23 months agorefdebug: Create script.
Corey Farrell [Wed, 10 Oct 2018 09:37:23 +0000 (05:37 -0400)]
refdebug: Create script.

This allows us to process AO2 statistics for total objects, memory
usage, memory overhead and lock usage.

* Install and into the Asterisk scripts folder.
* Enable support for without DEBUG_THREADS.

Steal a bit from the ao2 magic to flag when an object lock is used.
Remove 'lockobj' from since we can now record 'used' or
'unused' for those objects.

Add comments to explain thread safety of the 'struct __priv_data'

Change-Id: I84e9d679cc86d772cc97c888d9d856a17e0d3a4a

23 months agoMerge "res_statsd.c: Fix returned reload status."
George Joseph [Mon, 15 Oct 2018 16:26:34 +0000 (11:26 -0500)]
Merge "res_statsd.c: Fix returned reload status."

23 months agomain/astfd: Fix GCC8 format-truncation warning.
Corey Farrell [Fri, 12 Oct 2018 17:21:24 +0000 (13:21 -0400)]
main/astfd: Fix GCC8 format-truncation warning.

The field used to store call arguments was not large enough to hold the
arguments string that can be constructed for 'open'.  Expand it to
prevent this warning/error.

Change-Id: I514927f256481bc84df10a51b19d5b5fb1bc387e

23 months agoMerge "loader: Flag module as declined in all cases where it fails to load."
Jenkins2 [Wed, 10 Oct 2018 17:26:58 +0000 (12:26 -0500)]
Merge "loader: Flag module as declined in all cases where it fails to load."

23 months agores_statsd.c: Fix returned reload status.
Richard Mudgett [Tue, 9 Oct 2018 21:18:49 +0000 (16:18 -0500)]
res_statsd.c: Fix returned reload status.

The return status when there was no change in statsd.conf was incorrect.
This resulted in the wrong status message on the CLI when reloading the

* Fixed cleanup on initial load if initializing statsd failed.

Change-Id: Id24fae75f1a7ff584a444a5680e867d989792481

23 months agoMerge "core/frame: generate correct T.140 payload in ast_sendtext_data()"
George Joseph [Fri, 5 Oct 2018 19:32:43 +0000 (14:32 -0500)]
Merge "core/frame: generate correct T.140 payload in ast_sendtext_data()"

23 months agoMerge "func_periodic_hook.c: Cleanup module resources on failure."
George Joseph [Fri, 5 Oct 2018 19:32:21 +0000 (14:32 -0500)]
Merge "func_periodic_hook.c: Cleanup module resources on failure."

23 months agoMerge "codec_speex.c: Cleanup module loading to DECLINE and not FAILURE."
George Joseph [Fri, 5 Oct 2018 19:30:40 +0000 (14:30 -0500)]
Merge "codec_speex.c: Cleanup module loading to DECLINE and not FAILURE."

23 months agoMerge "res_smdi.c: Fix module ref counting and inverted test."
George Joseph [Fri, 5 Oct 2018 15:52:18 +0000 (10:52 -0500)]
Merge "res_smdi.c: Fix module ref counting and inverted test."

23 months agoMerge "res_statsd.c: Made use defaults if the statsd.conf file does not exist."
George Joseph [Fri, 5 Oct 2018 15:10:28 +0000 (10:10 -0500)]
Merge "res_statsd.c: Made use defaults if the statsd.conf file does not exist."

23 months agoMerge "res_smdi.c: Made use defaults if the smdi.conf file does not exist."
George Joseph [Fri, 5 Oct 2018 15:09:50 +0000 (10:09 -0500)]
Merge "res_smdi.c: Made use defaults if the smdi.conf file does not exist."

23 months agocore/frame: generate correct T.140 payload in ast_sendtext_data()
neutrino88 [Wed, 3 Oct 2018 21:51:49 +0000 (17:51 -0400)]
core/frame: generate correct T.140 payload in ast_sendtext_data()

ast_sendtext_data() would create an incorrect T.140 text frame which
length include the null terminator byte. It causes ultimately RTP
packets to be send with this trailing 0. The proposed fix just set the
correct length to the text frame

Reported by: Emmanuel BUU
Tested by: Emmanuel BUU

Change-Id: I7ab1b9ed1e21683b2b667ea0a59d9aba3c77dd96

23 months agoloader: Flag module as declined in all cases where it fails to load.
Corey Farrell [Thu, 4 Oct 2018 23:33:25 +0000 (19:33 -0400)]
loader: Flag module as declined in all cases where it fails to load.

This has no effect on startup since AST_MODULE_LOAD_FAILURE aborts
startup, but it's possible for this code to be returned on manual load
of a module after startup.

It is an error for a module to not have a load callback but this is not
a fatal system error.  In this case flag the module as declined, return
AST_MODULE_LOAD_FAILURE only if a required module is broken.

Expand doxygen documentation for AST_MODULE_LOAD_*.

Change-Id: I3c030bb917f6e5a0dfd9d91491a4661b348cabf8

23 months agofunc_periodic_hook.c: Cleanup module resources on failure.
Richard Mudgett [Thu, 4 Oct 2018 18:13:24 +0000 (13:13 -0500)]
func_periodic_hook.c: Cleanup module resources on failure.

