asterisk/apps/app_confbridge.c

4687 lines
142 KiB
C
Raw Permalink Normal View History

/*
* Asterisk -- An open source telephony toolkit.
*
* Copyright (C) 2007-2008, Digium, Inc.
*
* Joshua Colp <jcolp@digium.com>
* David Vossel <dvossel@digium.com>
*
* See http://www.asterisk.org for more information about
* the Asterisk project. Please do not directly contact
* any of the maintainers of this project for assistance;
* the project provides a web site, mailing lists and IRC
* channels for your use.
*
* This program is free software, distributed under the terms of
* the GNU General Public License Version 2. See the LICENSE file
* at the top of the source tree.
*/
/*! \file
*
* \brief Conference Bridge application
*
* \author\verbatim Joshua Colp <jcolp@digium.com> \endverbatim
* \author\verbatim David Vossel <dvossel@digium.com> \endverbatim
*
* This is a conference bridge application utilizing the bridging core.
* \ingroup applications
*/
/*! \li \ref app_confbridge.c uses the configuration file \ref confbridge.conf
* \addtogroup configuration_file Configuration Files
*/
/*!
* \page confbridge.conf confbridge.conf
* \verbinclude confbridge.conf.sample
*/
/*** MODULEINFO
<support_level>core</support_level>
***/
#include "asterisk.h"
#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#include <string.h>
#include <signal.h>
#include "asterisk/cli.h"
#include "asterisk/file.h"
#include "asterisk/channel.h"
#include "asterisk/pbx.h"
#include "asterisk/pbx.h"
#include "asterisk/module.h"
#include "asterisk/lock.h"
#include "asterisk/bridge.h"
#include "asterisk/musiconhold.h"
#include "asterisk/say.h"
#include "asterisk/audiohook.h"
#include "asterisk/astobj2.h"
#include "confbridge/include/confbridge.h"
#include "asterisk/paths.h"
#include "asterisk/manager.h"
#include "asterisk/test.h"
#include "asterisk/stasis.h"
#include "asterisk/stasis_bridges.h"
#include "asterisk/stasis_channels.h"
#include "asterisk/json.h"
media formats: re-architect handling of media for performance improvements In the old times media formats were represented using a bit field. This was fast but had a few limitations. 1. Asterisk was limited in how many formats it could handle. 2. Formats, being a bit field, could not include any attribute information. A format was strictly its type, e.g., "this is ulaw". This was changed in Asterisk 10 (see https://wiki.asterisk.org/wiki/display/AST/Media+Architecture+Proposal for notes on that work) which led to the creation of the ast_format structure. This structure allowed Asterisk to handle attributes and bundle information with a format. Additionally, ast_format_cap was created to act as a container for multiple formats that, together, formed the capability of some entity. Another mechanism was added to allow logic to be registered which performed format attribute negotiation. Everywhere throughout the codebase Asterisk was changed to use this strategy. Unfortunately, in software, there is no free lunch. These new capabilities came at a cost. Performance analysis and profiling showed that we spend an inordinate amount of time comparing, copying, and generally manipulating formats and their related structures. Basic prototyping has shown that a reasonably large performance improvement could be made in this area. This patch is the result of that project, which overhauled the media format architecture and its usage in Asterisk to improve performance. Generally, the new philosophy for handling formats is as follows: * The ast_format structure is reference counted. This removed a large amount of the memory allocations and copying that was done in prior versions. * In order to prevent race conditions while keeping things performant, the ast_format structure is immutable by convention and lock-free. Violate this tenet at your peril! * Because formats are reference counted, codecs are also reference counted. The Asterisk core generally provides built-in codecs and caches the ast_format structures created to represent them. Generally, to prevent inordinate amounts of module reference bumping, codecs and formats can be added at run-time but cannot be removed. * All compatibility with the bit field representation of codecs/formats has been moved to a compatibility API. The primary user of this