieted flag when receiving a dtmf tone during playback in speechbackground.
Matthew Nicholson [Wed, 19 May 2010 20:01:38 +0000 (20:01 +0000)]
Set quieted flag when receiving a dtmf tone during playback in speechbackground.

(closes issue #16966)
Reported by: asackheim

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@264334 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoInternal timing is now on by default, if you're using DAHDI 2.3 or above.
Tilghman Lesher [Wed, 19 May 2010 17:41:29 +0000 (17:41 +0000)]
Internal timing is now on by default, if you're using DAHDI 2.3 or above.

The reason for ensuring DAHDI 2.3 or above is that this version ensures that
a timer is always available, whereas in previous versions, it was possible
for DAHDI to be loaded, but have no drivers to actually generate timing.  If
internal_timing was turned on in this circumstance, a complete lack of audio
would result.  This is the reason why internal_timing was not on by default.
However, now that DAHDI ensures the availability of a timer, there is no
reason for this setting to be off (and in fact, it solves a great many initial
user problems).

(closes issue #15932)
 Reported by: dimas
 Patches:
       20100519__issue15932.diff.txt uploaded by tilghman (license 14)
 Tested by: tilghman

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@264248 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agofix incorrectly typed indications for [nz] stutter and dialrecall
Alec L Davis [Wed, 19 May 2010 08:23:07 +0000 (08:23 +0000)]
fix incorrectly typed indications for [nz] stutter and dialrecall

(closes issue #17359)
Reported by: alecdavis
Patches:
      bug17359.diff.txt uploaded by alecdavis (license 585)

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@264056 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoBecause progress is called multiple times, across several frames, we must persist...
Tilghman Lesher [Wed, 19 May 2010 06:32:27 +0000 (06:32 +0000)]
Because progress is called multiple times, across several frames, we must persist states when detecting multitone sequences.

(closes issue #16749)
 Reported by: dant
 Patches:
       dsp.c-bug16749-1.patch uploaded by dant (license 670)
 Tested by: dant

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@263949 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoModify directory name reading to be interrupted with operator or pound escape.
Jeff Peeler [Tue, 18 May 2010 18:54:58 +0000 (18:54 +0000)]
Modify directory name reading to be interrupted with operator or pound escape.

In the case of accidentally entering the wrong first three letters for the
reading, users could be very frustrated if the name listing is very long. This
allows interrupting the reading by pressing 0 or #. 0 will attempt to execute
a configured operator (o) extension and # will exit and proceed in the
dialplan.

ABE-2200

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@263769 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoFix logic error when checking for a devstate provider.
Mark Michelson [Mon, 17 May 2010 22:00:28 +0000 (22:00 +0000)]
Fix logic error when checking for a devstate provider.

When using strsep, if one of the list of specified separators is not found,
it is the first parameter to strsep which is now NULL, not the pointer returned
by strsep.

This issue isn't especially severe in that the worst it is likely to do is waste
some cycles when a device with no '/' and no ':' is passed to ast_device_state.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@263639 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoRemove arbitrary size limitation for hints.
Mark Michelson [Mon, 17 May 2010 21:48:46 +0000 (21:48 +0000)]
Remove arbitrary size limitation for hints.

(closes issue #17257)
Reported by: tim_ringenbach
Patches:
      hints_crash_fix.diff uploaded by tim ringenbach (license 540)

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@263637 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoManager cookies are not compatible with RFC2109.
Leif Madsen [Mon, 17 May 2010 14:35:18 +0000 (14:35 +0000)]
Manager cookies are not compatible with RFC2109.

The Version field in the cookies we're setting contain quotes around the version
number which is not compatible with RFC2109 and breaks some implementations.

(closes issue #17231)
Reported by: ecarruda
Patches:
      manager_rfc2109-trunk-v1.patch uploaded by ecarruda (license 559)
      manager_rfc2109-1.6.2-v1.patch uploaded by ecarruda (license 559)
Tested by: ecarruda, russell

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@263456 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoUpdate link to new version of core sounds.
Leif Madsen [Mon, 17 May 2010 14:04:57 +0000 (14:04 +0000)]
Update link to new version of core sounds.

The latest version of the core sounds files 1.4.19 now includes the missing
queue-minute sound file which is called by app_queue but which has been
missing.

(closes issue #17123)
Reported by: n8ideas

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@263374 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agobackport of DAHDI buffer policy dial string option
David Vossel [Mon, 17 May 2010 13:01:39 +0000 (13:01 +0000)]
backport of DAHDI buffer policy dial string option

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@263292 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoFix internal timing not working with Zaptel
Jeff Peeler [Thu, 13 May 2010 23:08:13 +0000 (23:08 +0000)]
Fix internal timing not working with Zaptel

dahdi_compat.h was not being included in channel.c when used with
Zaptel and wasn't in file.c at all.

(closes issue #15250)
Reported by: mneuhauser
Patches:
      dahdi_compat.patch uploaded by mneuhauser (license 425)
Tested by: IgorG

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@263112 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agofixes app_meetme dsp error
David Vossel [Wed, 12 May 2010 17:00:04 +0000 (17:00 +0000)]
fixes app_meetme dsp error

We attempted to detect silence after translating a frame
from signed linear.  This caused a flooding of errors.  To
resolve this the code to detect silence was moved before the
translation.