* Make load_module() cleanup if it failed to setup the module.

* Make unload_module() always return 0.  It is silly to fail unloading if
the hook function we try to unregister was not even registered.

Change-Id: I280fc6e8ba2a7ee2588ca01d870eebaf74b4ffe6

23 months agocodec_speex.c: Cleanup module loading to DECLINE and not FAILURE.
Richard Mudgett [Thu, 4 Oct 2018 16:49:09 +0000 (11:49 -0500)]
codec_speex.c: Cleanup module loading to DECLINE and not FAILURE.

If codec_speex fails to register a translator it would cause Asterisk to
exit instead of continue as a DECLINED module.

* Make unload_module() always return 0.  It is silly to fail unloading if
any translators we try to unregister were not even registered.

Change-Id: Ia262591f68333dad17673ba7104d11c88096f51a

23 months agoCI: Fix missing () in gates.jenkinsfile
George Joseph [Thu, 4 Oct 2018 18:03:05 +0000 (12:03 -0600)]
CI: Fix missing () in gates.jenkinsfile

Change-Id: I2f252e0f8c7f1a6328438fbd2be5d6574b7dfa5b

23 months agoMerge "CI: Add timestamps and timeouts to jenkinsfiles"
Jenkins2 [Thu, 4 Oct 2018 17:50:18 +0000 (12:50 -0500)]
Merge "CI: Add timestamps and timeouts to jenkinsfiles"

23 months agoMerge "jenkins: Fix cleanup command redirection."
Jenkins2 [Thu, 4 Oct 2018 16:28:39 +0000 (11:28 -0500)]
Merge "jenkins: Fix cleanup command redirection."

23 months agoCI: Add timestamps and timeouts to jenkinsfiles
George Joseph [Thu, 4 Oct 2018 15:13:22 +0000 (09:13 -0600)]
CI: Add timestamps and timeouts to jenkinsfiles

Change-Id: Ide83574dc957bc1df28e30a69079140050dfc35f

23 months agoMerge "ast_coredumper: Remove .gdbinit file on exit"
George Joseph [Thu, 4 Oct 2018 14:43:09 +0000 (09:43 -0500)]
Merge "ast_coredumper: Remove .gdbinit file on exit"

23 months agoMerge "CI: Use brace expansion instead of calling out to seq"
George Joseph [Thu, 4 Oct 2018 14:41:48 +0000 (09:41 -0500)]
Merge "CI: Use brace expansion instead of calling out to seq"

23 months agoMerge "CI: Look up configured kernel.core_pattern sysctl"
George Joseph [Thu, 4 Oct 2018 13:54:24 +0000 (08:54 -0500)]
Merge "CI: Look up configured kernel.core_pattern sysctl"

23 months agoMerge "ast_coredumper: Don't use "declare -n""
George Joseph [Thu, 4 Oct 2018 12:27:18 +0000 (07:27 -0500)]
Merge "ast_coredumper: Don't use "declare -n""

23 months agoMerge "astobj2: Comment on OBJ_NOLOCK in ao2_container_clone."
George Joseph [Thu, 4 Oct 2018 12:25:47 +0000 (07:25 -0500)]
Merge "astobj2: Comment on OBJ_NOLOCK in ao2_container_clone."

23 months agoMerge "http.c: Reload TLS even if http.conf hasn't changed"
George Joseph [Thu, 4 Oct 2018 12:23:49 +0000 (07:23 -0500)]
Merge "http.c: Reload TLS even if http.conf hasn't changed"

23 months agoast_coredumper: Remove .gdbinit file on exit
Sean Bright [Wed, 3 Oct 2018 22:02:27 +0000 (18:02 -0400)]
ast_coredumper: Remove .gdbinit file on exit

Change-Id: I1297de78628773ca368e687c6f148bf74857cae9

23 months agoCI: Look up configured kernel.core_pattern sysctl
Sean Bright [Wed, 3 Oct 2018 14:33:28 +0000 (10:33 -0400)]
CI: Look up configured kernel.core_pattern sysctl

Change-Id: I8246a0147df8d821fbbcabc1db1887104b8bedc4

23 months agojenkins: Fix cleanup command redirection.
Corey Farrell [Wed, 3 Oct 2018 20:51:01 +0000 (16:51 -0400)]
jenkins: Fix cleanup command redirection.

Fix redirection to /dev/null of cleanup commands.  The '2' was being
interpreted as part of the command instead of part of the redirect.

Change-Id: I2e3a591b165e0288c4b82b9ef475fdfd5392a90a

23 months agoast_coredumper: Don't use "declare -n"
George Joseph [Wed, 3 Oct 2018 20:29:31 +0000 (14:29 -0600)]
ast_coredumper: Don't use "declare -n"

Change-Id: I7ddfed4cd6549a0cd458e4d5cf9ac95d784de6cb

23 months agoMerge "CI: Use bindport instead of port in test http.conf"
George Joseph [Wed, 3 Oct 2018 18:56:54 +0000 (13:56 -0500)]
Merge "CI: Use bindport instead of port in test http.conf"

23 months agoMerge "core/frame: Fix ast_frdup() and ast_frisolate() for empty text frames"
George Joseph [Wed, 3 Oct 2018 18:55:24 +0000 (13:55 -0500)]
Merge "core/frame: Fix ast_frdup() and ast_frisolate() for empty text frames"

23 months agores_smdi.c: Fix module ref counting and inverted test.
Richard Mudgett [Tue, 2 Oct 2018 21:15:37 +0000 (16:15 -0500)]
res_smdi.c: Fix module ref counting and inverted test.