representation is chan_iax2, which must continue to maintain its bit-field usage of formats for interoperability concerns. * When a format is negotiated with attributes, or when a format cannot be represented by one of the cached formats, a new format object is created or cloned from an existing format. That format may have the same codec underlying it, but is a different format than a version of the format with different attributes or without attributes. * While formats are reference counted objects, the reference count maintained on the format should be manipulated with care. Formats are generally cached and will persist for the lifetime of Asterisk and do not explicitly need to have their lifetime modified. An exception to this is when the user of a format does not know where the format came from *and* the user may outlive the provider of the format. This occurs, for example, when a format is read from a channel: the channel may have a format with attributes (hence, non-cached) and the user of the format may last longer than the channel (if the reference to the channel is released prior to the format's reference). For more information on this work, see the API design notes: https://wiki.asterisk.org/wiki/display/AST/Media+Format+Rewrite Finally, this work was the culmination of a large number of developer's efforts. Extra thanks goes to Corey Farrell, who took on a large amount of the work in the Asterisk core, chan_sip, and was an invaluable resource in peer reviews throughout this project. There were a substantial number of patches contributed during this work; the following issues/patch names simply reflect some of the work (and will cause the release scripts to give attribution to the individuals who work on them). Reviews: https://reviewboard.asterisk.org/r/3814 https://reviewboard.asterisk.org/r/3808 https://reviewboard.asterisk.org/r/3805 https://reviewboard.asterisk.org/r/3803 https://reviewboard.asterisk.org/r/3801 https://reviewboard.asterisk.org/r/3798 https://reviewboard.asterisk.org/r/3800 https://reviewboard.asterisk.org/r/3794 https://reviewboard.asterisk.org/r/3793 https://reviewboard.asterisk.org/r/3792 https://reviewboard.asterisk.org/r/3791 https://reviewboard.asterisk.org/r/3790 https://reviewboard.asterisk.org/r/3789 https://reviewboard.asterisk.org/r/3788 https://reviewboard.asterisk.org/r/3787 https://reviewboard.asterisk.org/r/3786 https://reviewboard.asterisk.org/r/3784 https://reviewboard.asterisk.org/r/3783 https://reviewboard.asterisk.org/r/3778 https://reviewboard.asterisk.org/r/3774 https://reviewboard.asterisk.org/r/3775 https://reviewboard.asterisk.org/r/3772 https://reviewboard.asterisk.org/r/3761 https://reviewboard.asterisk.org/r/3754 https://reviewboard.asterisk.org/r/3753 https://reviewboard.asterisk.org/r/3751 https://reviewboard.asterisk.org/r/3750 https://reviewboard.asterisk.org/r/3748 https://reviewboard.asterisk.org/r/3747 https://reviewboard.asterisk.org/r/3746 https://reviewboard.asterisk.org/r/3742 https://reviewboard.asterisk.org/r/3740 https://reviewboard.asterisk.org/r/3739 https://reviewboard.asterisk.org/r/3738 https://reviewboard.asterisk.org/r/3737 https://reviewboard.asterisk.org/r/3736 https://reviewboard.asterisk.org/r/3734 https://reviewboard.asterisk.org/r/3722 https://reviewboard.asterisk.org/r/3713 https://reviewboard.asterisk.org/r/3703 https://reviewboard.asterisk.org/r/3689 https://reviewboard.asterisk.org/r/3687 https://reviewboard.asterisk.org/r/3674 https://reviewboard.asterisk.org/r/3671 https://reviewboard.asterisk.org/r/3667 https://reviewboard.asterisk.org/r/3665 https://reviewboard.asterisk.org/r/3625 https://reviewboard.asterisk.org/r/3602 https://reviewboard.asterisk.org/r/3519 https://reviewboard.asterisk.org/r/3518 https://reviewboard.asterisk.org/r/3516 https://reviewboard.asterisk.org/r/3515 https://reviewboard.asterisk.org/r/3512 https://reviewboard.asterisk.org/r/3506 https://reviewboard.asterisk.org/r/3413 https://reviewboard.asterisk.org/r/3410 https://reviewboard.asterisk.org/r/3387 https://reviewboard.asterisk.org/r/3388 https://reviewboard.asterisk.org/r/3389 https://reviewboard.asterisk.org/r/3390 https://reviewboard.asterisk.org/r/3321 https://reviewboard.asterisk.org/r/3320 https://reviewboard.asterisk.org/r/3319 