(closes issue #17133)
Reported by: jsdyer

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@262662 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoUse a less silly method for modifying a flex-generated file.
Jason Parker [Tue, 11 May 2010 19:55:42 +0000 (19:55 +0000)]
Use a less silly method for modifying a flex-generated file.

The sed syntax that was used wasn't actually valid, causing some versions to
choke.  This is the method that is used in 1.6.x+ for similar changes.

(closes issue #16696)
Reported by: bklang
Patches:
      16696-sedfix.diff uploaded by qwell (license 4)
Tested by: qwell

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@262421 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoFix issue #17302 a slightly different way (mad props to Qwell)
Tilghman Lesher [Tue, 11 May 2010 17:22:07 +0000 (17:22 +0000)]
Fix issue #17302 a slightly different way (mad props to Qwell)

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@262321 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoAllow compilation on Mac OS X 10.4 (Tiger)
Tilghman Lesher [Mon, 10 May 2010 16:34:21 +0000 (16:34 +0000)]
Allow compilation on Mac OS X 10.4 (Tiger)

(closes issue #17297)
 Reported by: jcovert
 Patches:
       20100506__issue17297.diff.txt uploaded by tilghman (license 14)

(closes issue #17302)
 Reported by: jcovert

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@262151 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoOnly allow the operator key to be accepted after leaving a voicemail.
Jeff Peeler [Thu, 6 May 2010 20:10:59 +0000 (20:10 +0000)]
Only allow the operator key to be accepted after leaving a voicemail.

Or rather disallow the operator key from being accepted when not offered,
such as after finishing a recording from within the mailbox options menu.

ABE-2121
SWP-1267

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@261735 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoRevert 261698, code in trunk leads me to believe unadvertised options are supported.
Jeff Peeler [Thu, 6 May 2010 18:47:28 +0000 (18:47 +0000)]
Revert 261698, code in trunk leads me to believe unadvertised options are supported.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@261699 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoRemove some hidden broken code in the voicemail mailbox options menu.
Jeff Peeler [Thu, 6 May 2010 18:39:06 +0000 (18:39 +0000)]
Remove some hidden broken code in the voicemail mailbox options menu.

After finishing a recording from within the mailbox options menu, pressing 0
exhibited strange behavior with operator=yes turned on. Pressing 0 was not
even advertised as an option and the options from the vm-saveoper prompt:
"Press 1 to accept this recording. Otherwise, please continue to hold" did
not function correctly. While this of course could be fixed, it didn't really
seem to make sense even if it was working properly.

ABE-2121
SWP-1267

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@261698 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoUse the versioned MOH tarballs, now that we have them.
Jason Parker [Thu, 6 May 2010 16:56:02 +0000 (16:56 +0000)]
Use the versioned MOH tarballs, now that we have them.

This makes for more reproducibility.  Prompted by a discussion in #asterisk-dev

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@261608 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoRegistration fix for SIP realtime.
Paul Belanger [Wed, 5 May 2010 16:42:22 +0000 (16:42 +0000)]
Registration fix for SIP realtime.

Make sure realtime fields are not empty.

(closes issue #17266)
Reported by: Nick_Lewis
Patches:
      chan_sip.c-realtime.patch uploaded by Nick Lewis (license 657)
Tested by: Nick_Lewis, sberney

Review: https://reviewboard.asterisk.org/r/643/

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@261274 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoAdd a tiny corner case to the previous commit
Tilghman Lesher [Tue, 4 May 2010 23:47:08 +0000 (23:47 +0000)]
Add a tiny corner case to the previous commit

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@261094 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoProtect against overflow, when calculating how long to wait for a frame.
Tilghman Lesher [Tue, 4 May 2010 23:36:53 +0000 (23:36 +0000)]
Protect against overflow, when calculating how long to wait for a frame.

(closes issue #17128)
 Reported by: under
 Patches:
       d.diff uploaded by under (license 914)

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@261093 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoVoicemail transfer to operator should occur immediately, not after main menu.
Jeff Peeler [Tue, 4 May 2010 18:46:46 +0000 (18:46 +0000)]
Voicemail transfer to operator should occur immediately, not after main menu.

There were two scenarios in the advanced options that while using the
operator=yes and review=yes options, the transfer occurred only after exiting
the main menu (after sending a reply or leaving a message for an extension).
Now after the audio is processed for the reply or message the transfer occurs
immediately as expected.