I think this module is so screwed up that it doesn't work anymore.  Even
with these attempts to fix things it still won't gracefully shut down.
The module refs will not go to zero to allow unloading the module.

* Fix module ref counting dealing with the SMDI interface object.  There
were several off-nominal paths that unbalanced the module ref count.  Also
the destructor freed the ao2 object itself which is bad.  Made the
smdi_read thread not hold its own ref to the SMDI interface object so when
all refs go away the destructor will stop the listener thread.

* Fixed the smdi_load() return code of 1 concerning the number of
listeners.  The test was inverted.

Change-Id: Ic288db51b58e395d6a2fc3847f77176c16988784

23 months agores_smdi.c: Made use defaults if the smdi.conf file does not exist.
Richard Mudgett [Tue, 2 Oct 2018 21:23:31 +0000 (16:23 -0500)]
res_smdi.c: Made use defaults if the smdi.conf file does not exist.

This module is an optional dependency of a couple of other modules.  If it
declines to load, it then forces other modules that can optionally use
this module to also decline.

* Made use the default configuration if the config file does not exist and
simplified some of the logic.

Change-Id: Ib93191f1fe28c0dd9ebe3d84c7762b32f83c4eb9

23 months agoastobj2: Comment on OBJ_NOLOCK in ao2_container_clone.
Corey Farrell [Tue, 2 Oct 2018 22:15:47 +0000 (18:15 -0400)]
astobj2: Comment on OBJ_NOLOCK in ao2_container_clone.

The test for OBJ_NOLOCK looks wrong but it isn't.  Add comments to
prevent confusion.

Change-Id: I9662b82eb39e7627a1f1944fd18f967a2b987344

23 months agoMerge "astobj2: Record lock usage to refs log when DEBUG_THREADS is enabled."
George Joseph [Wed, 3 Oct 2018 15:17:27 +0000 (10:17 -0500)]
Merge "astobj2: Record lock usage to refs log when DEBUG_THREADS is enabled."

23 months agoMerge "loader: Fix result of module reload error."
George Joseph [Wed, 3 Oct 2018 15:17:07 +0000 (10:17 -0500)]
Merge "loader: Fix result of module reload error."

23 months agoMerge "core: Disable astobj2 locking for some common objects."
George Joseph [Wed, 3 Oct 2018 15:16:20 +0000 (10:16 -0500)]
Merge "core: Disable astobj2 locking for some common objects."

23 months agoCI: Use brace expansion instead of calling out to seq
Sean Bright [Wed, 3 Oct 2018 14:05:55 +0000 (10:05 -0400)]
CI: Use brace expansion instead of calling out to seq

Also make the shebang in the first line.

Change-Id: I3fdd6f22e652e4fb5b5fe85df46fa34eb6d0cf08

23 months agoCI: Use bindport instead of port in test http.conf
Sean Bright [Wed, 3 Oct 2018 13:59:52 +0000 (09:59 -0400)]
CI: Use bindport instead of port in test http.conf

Change-Id: Ife9a6879da63a56e5b8348a2024eeed4e7b1615b

23 months agohttp.c: Reload TLS even if http.conf hasn't changed
Sean Bright [Wed, 3 Oct 2018 12:56:34 +0000 (08:56 -0400)]
http.c: Reload TLS even if http.conf hasn't changed

There is currently no way to indicate to Asterisk that TLS certificates
and/or keys have been updated other than by modifying http.conf or
restarting Asterisk.

There is already code in main/tcptls.c that determines if a reload is
actually necessary based on the hashes of the certicate and dependent
files, so this change merely gives us a way to request a reload without
explicitly modifying http.conf.

Change-Id: Ie795420dcc7eb3d91336820688a29adbcc321276

23 months agoMerge "Resolve warning about duplicate 'dialplan' CLI."
Joshua Colp [Wed, 3 Oct 2018 13:31:20 +0000 (08:31 -0500)]
Merge "Resolve warning about duplicate 'dialplan' CLI."

23 months agores_statsd.c: Made use defaults if the statsd.conf file does not exist.
Richard Mudgett [Tue, 2 Oct 2018 18:29:59 +0000 (13:29 -0500)]
res_statsd.c: Made use defaults if the statsd.conf file does not exist.

This module is an optional dependency of many modules.  If it declines to
load it then forces other modules that can optionally use this module to
also decline.

* Made use default configuration if there is a config error or the config
file does not exist.

Change-Id: If1068a582ec54ab7fb437265cb5370a97a825737

23 months agocore: Disable astobj2 locking for some common objects.
Corey Farrell [Tue, 2 Oct 2018 03:12:14 +0000 (23:12 -0400)]
core: Disable astobj2 locking for some common objects.