https://reviewboard.asterisk.org/r/3318 https://reviewboard.asterisk.org/r/3266 https://reviewboard.asterisk.org/r/3265 https://reviewboard.asterisk.org/r/3234 https://reviewboard.asterisk.org/r/3178 ASTERISK-23114 #close Reported by: mjordan media_formats_translation_core.diff uploaded by kharwell (License 6464) rb3506.diff uploaded by mjordan (License 6283) media_format_app_file.diff uploaded by kharwell (License 6464) misc-2.diff uploaded by file (License 5000) chan_mild-3.diff uploaded by file (License 5000) chan_obscure.diff uploaded by file (License 5000) jingle.diff uploaded by file (License 5000) funcs.diff uploaded by file (License 5000) formats.diff uploaded by file (License 5000) core.diff uploaded by file (License 5000) bridges.diff uploaded by file (License 5000) mf-codecs-2.diff uploaded by file (License 5000) mf-app_fax.diff uploaded by file (License 5000) mf-apps-3.diff uploaded by file (License 5000) media-formats-3.diff uploaded by file (License 5000) ASTERISK-23715 rb3713.patch uploaded by coreyfarrell (License 5909) rb3689.patch uploaded by mjordan (License 6283) ASTERISK-23957 rb3722.patch uploaded by mjordan (License 6283) mf-attributes-3.diff uploaded by file (License 5000) ASTERISK-23958 Tested by: jrose rb3822.patch uploaded by coreyfarrell (License 5909) rb3800.patch uploaded by jrose (License 6182) chan_sip.diff uploaded by mjordan (License 6283) rb3747.patch uploaded by jrose (License 6182) ASTERISK-23959 #close Tested by: sgriepentrog, mjordan, coreyfarrell sip_cleanup.diff uploaded by opticron (License 6273) chan_sip_caps.diff uploaded by mjordan (License 6283) rb3751.patch uploaded by coreyfarrell (License 5909) chan_sip-3.diff uploaded by file (License 5000) ASTERISK-23960 #close Tested by: opticron direct_media.diff uploaded by opticron (License 6273) pjsip-direct-media.diff uploaded by file (License 5000) format_cap_remove.diff uploaded by opticron (License 6273) media_format_fixes.diff uploaded by opticron (License 6273) chan_pjsip-2.diff uploaded by file (License 5000) ASTERISK-23966 #close Tested by: rmudgett rb3803.patch uploaded by rmudgetti (License 5621) chan_dahdi.diff uploaded by file (License 5000) ASTERISK-24064 #close Tested by: coreyfarrell, mjordan, opticron, file, rmudgett, sgriepentrog, jrose rb3814.patch uploaded by rmudgett (License 5621) moh_cleanup.diff uploaded by opticron (License 6273) bridge_leak.diff uploaded by opticron (License 6273) translate.diff uploaded by file (License 5000) rb3795.patch uploaded by rmudgett (License 5621) tls_fix.diff uploaded by mjordan (License 6283) fax-mf-fix-2.diff uploaded by file (License 5000) rtp_transfer_stuff uploaded by mjordan (License 6283) rb3787.patch uploaded by rmudgett (License 5621) media-formats-explicit-translate-format-3.diff uploaded by file (License 5000) format_cache_case_fix.diff uploaded by opticron (License 6273) rb3774.patch uploaded by rmudgett (License 5621) rb3775.patch uploaded by rmudgett (License 5621) rtp_engine_fix.diff uploaded by opticron (License 6273) rtp_crash_fix.diff uploaded by opticron (License 6273) rb3753.patch uploaded by mjordan (License 6283) rb3750.patch uploaded by mjordan (License 6283) rb3748.patch uploaded by rmudgett (License 5621) media_format_fixes.diff uploaded by opticron (License 6273) rb3740.patch uploaded by mjordan (License 6283) rb3739.patch uploaded by mjordan (License 6283) rb3734.patch uploaded by mjordan (License 6283) rb3689.patch uploaded by mjordan (License 6283) rb3674.patch uploaded by coreyfarrell (License 5909) rb3671.patch uploaded by coreyfarrell (License 5909) rb3667.patch uploaded by coreyfarrell (License 5909) rb3665.patch uploaded by mjordan (License 6283) rb3625.patch uploaded by coreyfarrell (License 5909) rb3602.patch uploaded by coreyfarrell (License 5909) format_compatibility-2.diff uploaded by file (License 5000) core.diff uploaded by file (License 5000) git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@419044 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2014-07-20 22:06:33 +00:00
#include "asterisk/format_cache.h"
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
#include "asterisk/taskprocessor.h"