ABE-2107
ABE-2108

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@260923 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoFix FILTER() examples to work in 1.4
Tim Ringenbach [Tue, 4 May 2010 17:40:59 +0000 (17:40 +0000)]
Fix FILTER() examples to work in 1.4

Review: https://reviewboard.asterisk.org/r/644/

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@260887 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoFix fallout from removing from configure script. Pointed out by philipp64 on #aster...
Jason Parker [Tue, 4 May 2010 15:49:27 +0000 (15:49 +0000)]
Fix fallout from removing  from configure script.  Pointed out by philipp64 on #asterisk-dev

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@260801 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoShould have removed /usr/lib/ part. Thanks Qwell.
Paul Belanger [Mon, 3 May 2010 16:54:41 +0000 (16:54 +0000)]
Should have removed /usr/lib/ part. Thanks Qwell.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@260662 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agonon-root make install PREFIX=/tmp fails.
Paul Belanger [Mon, 3 May 2010 16:41:30 +0000 (16:41 +0000)]
non-root make install PREFIX=/tmp fails.
Prepend libdir when executing mkpkgconfig allowing non-root installs to work.

(closes issue #17268)
Reported by: pabelanger
Patches:
      issue17268.patch uploaded by pabelanger (license 224)
Tested by: pabelanger

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@260661 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoMinor typo pointed out by pabelanger on IRC.
Leif Madsen [Mon, 3 May 2010 14:57:39 +0000 (14:57 +0000)]
Minor typo pointed out by pabelanger on IRC.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@260569 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoEnsure channel state is not incorrectly set in the case of a very early answer.
Jeff Peeler [Fri, 30 Apr 2010 22:22:46 +0000 (22:22 +0000)]
Ensure channel state is not incorrectly set in the case of a very early answer.

The needringing bit was being read in dahdi_read after answering thereby
setting the state to ringing from up. This clears needringing upon answering
so that is no longer possible.

(closes issue #17067)
Reported by: tzafrir
Patches:
      needringing.diff uploaded by tzafrir (license 46)

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@260434 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoFix potential crash from race condition due to accessing channel data without the...
Mark Michelson [Fri, 30 Apr 2010 20:08:15 +0000 (20:08 +0000)]
Fix potential crash from race condition due to accessing channel data without the channel locked.

In res_musiconhold.c, there are several places where a channel's
stream's existence is checked prior to calling ast_closestream on it. The issue
here is that in several cases, the channel was not locked while checking the
stream. The result was that if two threads checked the state of the channel's
stream at approximately the same time, then there could be a situation where
both threads attempt to call ast_closestream on the channel's stream. The result
here is that the refcount for the stream would go below 0, resulting in a crash.

I have added proper channel locking to res_musiconhold.c to ensure that
we do not try to check chan->stream without the channel locked. A Digium customer
has been using this patch for several weeks and has not had any crashes since
applying the patch.

ABE-2147

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@260345 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoDTMF CallerID detection problems.
Richard Mudgett [Thu, 29 Apr 2010 22:11:47 +0000 (22:11 +0000)]
DTMF CallerID detection problems.

The code handling DTMF CallerID drops digits on long CallerID numbers and
may timeout waiting for the first ring with shorter numbers.

The DTMF emulation mode was not turned off when processing DTMF CallerID.
When the emulation code gets behind in processing the DTMF digits it can
skip a digit.

For shorter numbers, the timeout may have been too short.  I increased it
from 2 seconds to 4 seconds.  Four seconds is a typical time between rings
for many countries.

(closes issue #16460)
Reported by: sum
Patches:
      issue16460.patch uploaded by rmudgett (license 664)
      issue16460_v1.6.2.patch uploaded by rmudgett (license 664)
Tested by: sum, rmudgett

Review: https://reviewboard.asterisk.org/r/634/

JIRA SWP-562
JIRA AST-334
JIRA SWP-901

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@260195 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoFixes crash in audiohook_write_list
David Vossel [Thu, 29 Apr 2010 15:31:02 +0000 (15:31 +0000)]
Fixes crash in audiohook_write_list

The middle_frame in the audiohook_write_list function was
being freed if a audiohook manipulator returned a failure.
This is incorrect logic.  This patch resolves this and
adds detailed descriptions of how this function should work
and why manipulator failures must be ignored.

(closes issue #17052)
Reported by: dvossel
Tested by: dvossel

(closes issue #16196)
Reported by: atis

Review: https://reviewboard.asterisk.org/r/623/

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@260049 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoresolves deadlocks in chan_local
David Vossel [Wed, 28 Apr 2010 21:16:03 +0000 (21:16 +0000)]
resolves deadlocks in chan_local

Issue_1.
In the local_hangup() 3 locks must be held at the same time... pvt, pvt->chan,
and pvt->owner.  Proper deadlock avoidance is done when the channel to hangup
is the outbound chan_local channel, but when it is not the outbound channel we
have an issue... We attempt to do deadlock avoidance only on the tech pvt, when
both the tech pvt and the pvt->owner are locked coming into that loop.  By
never giving up the pvt->owner channel deadlock avoidance is not entirely possible.
This patch resolves that by doing deadlock avoidance on both the pvt->owner and the pvt
when trying to get the pvt->chan lock.

Issue_2.
ast_prod() is used in ast_activate_generator() to queue a frame on the channel
and make the channel's read function get called.  This function is used in
ast_activate_generator() while the channel is locked, which mean's the channel
will have a lock both from the generator code and the frame_queue code by the
time it gets to chan_local.c's local_queue_frame code... local_queue_frame
contains some of the same crazy deadlock avoidance that local_hangup requires,
and this recursive lock prevents that deadlock avoidance from happening correctly.
This patch removes ast_prod() from the channel lock so only one lock is held during
the local_queue_frame function.