* ACO options
* Indications
* Module loader ref_debug object
* Media index info and variants
* xmldoc items

These allocation locations were identified using on the
master branch.

Change-Id: Ie999b9941760be3d1946cdb6e30cb85fd97504d8

23 months agoMerge "loader: Improve error handling."
George Joseph [Tue, 2 Oct 2018 19:24:09 +0000 (14:24 -0500)]
Merge "loader: Improve error handling."

23 months agoResolve warning about duplicate 'dialplan' CLI.
Corey Farrell [Thu, 13 Sep 2018 18:03:15 +0000 (14:03 -0400)]
Resolve warning about duplicate 'dialplan' CLI.

Change-Id: I029db1b4a32ccfb38374d6fe944dc430866f4b30

23 months agoloader: Fix result of module reload error.
Corey Farrell [Tue, 2 Oct 2018 06:33:44 +0000 (02:33 -0400)]
loader: Fix result of module reload error.

When a module reload fails we never set AST_MODULE_RELOAD_ERROR.  This
caused reload failures to incorrectly report 'No module found'.

Change-Id: I5f3953e0f7d135e53ec797f24c97ee3f73f232e7

23 months agoloader: Improve error handling.
Corey Farrell [Fri, 28 Sep 2018 15:13:39 +0000 (11:13 -0400)]
loader: Improve error handling.

* Display list of unavailable dependencies when they cause another
  module to fail loading.
* When a module declines to load find all modules which depend on it so
  they can be declined and listed together.
* Prevent retry of declined modules during startup.
* When a module fails to dlopen try loading it with RTLD_LAZY so we can
  attempt to display the list of missing dependencies.

These changes are meant to reduce logger spam that is caused when a
module has many dependencies and declines to load.  This also fixes some
error paths which failed to recognize required modules.

Module load/start errors are delayed until the end of loader startup.

Change-Id: I046052c71331c556c09d39f47a3b92975f3e1758

23 months agocore/frame: Fix ast_frdup() and ast_frisolate() for empty text frames
neutrino88 [Tue, 25 Sep 2018 21:19:36 +0000 (17:19 -0400)]
core/frame: Fix ast_frdup() and ast_frisolate() for empty text frames

If a channel creates an AST_TEXT_FRAME with datalen == 0, the ast_frdup()
and ast_frisolate() functions could create a clone frame with an invalid
data.ptr which would cause a crash.  The proposed fix is to make sure that
for such empty text frames, ast_frdup() and ast_frisolate() return cloned
text frames with a valid data.ptr.

Reported by: Emmanuel BUU
Tested by: Emmanuel BUU

Change-Id: Ib882dd028598f13c4c233edbfdd7e54ad44a68e9

23 months agoMerge "Append CHANGES/UPGRADE.txt for module loader changes."
Joshua Colp [Tue, 2 Oct 2018 13:47:48 +0000 (08:47 -0500)]
Merge "Append CHANGES/UPGRADE.txt for module loader changes."

23 months agoMerge "app_queue.c: Fix json ref leak"
Joshua Colp [Tue, 2 Oct 2018 12:58:26 +0000 (07:58 -0500)]
Merge "app_queue.c: Fix json ref leak"

23 months agoastobj2: Record lock usage to refs log when DEBUG_THREADS is enabled.
Corey Farrell [Mon, 1 Oct 2018 04:11:44 +0000 (00:11 -0400)]
astobj2: Record lock usage to refs log when DEBUG_THREADS is enabled.

When DEBUG_THREADS is enabled we can know if the astobj2 mutex / rwlock
was ever used, so it can be recorded in the REF_DEBUG destructor entry.

Create contrib/scripts/ to process locking used by
allocator.  This can be used to identify places where
AO2_ALLOC_OPT_LOCK_NOLOCK should be used to reduce memory usage.

Change-Id: I2e3cd23336a97df2692b545f548fd79b14b53bf4

23 months agoMerge "app_page: Add dependency against app_confbridge."
George Joseph [Mon, 1 Oct 2018 19:19:48 +0000 (14:19 -0500)]
Merge "app_page: Add dependency against app_confbridge."

23 months agoapp_page: Add dependency against app_confbridge.
Corey Farrell [Mon, 1 Oct 2018 17:11:41 +0000 (13:11 -0400)]
app_page: Add dependency against app_confbridge.

Change-Id: I1946509f518961d23fb21229d91676ee3e441921

23 months agoapp_queue.c: Fix json ref leak
Richard Mudgett [Fri, 28 Sep 2018 18:55:43 +0000 (13:55 -0500)]
app_queue.c: Fix json ref leak

Declining the queue_member_status_type stasis message in stasis.conf
causes these messages to leak json objects.

* Add missing ast_json_unref() if the type is NULL in


Change-Id: I691ecf49bd1f7d9c29182e1eee8c4bb7103be9fc

23 months agoMerge "astobj2: Reduce memory overhead."
Jenkins2 [Mon, 1 Oct 2018 14:02:37 +0000 (09:02 -0500)]
Merge "astobj2: Reduce memory overhead."

23 months agoMerge "lock: Improve performance of DEBUG_THREADS."
Joshua Colp [Mon, 1 Oct 2018 13:32:49 +0000 (08:32 -0500)]
Merge "lock: Improve performance of DEBUG_THREADS."