app_confbridge: Enable sending events to participants ConfBridge can now send events to participants via in-dialog MESSAGEs. All current Confbridge events are supported, such as ConfbridgeJoin, ConfbridgeLeave, etc. In addition to those events, a new event ConfbridgeWelcome has been added that will send a list of all current participants to a new participant. For all but the ConfbridgeWelcome event, the JSON message contains information about the bridge, such as its id and name, and information about the channel that triggered the event such as channel name, callerid info, mute status, and the MSID labels for their audio and video tracks. You can use the labels to correlate callerid and mute status to specific video elements in a webrtc client. To control this behavior, the following options have been added to confbridge.conf: bridge_profile/enable_events: This must be enabled on any bridge where events are desired. user_profile/send_events: This must be set for a user profile to send events. Different user profiles connected to the same bridge can have different settings. This allows admins to get events but not normal users for instance. user_profile/echo_events: In some cases, you might not want the user triggering the event to get the event sent back to them. To prevent it, set this to false. A change was also made to res_pjsip_sdp_rtp to save the generated msid to the stream so it can be re-used. This allows participant A's video stream to appear as the same label to all other participants. Change-Id: I26420aa9f101f0b2387dc9e2fd10733197f1318e
2018-05-31 21:22:13 +00:00
#include "asterisk/stream.h"
#include "asterisk/message.h"
/*** DOCUMENTATION
<application name="ConfBridge" language="en_US">
<synopsis>
Conference bridge application.
</synopsis>
<syntax>
<parameter name="conference" required="true">
<para>Name of the conference bridge. You are not limited to just
numbers.</para>
</parameter>
<parameter name="bridge_profile">
<para>The bridge profile name from confbridge.conf. When left blank,
a dynamically built bridge profile created by the CONFBRIDGE dialplan
function is searched for on the channel and used. If no dynamic
profile is present, the 'default_bridge' profile found in
confbridge.conf is used. </para>
<para>It is important to note that while user profiles may be unique
for each participant, mixing bridge profiles on a single conference
is _NOT_ recommended and will produce undefined results.</para>
</parameter>
<parameter name="user_profile">
<para>The user profile name from confbridge.conf. When left blank,
a dynamically built user profile created by the CONFBRIDGE dialplan
function is searched for on the channel and used. If no dynamic
profile is present, the 'default_user' profile found in
confbridge.conf is used.</para>
</parameter>
<parameter name="menu">
<para>The name of the DTMF menu in confbridge.conf to be applied to
this channel. When left blank, a dynamically built menu profile
created by the CONFBRIDGE dialplan function is searched for on
the channel and used. If no dynamic profile is present, the
'default_menu' profile found in confbridge.conf is used.</para>
</parameter>
</syntax>
<description>
<para>Enters the user into a specified conference bridge. The user can
exit the conference by hangup or DTMF menu option.</para>
<para>This application sets the following channel variable upon completion:</para>
<variablelist>
<variable name="CONFBRIDGE_RESULT">
<value name="FAILED">The channel encountered an error and could not enter the conference.</value>
<value name="HANGUP">The channel exited the conference by hanging up.</value>
<value name="KICKED">The channel was kicked from the conference.</value>
<value name="ENDMARKED">The channel left the conference as a result of the last marked user leaving.</value>
<value name="DTMF">The channel pressed a DTMF sequence to exit the conference.</value>
<value name="TIMEOUT">The channel reached its configured timeout.</value>
</variable>
</variablelist>
</description>
<see-also>
<ref type="application">ConfKick</ref>
<ref type="function">CONFBRIDGE</ref>
<ref type="function">CONFBRIDGE_INFO</ref>
<ref type="function">CONFBRIDGE_CHANNELS</ref>
</see-also>
</application>
<application name="ConfKick" language="en_US">
<since>
<version>16.19.0</version>
<version>18.5.0</version>
<version>19.0.0</version>
</since>
<synopsis>
Kicks channel(s) from the requested ConfBridge.