(closes issue #17185)
Reported by: schmoozecom
Patches:
      issue_17185_v1.diff uploaded by dvossel (license 671)
      issue_17185_v2.diff uploaded by dvossel (license 671)
Tested by: schmoozecom, GameGamer43

Review: https://reviewboard.asterisk.org/r/631/

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@259858 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoUpdate config.guess.
Leif Madsen [Wed, 28 Apr 2010 21:07:48 +0000 (21:07 +0000)]
Update config.guess.

Updating config.guess because after installing Ubuntu Server 9.10 and
running all the update scripts, running ./configure would not continue
because it was unable to determine what kind of system I had. After
updating config.guess things started working again.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@259852 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoAdd AC_CONFIG_AUX_DIR to configure script, so systems without install can use install...
Jason Parker [Wed, 28 Apr 2010 20:30:21 +0000 (20:30 +0000)]
Add AC_CONFIG_AUX_DIR to configure script, so systems without install can use install-sh from our source dir.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@259847 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoMissed this when removing $ID
Jason Parker [Wed, 28 Apr 2010 20:25:36 +0000 (20:25 +0000)]
Missed this when removing $ID

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@259833 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoRemove usage of `id` since it isn't useful and was causing breakge.
Jason Parker [Wed, 28 Apr 2010 19:17:38 +0000 (19:17 +0000)]
Remove usage of `id` since it isn't useful and was causing breakge.

Solaris `id` doesn't support the -u argument.  Instead of figuring out how to
fix this to work on Solaris, I decided to check why it was necessary and where
else it was used.  It was only used in one place, and it hasn't been needed
for a very long time (I question whether it was ever needed).

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@259748 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoDo not play goodbye prompt after timeout of message review.
Jeff Peeler [Wed, 28 Apr 2010 17:13:29 +0000 (17:13 +0000)]
Do not play goodbye prompt after timeout of message review.

ABE-2124

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@259664 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoDAHDI "WARNING" message is confusing and vague
Richard Mudgett [Tue, 27 Apr 2010 21:53:07 +0000 (21:53 +0000)]
DAHDI "WARNING" message is confusing and vague

"WARNING[28406]: chan_dahdi.c:6873 ss_thread: CallerID feed failed: Success"

Changed the warning to "Failed to decode CallerID on channel 'name'".  The
message before it is likely more specific about why the CallerID decode
failed.

SWP-501
AST-283

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@259531 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoUpdate sounds files.
Leif Madsen [Tue, 27 Apr 2010 21:48:47 +0000 (21:48 +0000)]
Update sounds files.

* Add additional sounds prompts for say_enumeration
* Update the English conference sounds prompts so they are better
  quality and all sound more consistent
* Clean up the core-sounds-XX.txt and extra-sounds-XX.txt files to
  include all present sound files

Both core (en, fr, es) and extra (en, fr) sounds files have been updated.

(closes issue #16200)
Reported by: murf

(closes issue #17137)
Reported by: lmadsen

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@259526 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoAdd gar to the check for AR for those silly OSes (Solaris) that don't have ar.
Jason Parker [Tue, 27 Apr 2010 21:15:46 +0000 (21:15 +0000)]
Add gar to the check for AR for those silly OSes (Solaris) that don't have ar.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@259441 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoSupport the silly OSes that don't have ar and strip.
Jason Parker [Tue, 27 Apr 2010 19:29:26 +0000 (19:29 +0000)]
Support the silly OSes that don't have ar and strip.

Since AC_PATH_TOOL is equiv to AC_CHECK_TOOL when path isn't specified, and
AC_PATH_TOOLS doesn't exist, we'll just switch to AC_CHECK_TOOLS.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@259352 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agohidecalleridname parameter in chan_dahdi.conf
Richard Mudgett [Tue, 27 Apr 2010 18:14:54 +0000 (18:14 +0000)]
hidecalleridname parameter in chan_dahdi.conf

Issue #7321 implements a new chan_dahdi configuration option.  However, a
change mentioned in the issue was never implemented.  This is the change
that will allow the feature to work.

I added a note to chan_dahdi.conf.sample about the feature.

(closes issue #17143)
Reported by: djensen99
Patches:
      diff.txt uploaded by djensen99 (license NA) (One line change)
Tested by: djensen99

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@259270 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoLet compilation succeed warning-free when DONT_OPTIMIZE is turned off.
Mark Michelson [Mon, 26 Apr 2010 21:44:43 +0000 (21:44 +0000)]
Let compilation succeed warning-free when DONT_OPTIMIZE is turned off.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@259104 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoPrevent Newchannel manager events for dummy channels.
Mark Michelson [Mon, 26 Apr 2010 21:03:08 +0000 (21:03 +0000)]
Prevent Newchannel manager events for dummy channels.

No Newchannel manager event will be fired for channels that are
allocated to not match a registered technology type. Thus bogus
channels allocated solely for variable substitution or CDR
operations do not result in a Newchannel event.