23 months agoMerge " Check for unbound version >= 1.5"
Joshua Colp [Mon, 1 Oct 2018 12:08:52 +0000 (07:08 -0500)]
Merge "  Check for unbound version >= 1.5"

23 months agoMerge "res_pjsip: improve realtime performance on CLI 'pjsip show contacts'"
Joshua Colp [Mon, 1 Oct 2018 11:51:37 +0000 (06:51 -0500)]
Merge "res_pjsip: improve realtime performance on CLI 'pjsip show contacts'"

23 months agoMerge "jansson-bundled: Add patches to improve json_pack error reporting."
Joshua Colp [Mon, 1 Oct 2018 11:24:46 +0000 (06:24 -0500)]
Merge "jansson-bundled: Add patches to improve json_pack error reporting."

23 months agoMerge "app_confbridge: Use bridge join hook to send join and leave events"
Joshua Colp [Mon, 1 Oct 2018 11:24:21 +0000 (06:24 -0500)]
Merge "app_confbridge:  Use bridge join hook to send join and leave events"

23 months agoMerge "res_stasis: Fix stale data in ARI bridges"
Joshua Colp [Mon, 1 Oct 2018 09:34:30 +0000 (04:34 -0500)]
Merge "res_stasis: Fix stale data in ARI bridges"

23 months agoAppend CHANGES/UPGRADE.txt for module loader changes.
Corey Farrell [Mon, 1 Oct 2018 08:24:00 +0000 (04:24 -0400)]
Append CHANGES/UPGRADE.txt for module loader changes.

Change-Id: Ib8db4e14187f5c11ecbff532df17d30c5d36fa3e

23 months agores_pjsip: improve realtime performance on CLI 'pjsip show contacts'
Alexei Gradinari [Tue, 25 Sep 2018 22:33:32 +0000 (18:33 -0400)]
res_pjsip: improve realtime performance on CLI 'pjsip show contacts'

CLI command 'pjsip show contacts' inefficiently make a lot of DB requests.

For example if there are 10k aors then asterisk requests these 10k records
of aor and then does 10k requests of contact - one request per aor.

Even if use 'like <pattern>' the asterisk requests all aor's and contact's
records and then filters them by itself.

This patch gathers contact's container by
- retrieving all dynamic contacts by regex (filtered by reg_server)
- retrieving all aors with permanent contacts
- finally filters container by regex

ASTERISK-28077 #close

Change-Id: Id0ad65d14952a02fb213273a90f3f680a8149618

23 months agojansson-bundled: Add patches to improve json_pack error reporting.
Corey Farrell [Fri, 28 Sep 2018 19:45:36 +0000 (15:45 -0400)]
jansson-bundled: Add patches to improve json_pack error reporting.

Change-Id: I045e420d5e73e60639079246e810da6ae21ae22b

23 months agolock: Improve performance of DEBUG_THREADS.
Corey Farrell [Fri, 28 Sep 2018 00:32:21 +0000 (20:32 -0400)]
lock: Improve performance of DEBUG_THREADS.

Add a volatile flag to lock tracking structures so we only need to use
the global lock when first initializing tracking.

Additionally add support for DEBUG_THREADS_LOOSE_ABI.  This is used by
astobj2.c to eliminate storage for tracking fields when DEBUG_THREADS is
not defined.

Change-Id: Iabd650908901843e9fff47ef1c539f0e1b8cb13b

23 months agoMerge "config.c: Cleanup AST_INCLUDE_GLOB"
George Joseph [Fri, 28 Sep 2018 18:16:17 +0000 (13:16 -0500)]
Merge "config.c: Cleanup AST_INCLUDE_GLOB"

23 months agoMerge "res_odbc: fix missing SQL error diagnostic"
Kevin Harwell [Fri, 28 Sep 2018 15:39:21 +0000 (10:39 -0500)]
Merge "res_odbc: fix missing SQL error diagnostic"

23 months agoMerge "astobj2: Fix shutdown order."
George Joseph [Fri, 28 Sep 2018 13:38:30 +0000 (08:38 -0500)]
Merge "astobj2: Fix shutdown order."

23 months agoMerge "app_queue: Fix Attended transfer hangup with removing pending member."
George Joseph [Fri, 28 Sep 2018 12:49:16 +0000 (07:49 -0500)]
Merge "app_queue: Fix Attended transfer hangup with removing pending member."

23 months agoapp_confbridge: Use bridge join hook to send join and leave events
George Joseph [Thu, 27 Sep 2018 18:19:28 +0000 (12:19 -0600)]
app_confbridge:  Use bridge join hook to send join and leave events

The first attempt at publishing confbridge events to participants
involved publishing them at the same time stasis events were
created.  This caused issues with bridge and channel locks.  The
second attempt involved publishing them when the stasis events
were received by the code that published the confbridge AMI events.
This caused timing issues because, depending on resources available,
the event could be received before channels actually joined the
bridge and would therefore fail to send messages to the participant.

This attempt reverts to the original mechanism with one exception.
The join and leave events are published via bridge join and leave
hooks.  This guarantees the states of the channels and bridge and
provides deterministic timing for event publishing.