</synopsis>
<syntax>
<parameter name="conference" required="true" />
<parameter name="channel">
<para>The channel to kick, <literal>all</literal>
to kick all users, or <literal>participants</literal>
to kick all non-admin participants. Default is all.</para>
</parameter>
</syntax>
<description>
<para>Kicks the requested channel(s) from a conference bridge.</para>
<variablelist>
<variable name="CONFKICKSTATUS">
<value name="FAILURE">
Could not kick any users with the provided arguments.
</value>
<value name="SUCCESS">
Successfully kicked users from specified conference bridge.
</value>
</variable>
</variablelist>
</description>
<see-also>
<ref type="application">ConfBridge</ref>
<ref type="function">CONFBRIDGE</ref>
<ref type="function">CONFBRIDGE_INFO</ref>
<ref type="function">CONFBRIDGE_CHANNELS</ref>
</see-also>
</application>
<function name="CONFBRIDGE" language="en_US">
<synopsis>
Set a custom dynamic bridge, user, or menu profile on a channel for the
ConfBridge application using the same options available in confbridge.conf.
</synopsis>
<syntax>
<parameter name="type" required="true">
<para>To what type of conference profile the option applies.</para>
<enumlist>
<enum name="bridge"></enum>
<enum name="menu"></enum>
<enum name="user"></enum>
</enumlist>
</parameter>
<parameter name="option" required="true">
<para>Option refers to a <filename>confbridge.conf</filename> option
that is being set dynamically on this channel, or <literal>clear</literal>
to remove already applied profile options from the channel.</para>
</parameter>
</syntax>
<description>
<para>A custom profile uses the default profile type settings defined in
<filename>confbridge.conf</filename> as defaults if the profile template
is not explicitly specified first.</para>
<para>For <literal>bridge</literal> profiles the default template is <literal>default_bridge</literal>.</para>
<para>For <literal>menu</literal> profiles the default template is <literal>default_menu</literal>.</para>
<para>For <literal>user</literal> profiles the default template is <literal>default_user</literal>.</para>
<para>---- Example 1 ----</para>
<para>In this example the custom user profile set on the channel will
automatically be used by the ConfBridge application.</para>
<example title="Example 1">
exten => 1,1,Answer()
</example>
<para>; In this example the effect of the following line is</para>
<para>; implied:</para>
<example title="Example 1b">
same => n,Set(CONFBRIDGE(user,template)=default_user)
same => n,Set(CONFBRIDGE(user,announce_join_leave)=yes)
same => n,Set(CONFBRIDGE(user,startmuted)=yes)
same => n,ConfBridge(1)
</example>
<para>---- Example 2 ----</para>
<para>This example shows how to use a predefined user profile in
<filename>confbridge.conf</filename> as a template for a dynamic profile.