(closes issue #16957)
Reported by: atis

Review: https://reviewboard.asterisk.org/r/601

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@259018 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoWhen StopMonitor is called, ensure that it will not be restarted by a channel event.
Tilghman Lesher [Sun, 25 Apr 2010 18:09:05 +0000 (18:09 +0000)]
When StopMonitor is called, ensure that it will not be restarted by a channel event.
(closes issue #16590)
 Reported by: kkm
 Patches:
       resmonitor-16590-trunk.239289.diff uploaded by kkm (license 888)

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@258775 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoFix broken CDR behavior.
Matthew Nicholson [Thu, 22 Apr 2010 21:49:07 +0000 (21:49 +0000)]
Fix broken CDR behavior.

This change allows a CDR record previously marked with disposition ANSWERED to be set as BUSY or NO ANSWER.

Additionally this change partially reverts r235635 and does not set the AST_CDR_FLAG_ORIGINATED flag on CDRs generated from ast_call().  To preserve proper CDR behavior, the AST_CDR_FLAG_DIALED flag is now cleared from all brige CDRs in ast_bridge_call().

(closes issue #16797)
Reported by: VarnishedOtter
Tested by: mnicholson

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@258670 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoFix looping forever when no input received in certain voicemail menu scenarios.
Jeff Peeler [Wed, 21 Apr 2010 21:45:36 +0000 (21:45 +0000)]
Fix looping forever when no input received in certain voicemail menu scenarios.

Specifically, prompting for an extension (when leaving or forwarding a message)
or when prompting for a digit (when saving a message or changing folders).

ABE-2122
SWP-1268

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@258432 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoPlay correct prompt when voicemail store failure occurs after attempted forward.
Jeff Peeler [Tue, 20 Apr 2010 16:16:33 +0000 (16:16 +0000)]
Play correct prompt when voicemail store failure occurs after attempted forward.

If a user's mailbox was full and a message was attempted to be forwarded to
said box, warnings on the console would indicate failure. However, the played
prompt was that of success (vm-msgsaved). Now storage failure is taken into
account and the correct prompt (vm-mailboxfull) is played when appropriate.

ABE-2123
SWP-1262

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@258029 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agomake app_voicemail compile with IMAP_STORAGE
Jeff Peeler [Mon, 19 Apr 2010 19:09:46 +0000 (19:09 +0000)]
make app_voicemail compile with IMAP_STORAGE

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@257856 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoMake the mixmonitor thread process audio frames faster
Dwayne M. Hubbard [Fri, 16 Apr 2010 21:15:43 +0000 (21:15 +0000)]
Make the mixmonitor thread process audio frames faster

Mantis issue 17078 reports MixMonitor recordings have shorter durations than
the call duration.  This was because the mixmonitor thread was not processing
frames from the audiohook fast enough.  The mixmonitor thread would slowly fall
behind the most recent audio frame and when the channel hangs up, the mixmonitor
thread would exit without processing the same number of frames as the channel;
leaving the mixmonitor recording shorter than actual call duration.

This revision fixes this issue by moving the ast_audiohook_trigger_wait() and
the subsequent audiohook.status check into the block where the
ast_audiohook_read_frame() function returns NULL.

(closes issue #17078)
Reported by: geoff2010
Patches:
      dw-M17078.patch uploaded by dhubbard (license 733)
Tested by: dhubbard, geoff2010

Review: https://reviewboard.asterisk.org/r/611/

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@257686 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoAllow application options with arguments to contain parentheses, through a variety...
Tilghman Lesher [Thu, 15 Apr 2010 21:23:24 +0000 (21:23 +0000)]
Allow application options with arguments to contain parentheses, through a variety of escaping techniques.

Fixes SWP-1194 (ABE-2143).

Review: https://reviewboard.asterisk.org/r/604/

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@257544 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoDon't recreate peer, when responding to a repeated deregistration attempt.
Tilghman Lesher [Thu, 15 Apr 2010 20:24:50 +0000 (20:24 +0000)]
Don't recreate peer, when responding to a repeated deregistration attempt.

When a reply to a deregistration is lost in transmit, the client retries the
deregistration.  Previously, this would cause a realtime/autocreate peer to be
loaded back into memory, after it had already been correctly purged.  Instead,
we just want to resend the reply without loading the peer.

(closes issue #16908)
 Reported by: kkm
 Patches:
       20100412__issue16908.diff.txt uploaded by tilghman (license 14)
 Tested by: kkm

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@257467 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoUpdate backtrace.txt documentation.
Leif Madsen [Thu, 15 Apr 2010 19:40:33 +0000 (19:40 +0000)]
Update backtrace.txt documentation.

Update the backtrace.txt documentation so it conforms to the same layout as
other documents we've been working on recently. Additionally, add a bunch of
new information about gathering backtraces for crashes and deadlocks, along
with ways of verifying your file before uploading it. Create a couple of one
line commands for people to generate the files we need.