Change-Id: I2660074f8a30a5224cb953d5e047ee84484a9036

23 months agoMerge "res_rtp_asterisk.c: Add "seqno" strictrtp option"
George Joseph [Fri, 28 Sep 2018 12:27:24 +0000 (07:27 -0500)]
Merge "res_rtp_asterisk.c: Add "seqno" strictrtp option"

23 months agoastobj2: Reduce memory overhead.
Corey Farrell [Thu, 27 Sep 2018 09:51:43 +0000 (05:51 -0400)]
astobj2: Reduce memory overhead.

Reduce options to 2-bit field, magic to 30 bit field.  Move ref_counter
next to options and explicitly use int32_t so the fields will pack.

This reduces memory overhead for every ao2 object by 8 bytes on x86_64.

Change-Id: Idc1baabb35ec3b3d8de463c4fa3011eaf7fcafb5

23 months agoconfig.c: Cleanup AST_INCLUDE_GLOB
Sean Bright [Thu, 27 Sep 2018 20:01:58 +0000 (16:01 -0400)]
config.c: Cleanup AST_INCLUDE_GLOB

* In main/config.c, AST_INCLUDE_GLOB is fixed to '1' making the #ifdefs

* In utils/extconf.c, AST_INCLUDE_GLOB is never defined so there is a
  lot of dead code.

Change-Id: I1bad1a46d7466ddf90d52cc724e997195495226c

23 months agoMerge "res_rtp_asterisk: Raise event when RTP port is allocated"
George Joseph [Thu, 27 Sep 2018 14:20:24 +0000 (09:20 -0500)]
Merge "res_rtp_asterisk: Raise event when RTP port is allocated"

23 months agoMerge "CI: Add --test-timeout option to"
Joshua Colp [Thu, 27 Sep 2018 11:22:40 +0000 (06:22 -0500)]
Merge "CI:  Add --test-timeout option to"

23 months agoastobj2: Fix shutdown order.
Corey Farrell [Thu, 27 Sep 2018 10:33:22 +0000 (06:33 -0400)]
astobj2: Fix shutdown order.

When REF_DEBUG and AO2_DEBUG are both enabled we closed the refs log
before we shutdown astobj2_container.  This caused the AO2_DEBUG
container registration container to be reported as a leak.

Change-Id: If9111c4c21c68064b22c546d5d7a41fac430430e

23 months agoapp_queue: Fix Attended transfer hangup with removing pending member.
Cao Minh Hiep [Thu, 6 Sep 2018 02:14:12 +0000 (11:14 +0900)]
app_queue: Fix Attended transfer hangup with removing pending member.

This issue related to setting of holdtime, announcements, member delays.
It works well if we set the member delays to "0" and no announcements
and no holdtime.This issue will happen if we set member delays to "1",
"2"... or announcements or holdtime and hangs up the call during
processing it.

And here is the reason:
(At the step of answering a phone.)
It takes care any holdtime, announcements, member delays,
or other options after a call has been answered if it exists.

Normally, After the call has been aswered,
and we wait for the processing one of the cases of the member delays
or hold time or announcements finished, "if (ast_check_hangup(peer))"
will be not executed, then queue will be updated at update_queue().
Here, pending member will be removed.

However, after the call has been aswered,
if we hangs up the call during one of the cases of the member delays
or hold time or announcements, "if (ast_check_hangup(peer))"
will be executed.
outgoing = NULL and at hangupcalls, pending members will not be removed.

* This fixed patch will remove the pending member from container
before hanging up the call with outgoing is NULL.


Reported by: Cao Minh Hiep
Tested by: Cao Minh Hiep

Change-Id: Ib780fbf48ace9d2d8eaa1270b9d530a4fc14c855

23 months agores_stasis: Fix stale data in ARI bridges
Moritz Fain [Tue, 26 Jun 2018 14:17:37 +0000 (16:17 +0200)]
res_stasis: Fix stale data in ARI bridges

Fixed an issue that resulted in "Allocation failed" each time an ARI
request was made to start playing MOH on a bridge.

In bridge_moh_create() we were attaching the after bridge callbacks to
chan which is the ;1 channel of the unreal channel pair.  We should have
attached them to the ;2 channel which is pushed into the bridge by
ast_unreal_channel_push_to_bridge().  The callbacks are called when the
specific channel leaves the bridging system.  Since the ;1 channel is
never put into a bridge the callbacks never get called.  The callbacks
then never remove the moh_wrapper from the app_bridges_moh container.  As
a result we cannot find the channel associated with the wrapper to start
MOH because it has hungup.  This is the reason causing the reported issue.

* Rather than using after bridge callbacks to cleanup, we now have
moh_channel_thread() doing the cleanup when the channel hangs up.

* Fixed moh_channel_thread() accumulating control frames on the stasis
bridge MOH channel until MOH is stopped.  Control frames are no longer
accumulated while MOH is playing.

* Fixed channel ref counting issue.  stasis_app_bridge_moh_channel() may
or may not return a channel ref.  As a result ast_ari_bridges_start_moh()
wouldn't know it may have a channel ref to release.
stasis_app_bridge_moh_channel() will now return a ref with the channel it

* Eliminated RAII_VAR in bridge_moh_create().