Here we make an admin/marked user out of the <literal>my_user</literal>
profile that you define in <filename>confbridge.conf</filename>.</para>
<example title="Example 2">
exten => 1,1,Answer()
same => n,Set(CONFBRIDGE(user,template)=my_user)
same => n,Set(CONFBRIDGE(user,admin)=yes)
same => n,Set(CONFBRIDGE(user,marked)=yes)
same => n,ConfBridge(1)
</example>
</description>
</function>
<function name="CONFBRIDGE_INFO" language="en_US">
<synopsis>
Get information about a ConfBridge conference.
</synopsis>
<syntax>
<parameter name="type" required="true">
<para>What conference information is requested.</para>
<enumlist>
<enum name="admins">
<para>Get the number of admin users in the conference.</para>
</enum>
<enum name="locked">
<para>Determine if the conference is locked. (0 or 1)</para>
</enum>
<enum name="marked">
<para>Get the number of marked users in the conference.</para>
</enum>
<enum name="muted">
<para>Determine if the conference is muted. (0 or 1)</para>
</enum>
<enum name="parties">
<para>Get the number of users in the conference.</para>
</enum>
</enumlist>
</parameter>
<parameter name="conf" required="true">
<para>The name of the conference being referenced.</para>
</parameter>
</syntax>
<description>
<para>This function returns a non-negative integer for valid conference
names and an empty string for invalid conference names.</para>
</description>
<see-also>
<ref type="function">CONFBRIDGE_CHANNELS</ref>
</see-also>
</function>
<function name="CONFBRIDGE_CHANNELS" language="en_US">
<since>
<version>16.26.0</version>
<version>18.12.0</version>
<version>19.4.0</version>
</since>
<synopsis>
Get a list of channels in a ConfBridge conference.
</synopsis>
<syntax>
<parameter name="type" required="true">
<para>What conference information is requested.</para>
<enumlist>
<enum name="admins">
<para>Get the number of admin users in the conference.</para>
</enum>
<enum name="marked">
<para>Get the number of marked users in the conference.</para>
</enum>
<enum name="parties">
<para>Get the number of total users in the conference.</para>
</enum>
<enum name="active">
<para>Get the number of active users in the conference.</para>
</enum>
<enum name="waiting">
<para>Get the number of waiting users in the conference.</para>
</enum>
</enumlist>
</parameter>
<parameter name="conf" required="true">
<para>The name of the conference being referenced.</para>
</parameter>
</syntax>
<description>
<para>This function returns a comma-separated list of channels in a ConfBridge conference, optionally filtered by a type of participant.</para>
</description>
<see-also>
<ref type="function">CONFBRIDGE_INFO</ref>
</see-also>
</function>
<manager name="ConfbridgeList" language="en_US">
<synopsis>
List participants in a conference.
</synopsis>
<syntax>
<xi:include xpointer="xpointer(/docs/manager[@name='Login']/syntax/parameter[@name='ActionID'])" />
<parameter name="Conference" required="true">
<para>Conference number.</para>
</parameter>
</syntax>
<description>
<para>Lists all users in a particular ConfBridge conference.