(closes issue #17190)
Reported by: lmadsen
Patches:
      backtrace.txt.patch-2 uploaded by lmadsen (license 10)
Tested by: lmadsen, pabelanger

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@257426 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoUpdate address of the bug tracker.
Leif Madsen [Thu, 15 Apr 2010 13:41:45 +0000 (13:41 +0000)]
Update address of the bug tracker.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@257342 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoWhen forwarding a message, ensure that prepending works correctly.
Tilghman Lesher [Wed, 14 Apr 2010 23:08:11 +0000 (23:08 +0000)]
When forwarding a message, ensure that prepending works correctly.

This is a regression in 1.4, only.

(closes issue #17103)
 Reported by: mglazer
 Patches:
       20100408__issue17103.diff.txt uploaded by tilghman (license 14)
 Tested by: tilghman

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@257266 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoAdd an option to restore past broken behavor of the Events manager action
Matthew Nicholson [Tue, 13 Apr 2010 16:46:30 +0000 (16:46 +0000)]
Add an option to restore past broken behavor of the Events manager action

Before r238915, certain values for the EventMask parameter of the Events action would result in no response being returned.  This patch adds an option to restore that broken behavior.  Also while fixing this bug I discovered that passing an empty EventMasks parameter would also result in no response being returned, this has been fixed as well while being preserved when the broken behavior is requested.

(closes issue #17023)
Reported by: nblasgen

Review: https://reviewboard.asterisk.org/r/602/

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@257070 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoAdd How-To document on collecting debugging info for issues.asterisk.org
Leif Madsen [Mon, 12 Apr 2010 17:29:26 +0000 (17:29 +0000)]
Add How-To document on collecting debugging info for issues.asterisk.org

Paul Belanger has been helping a lot with bug tracking recently and created
this document that we can now point to when additional debugging information
is required. This document will help those filing issues to know how to get
the information required when filing their issues. This will make things
easier on the developers.

Initial text and changes by pabelanger. Tweaks and editing by myself.

(closes issue #17159)
Reported by: pabelanger
Patches:
      HOWTO_collect_debug_information.txt.patch uploaded by lmadsen (license 10)
Tested by: tzafrir, pabelanger, lmadsen

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@256900 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoDAHDI/PRI call to pri_channel_bridge() not protected by PRI lock.
Richard Mudgett [Tue, 6 Apr 2010 00:10:16 +0000 (00:10 +0000)]
DAHDI/PRI call to pri_channel_bridge() not protected by PRI lock.

SWP-1231
ABE-2163

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@256225 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoResolve a deadlock that occurs due to a pointless call to ast_bridged_channel()
Russell Bryant [Fri, 2 Apr 2010 23:45:56 +0000 (23:45 +0000)]
Resolve a deadlock that occurs due to a pointless call to ast_bridged_channel()

(closes issue #16840)
Reported by: bzing2
Patches:
      patch.txt uploaded by bzing2 (license 902)
      issue_16840.rev1.diff uploaded by russell (license 2)
Tested by: bzing2, russell

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@256014 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoRemove extremely verbose debug message.
Russell Bryant [Fri, 2 Apr 2010 23:30:15 +0000 (23:30 +0000)]
Remove extremely verbose debug message.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@256009 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoEnsure line terminators in email are consistent.
Tilghman Lesher [Wed, 31 Mar 2010 19:09:46 +0000 (19:09 +0000)]
Ensure line terminators in email are consistent.

Fixes an issue with certain Mail Transport Agents, where attachments are not
interpreted correctly.

(closes issue #16557)
 Reported by: jcovert
 Patches:
       20100308__issue16557__1.4.diff.txt uploaded by tilghman (license 14)
       20100308__issue16557__1.6.0.diff.txt uploaded by tilghman (license 14)
       20100308__issue16557__trunk.diff.txt uploaded by tilghman (license 14)
 Tested by: ebroad, zktech

Reviewboard: https://reviewboard.asterisk.org/r/544/

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@255591 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoAdd documentation clarifying when 't' and 'T' can be used.
Leif Madsen [Wed, 31 Mar 2010 17:42:58 +0000 (17:42 +0000)]
Add documentation clarifying when 't' and 'T' can be used.

(closes issue #17021)
Reported by: kovzol
Tested by: lmadsen, kovzol, davidw, ebroad

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@255503 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoDon't kill Asterisk if the H323 listener does not start.
Russell Bryant [Tue, 30 Mar 2010 20:56:00 +0000 (20:56 +0000)]
Don't kill Asterisk if the H323 listener does not start.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@255409 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoDon't make Asterisk not start if pbx_dundi fails to initialize.
Russell Bryant [Tue, 30 Mar 2010 16:06:06 +0000 (16:06 +0000)]
Don't make Asterisk not start if pbx_dundi fails to initialize.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@255322 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoDon't remove local copies of utils in uninstall.
Jason Parker [Thu, 25 Mar 2010 20:41:15 +0000 (20:41 +0000)]
Don't remove local copies of utils in uninstall.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@254800 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoFix DEBUG_THREADS issue with out-of-tree modules.
Jason Parker [Thu, 25 Mar 2010 19:39:23 +0000 (19:39 +0000)]
Fix DEBUG_THREADS issue with out-of-tree modules.