ASTERISK-26094 #close

Change-Id: Ibff479e167b3320c68aaabfada7e1d0ef7bd548c

23 months agores_rtp_asterisk.c: Add "seqno" strictrtp option
Ben Ford [Mon, 10 Sep 2018 16:28:09 +0000 (11:28 -0500)]
res_rtp_asterisk.c: Add "seqno" strictrtp option

When networks experience disruptions, there can be large gaps of time
between receiving packets. When strictrtp is enabled, this created
issues where a flood of packets could come in and be seen as an attack.
Another option - seqno - has been added to the strictrtp option that
ignores the time interval and goes strictly by sequence number for

Change-Id: I8a42b8d193673899c8fc22fe7f98ea87df89be71

23 months agoMerge "jansson: Backport fixes to bundled, use json_vsprintf if available."
George Joseph [Wed, 26 Sep 2018 16:09:50 +0000 (11:09 -0500)]
Merge "jansson: Backport fixes to bundled, use json_vsprintf if available."

23 months agoMerge "chan_sip: SipNotify on Chan_Sip vi AMI behave different to CLI"
George Joseph [Wed, 26 Sep 2018 14:34:10 +0000 (09:34 -0500)]
Merge "chan_sip: SipNotify on Chan_Sip vi AMI behave different to CLI"

23 months agores_odbc: fix missing SQL error diagnostic
Alexei Gradinari [Thu, 20 Sep 2018 18:59:54 +0000 (14:59 -0400)]
res_odbc: fix missing SQL error diagnostic

On SQL error there is not diagnostic information about this error.
There is only
WARNING res_odbc.c: SQL Execute error -1!

The function ast_odbc_print_errors calls a SQLGetDiagField to get the number
of available diagnostic records, but the SQLGetDiagField returns 0.
However SQLGetDiagRec could return one diagnostic records in this case.

Looking at many example of getting diagnostics error information
I found out that the best way it's to use only SQLGetDiagRec
while it returns SQL_SUCCESS.

Also this patch adds calls of ast_odbc_print_errors on SQL_ERROR
to res_config_odbc.

ASTERISK-28065 #close

Change-Id: Iba5ae5470ac49ecd911dd084effbe9efac68ccc1

23 months agoCI: Add --test-timeout option to
George Joseph [Wed, 26 Sep 2018 13:12:28 +0000 (07:12 -0600)]
CI:  Add --test-timeout option to

The default is 600 seconds.
Also added timeouts to the *TestGroups.json files.

Change-Id: I8ab6a69e704b6a10f06a0e52ede02312a2b72fe0

23 months agoMerge "rtp_engine: rtcp_report_to_json can overflow the ssrc integer value"
George Joseph [Wed, 26 Sep 2018 13:02:28 +0000 (08:02 -0500)]
Merge "rtp_engine: rtcp_report_to_json can overflow the ssrc integer value"

23 months agochan_sip: SipNotify on Chan_Sip vi AMI behave different to CLI
pk16208 [Tue, 18 Sep 2018 13:01:02 +0000 (15:01 +0200)]
chan_sip: SipNotify on Chan_Sip vi AMI behave different to CLI

With tls and udp enabled asterisk generates a warning about sending
message via udp instead of tls.
sip notify command via cli works as expected and without warning.

asterisk has to set the connection information accordingly to connection
and not on presumption

ASTERISK-28057 #close

Change-Id: Ib43315aba1f2c14ba077b52d8c5b00be0006656e

23 months Check for unbound version >= 1.5
George Joseph [Mon, 24 Sep 2018 22:56:07 +0000 (16:56 -0600)]  Check for unbound version >= 1.5

In order to do this and provide good feedback, a new macro was
created (AST_EXT_LIB_EXTRA_CHECK) which does the normal check and
path setups for the library then compiles, links and runs a supplied
code fragment to do the final determination.  In this case, the
final code fragment compares UNBOUND_VERSION_MAJOR
and UNBOUND_VERSION_MINOR to determine if they're greater than or
equal to 1.5.

Since we require version 1.5, some code in res_resolver_unbound
was also simplified.

Reported by: Samuel Galarneau

Change-Id: Iee94ad543cd6f8b118df8c4c7afd9c4e2ca1fa72

23 months agores_rtp_asterisk: Raise event when RTP port is allocated
Joshua Colp [Mon, 24 Sep 2018 17:43:17 +0000 (17:43 +0000)]
res_rtp_asterisk: Raise event when RTP port is allocated

This change raises a testsuite event to provide what port
Asterisk has actually allocated for RTP. This ensures that
testsuite tests can remove any assumption of ports and instead
use the actual port in use.


Change-Id: I91bd45782e84284e01c89acf4b2da352e14ae044

23 months agojansson: Backport fixes to bundled, use json_vsprintf if available.
Corey Farrell [Tue, 17 Jul 2018 03:55:02 +0000 (23:55 -0400)]
jansson: Backport fixes to bundled, use json_vsprintf if available.

Use json_vsprintf from versions which contain fix for va_copy leak.

Apply fixes from jansson master:
* va_copy leak fix.
* Avoid potential invalid memory read in json_pack.
* Rename variable that shadowed another.