Take 2, without ABI breakage this time.

Review: https://reviewboard.asterisk.org/r/588/

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@254714 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoUpdate Asterisk 1.4 to use menuselect trunk.
Russell Bryant [Thu, 25 Mar 2010 18:51:13 +0000 (18:51 +0000)]
Update Asterisk 1.4 to use menuselect trunk.

Review: https://reviewboard.asterisk.org/r/590/

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@254639 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoAdd doxygen for acl.h
Mark Michelson [Thu, 25 Mar 2010 17:33:35 +0000 (17:33 +0000)]
Add doxygen for acl.h

Review: https://reviewboard.asterisk.org/r/528

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@254552 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoSeveral fixes regarding RFC2833 DTMF detection.
Mark Michelson [Thu, 25 Mar 2010 15:59:56 +0000 (15:59 +0000)]
Several fixes regarding RFC2833 DTMF detection.

Here is a copy and paste of the details from my request on
reviewboard that dealt with these changes:

Fix 1. The first change in place is to fix Mantis issue 15811, which deals with a situation where Asterisk will incorrectly interpret out of order RFC2833 frames as duplicate DTMF digits. For instance, we would receive a sequence like:

seqno 1: DTMF 1
seqno 2: DTMF 1
seqno 3: DTMF 1
seqno 4: DTMF 1
seqno 6: DTMF 1 (end)
seqno 5: DTMF 1
seqno 7: DTMF 1 (end)
seqno 8: DTMF 1 (end)

Prior to this patch when we received the frame with seqno 5, we would interpret this as a new DTMF 1. With this patch, we will check the seqno of the incoming digit and not process the frame if the seqno is lower than the last recorded seqno. Note that we do not record the seqno of the dropped DTMF frame for future processing. While the above situation is what was designed to be fixed, the patch is written in such a way that the following would also be fixed too:

seqno  9: DTMF 1
seqno 10: DTMF 1 (end)
seqno 11: DTMF 1 (end)
seqno 13: DTMF 2
seqno 12: DTMF 1 (end)
seqno 14: DTMF 2
seqno 15: DTMF 2 (end)
seqno 16: DTMF 2 (end)
seqno 17: DTMF 2 (end)

In this second situation, the beginning of the DTMF 2 arrives before the final end frame of the DTMF 1. With the patch, seqno 12 is no processed and thus we properly interpret the DTMF.

Fix 2. The second change in place is to fix an issue like the following:

seqno 1: DTMF 1
seqno 2: DTMF 1
seqno 3: DTMF 1 (end) *packet lost*
seqno 4: DTMF 1 (end) *packet lost*
seqno 5: DTMF 1 (end) *packet lost*
seqno 6: DTMF 2

When we receive seqno 6, we had code in place that was supposed to properly end the previously unended DTMF 1. The problem was that the code was essentially a no-op. The code would set up an end frame for the DTMF 1 but would immediately overwrite the frame with the begin for DTMF 2. I changed process_dtmf_rfc2833() so that instead of returning a single frame, it is given as an output parameter a list of frames. Each frame that needs to be returned is appended to this list.

Fix 3. The final change is a minor one where an AST_CONTROL_SRCCHANGE frame could get lost. If we process a cisco DTMF or an RFC 3389 frame and no frame was returned, then we would return &ast_null_frame. The problem is that earlier in the function, we may have generated an AST_CONTROL_SRCCHANGE frame and put it in the list of frames we wish to return. This frame would be lost in such a case. The patch fixes this problem

Review: https://reviewboard.asterisk.org/r/558

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@254452 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoHandle new SRCCHANGE control message here too
Terry Wilson [Thu, 25 Mar 2010 15:57:29 +0000 (15:57 +0000)]
Handle new SRCCHANGE control message here too

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@254451 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoEnsure that monitor recordings are written to the correct location (again)
Jeff Peeler [Wed, 24 Mar 2010 00:37:23 +0000 (00:37 +0000)]
Ensure that monitor recordings are written to the correct location (again)

This is an extension to 248860. As such the dialplan test has been extended:

; non absolute path, not combined
exten => 5040, 1, monitor(wav,tmp/jeff/monitor_test)
exten => 5040, n, dial(sip/5001)
; absolute path, not combined
exten => 5041, 1, monitor(wav,/tmp/jeff/monitor_test2)
exten => 5041, n, dial(sip/5001)
; no path, not combined
exten => 5042, 1, monitor(wav,monitor_test3)
exten => 5042, n, dial(sip/5001)
; combined: changemonitor from non absolute to no path (leaves tmp/jeff)
exten => 5043, 1, monitor(wav,tmp/jeff/monitor_test4,m)
exten => 5043, n, changemonitor(monitor_test5)
exten => 5043, n, dial(sip/5001)
; combined: changemonitor from no path to non absolute path
exten => 5044, 1, monitor(wav,monitor_test6,m)
exten => 5044, n, changemonitor(tmp/jeff/monitor_test7) ; this wasn't possible before
exten => 5044, n, dial(sip/5001)
; non absolute path, combined
exten => 5045, 1, monitor(wav,tmp/jeff/monitor_test8,m)
exten => 5045, n, dial(sip/5001)
; absolute path, combined
exten => 5046, 1, monitor(wav,/tmp/jeff/monitor_test9,m)
exten => 5046, n, dial(sip/5001)
; no path, combined
exten => 5047, 1, monitor(wav,monitor_test10,m)
exten => 5047, n, dial(sip/5001)
; combined: changemonitor from non absolute to absolute (leaves tmp/jeff)
exten => 5048, 1, monitor(wav,tmp/jeff/monitor_test11,m)
exten => 5048, n, changemonitor(/tmp/jeff/monitor_test12)
exten => 5048, n, dial(sip/5001)
; combined: changemonitor from absolute to non absolute (leaves /tmp/jeff)
exten => 5049, 1, monitor(wav,/tmp/jeff/monitor_test13,m)
exten => 5049, n, changemonitor(tmp/jeff/monitor_test14)
exten => 5049, n, dial(sip/5001)
; combined: changemonitor from no path to absolute
exten => 5050, 1, monitor(wav,monitor_test15,m)
exten => 5050, n, changemonitor(/tmp/jeff/monitor_test16)
exten => 5050, n, dial(sip/5001)
; combined: changemonitor from absolute to no path (leaves /tmp/jeff)
exten => 5051, 1, monitor(wav,/tmp/jeff/monitor_test17,m)
exten => 5051, n, changemonitor(monitor_test18)
exten => 5051, n, dial(sip/5001)
; not combined: changemonitor from non absolute to no path (leaves tmp/jeff)
exten => 5052, 1, monitor(wav,tmp/jeff/monitor_test19)
exten => 5052, n, changemonitor(monitor_test20)
exten => 5052, n, dial(sip/5001)
; not combined: changemonitor from no path to non absolute
exten => 5053, 1, monitor(wav,monitor_test21)
exten => 5053, n, changemonitor(tmp/jeff/monitor_test22)
exten => 5053, n, dial(sip/5001)
; not combined: changemonitor from non absolute to absolute (leaves tmp/jeff)
exten => 5054, 1, monitor(wav,tmp/jeff/monitor_test23)
exten => 5054, n, changemonitor(/tmp/jeff/monitor_test24)
exten => 5054, n, dial(sip/5001)
; not combined: changemonitor from absolute to non absolute (leaves /tmp/jeff)
exten => 5055, 1, monitor(wav,/tmp/jeff/monitor_test24)
exten => 5055, n, changemonitor(tmp/jeff/monitor_test25)
exten => 5055, n, dial(sip/5001)
; not combined: changemonitor from no path to absolute
exten => 5056, 1, monitor(wav,monitor_test26)
exten => 5056, n, changemonitor(/tmp/jeff/monitor_test27)
exten => 5056, n, dial(sip/5001)
; not combined: changemonitor from absolute to no path (leaves /tmp/jeff)
exten => 5057, 1, monitor(wav,/tmp/jeff/monitor_test28)
exten => 5057, n, changemonitor(monitor_test29)
exten => 5057, n, dial(sip/5001)

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@254235 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoRevert revisions 254046 and 254098.
Jason Parker [Tue, 23 Mar 2010 22:45:55 +0000 (22:45 +0000)]
Revert revisions 254046 and 254098.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@254161 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoAdd note about the out-of-tree module ABI changes.
Jason Parker [Tue, 23 Mar 2010 21:27:04 +0000 (21:27 +0000)]
Add note about the out-of-tree module ABI changes.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@254098 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoAllow out-of-tree modules to load, regardless of DEBUG_THREADS/DEBUG_CHANNEL_LOCKS...
Jason Parker [Tue, 23 Mar 2010 21:07:54 +0000 (21:07 +0000)]
Allow out-of-tree modules to load, regardless of DEBUG_THREADS/DEBUG_CHANNEL_LOCKS differences.

This can be guaranteed by forcing the ABI to no longer change when these compiler flags are set.
An unfortunate side-effect to this is that there is an ABI change here.  However, there is some
mitigation.  Existing modules *will* fail to load since they would require functions that no
longer exist.

Review: https://reviewboard.asterisk.org/r/508/

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@254046 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoUnconditionally copy the caller's account code to the called party.
Matthew Nicholson [Mon, 22 Mar 2010 19:50:00 +0000 (19:50 +0000)]
Unconditionally copy the caller's account code to the called party.

(related to issue #16331)

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@253799 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoFix final link on FreeBSD by adding the PTHREAD_CFLAGS.
Russell Bryant [Sun, 21 Mar 2010 14:26:43 +0000 (14:26 +0000)]
Fix final link on FreeBSD by adding the PTHREAD_CFLAGS.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@253670 f38db490-d61c-443f-a65b-d21fe96a405b

8 years agoResolve a number of FreeBSD build issues.
Russell Bryant [Sat, 20 Mar 2010 19:17:28 +0000 (19:17 +0000)]
Resolve a number of FreeBSD build issues.

git-svn-id: http://svn.asterisk.org/svn/asterisk/branches/1.4@253631 f38db490-d61c-443f-a65b-d21fe96a405b