Change-Id: I7522e462d2a52f53010ffa1e7d705c666ec35539

23 months agojson: Take advantage of new API's.
Corey Farrell [Tue, 17 Jul 2018 03:55:02 +0000 (23:55 -0400)]
json: Take advantage of new API's.

* Use "o*" format specifier for optional fields in ast_json_party_id.
* Stop using ast_json_deep_copy on immutable objects, it is now thread
  safe to just use ast_json_ref.

Additional changes to ast_json_pack calls in the vicinity:
* Use "O" when an object needs to be bumped.  This was previously
  avoided as it was not thread safe.
* Use "o?" and "O?" to replace NULL with ast_json_null().  The
  "?" is a new feature of ast_json_pack starting with Asterisk 16.

Change-Id: I8382d28d7d83ee0ce13334e51ae45dbc0bdaef48

23 months agoMerge "app_voicemail: Fix stack overrun in append_mailbox"
George Joseph [Mon, 24 Sep 2018 18:50:36 +0000 (13:50 -0500)]
Merge "app_voicemail:  Fix stack overrun in append_mailbox"

23 months agoMerge "app_voicemail: Cleanup mailbox topic and cache"
George Joseph [Mon, 24 Sep 2018 16:46:50 +0000 (11:46 -0500)]
Merge "app_voicemail:  Cleanup mailbox topic and cache"

23 months agoMerge "chan_sip.c: chan_sip unstable with TLS after asterisk start or reloads"
George Joseph [Mon, 24 Sep 2018 15:44:51 +0000 (10:44 -0500)]
Merge "chan_sip.c: chan_sip unstable with TLS after asterisk start or reloads"

23 months agoMerge "res_remb_modifier: Add module for controlling REMB from CLI."
George Joseph [Mon, 24 Sep 2018 15:11:54 +0000 (10:11 -0500)]
Merge "res_remb_modifier: Add module for controlling REMB from CLI."

23 months agoapp_voicemail: Cleanup mailbox topic and cache
George Joseph [Thu, 20 Sep 2018 15:15:48 +0000 (09:15 -0600)]
app_voicemail:  Cleanup mailbox topic and cache

app_voicemail wasn't properly cleaning up the stasis cache or the
mwi topic pool when the module was unloaded or when a user was
deleted as a result of a reload.  This resulted in leaks in both

* app_voicemail now calls ast_delete_mwi_state_full when it frees
  a user structure and ast_delete_mwi_state_full in turn now calls
  the new stasis_topic_pool_delete_topic function to clear the topic
  from the pool.

Change-Id: Ide23144a4a810e7e0faad5a8e988d15947965df8

23 months agoMerge "stasis: Add function to delete topic from pool"
George Joseph [Mon, 24 Sep 2018 14:29:14 +0000 (09:29 -0500)]
Merge "stasis:  Add function to delete topic from pool"

23 months agoMerge "res_rtp_asterisk: Fix crash on ast_rtp_new failure."
George Joseph [Mon, 24 Sep 2018 14:27:01 +0000 (09:27 -0500)]
Merge "res_rtp_asterisk: Fix crash on ast_rtp_new failure."

23 months agoMerge "channel.c: Address stack overflow in does_id_conflict()"
George Joseph [Mon, 24 Sep 2018 14:23:10 +0000 (09:23 -0500)]
Merge "channel.c:  Address stack overflow in does_id_conflict()"

2 years agortp_engine: rtcp_report_to_json can overflow the ssrc integer value
Kevin Harwell [Mon, 17 Sep 2018 20:35:05 +0000 (15:35 -0500)]
rtp_engine: rtcp_report_to_json can overflow the ssrc integer value

When writing an RTCP report to json the code attempts to pack the "ssrc" and
"source_ssrc" unsigned integer values as a signed int value type. This of course
means if the ssrc's unsigned value is greater than that which can fit into a
signed integer value it gets converted to a negative number. Subsequently, the
negative value goes out in the json report.

This patch now packs the value as a json_int_t, which is the widest integer type
available on a given system. This should make it so the value no longer

Note, this was caught by two failing tests hep/rtcp-receiver/ and

Change-Id: I2af275286ee5e795b79f0c3d450d9e4b28e958b0

2 years agoapp_voicemail: Fix stack overrun in append_mailbox
George Joseph [Fri, 21 Sep 2018 19:32:52 +0000 (13:32 -0600)]
app_voicemail:  Fix stack overrun in append_mailbox

The append_mailbox function wasn't calculating the correct length
to pass to ast_alloca and it wasn't handling the case where context
might be empty.

Found by the Address Sanitizer.

Change-Id: I7eb51c7bd18a7a8dbdba261462a95cc69e84f161

2 years agochannel.c: Address stack overflow in does_id_conflict()
George Joseph [Fri, 21 Sep 2018 20:23:34 +0000 (14:23 -0600)]
channel.c:  Address stack overflow in does_id_conflict()

does_id_conflict() was passing a pointer to an int to a callback
that expected a pointer to a size_t.

Found by the Address Sanitizer.

Change-Id: I0ff542067eef63a14a60301654d65d34fe2ad503