ConfbridgeList will follow as separate events, followed by a final event called
ConfbridgeListComplete.</para>
</description>
</manager>
<managerEvent language="en_US" name="ConfbridgeList">
<managerEventInstance class="EVENT_FLAG_REPORTING">
<synopsis>Raised as part of the ConfbridgeList action response list.</synopsis>
<syntax>
<parameter name="Conference">
<para>The name of the Confbridge conference.</para>
</parameter>
<parameter name="Admin">
<para>Identifies this user as an admin user.</para>
<enumlist>
<enum name="Yes"/>
<enum name="No"/>
</enumlist>
</parameter>
<parameter name="MarkedUser">
<para>Identifies this user as a marked user.</para>
<enumlist>
<enum name="Yes"/>
<enum name="No"/>
</enumlist>
</parameter>
<parameter name="WaitMarked">
<para>Must this user wait for a marked user to join?</para>
<enumlist>
<enum name="Yes"/>
<enum name="No"/>
</enumlist>
</parameter>
<parameter name="EndMarked">
<para>Does this user get kicked after the last marked user leaves?</para>
<enumlist>
<enum name="Yes"/>
<enum name="No"/>
</enumlist>
</parameter>
<parameter name="Waiting">
<para>Is this user waiting for a marked user to join?</para>
<enumlist>
<enum name="Yes"/>
<enum name="No"/>
</enumlist>
</parameter>
<parameter name="Muted">
<para>The current mute status.</para>
<enumlist>
<enum name="Yes"/>
<enum name="No"/>
</enumlist>
</parameter>
<parameter name="Talking">
<para>Is this user talking?</para>
<enumlist>
<enum name="Yes"/>
<enum name="No"/>
</enumlist>
</parameter>
<parameter name="AnsweredTime">
<para>The number of seconds the channel has been up.</para>
</parameter>
<channel_snapshot/>
</syntax>
</managerEventInstance>
</managerEvent>
<manager name="ConfbridgeListRooms" language="en_US">
<synopsis>
List active conferences.
</synopsis>
<syntax>
<xi:include xpointer="xpointer(/docs/manager[@name='Login']/syntax/parameter[@name='ActionID'])" />
</syntax>
<description>
<para>Lists data about all active conferences.
ConfbridgeListRooms will follow as separate events, followed by a final event called
ConfbridgeListRoomsComplete.</para>
</description>
</manager>
<managerEvent language="en_US" name="ConfbridgeListRooms">
<managerEventInstance class="EVENT_FLAG_REPORTING">
<synopsis>Raised as part of the ConfbridgeListRooms action response list.</synopsis>
<syntax>
<parameter name="Conference">
<para>The name of the Confbridge conference.</para>
</parameter>
<parameter name="Parties">
<para>Number of users in the conference.</para>
<para>This includes both active and waiting users.</para>
</parameter>
<parameter name="Marked">
<para>Number of marked users in the conference.</para>
</parameter>
<parameter name="Locked">
<para>Is the conference locked?</para>
<enumlist>
<enum name="Yes"/>
<enum name="No"/>
</enumlist>
</parameter>
<parameter name="Muted">
<para>Is the conference muted?</para>
<enumlist>
<enum name="Yes"/>
<enum name="No"/>
</enumlist>
</parameter>
</syntax>
</managerEventInstance>
</managerEvent>
<manager name="ConfbridgeMute" language="en_US">
<synopsis>
Mute a Confbridge user.
</synopsis>
<syntax>
<xi:include xpointer="xpointer(/docs/manager[@name='Login']/syntax/parameter[@name='ActionID'])" />
<parameter name="Conference" required="true" />
<parameter name="Channel" required="true">
<para>If this parameter is not a complete channel name, the first channel with this prefix will be used.</para>
<para>If this parameter is "all", all channels will be muted.</para>
<para>If this parameter is "participants", all non-admin channels will be muted.</para>
</parameter>
</syntax>
<description>
</description>
</manager>
<manager name="ConfbridgeUnmute" language="en_US">
<synopsis>
Unmute a Confbridge user.
</synopsis>
<syntax>
<xi:include xpointer="xpointer(/docs/manager[@name='Login']/syntax/parameter[@name='ActionID'])" />
<parameter name="Conference" required="true" />
<parameter name="Channel" required="true">
<para>If this parameter is not a complete channel name, the first channel with this prefix will be used.</para>
<para>If this parameter is "all", all channels will be unmuted.</para>
<para>If this parameter is "participants", all non-admin channels will be unmuted.</para>
</parameter>
</syntax>
<description>
</description>
</manager>
<manager name="ConfbridgeKick" language="en_US">
<synopsis>
Kick a Confbridge user.
</synopsis>
<syntax>
<xi:include xpointer="xpointer(/docs/manager[@name='Login']/syntax/parameter[@name='ActionID'])" />
<parameter name="Conference" required="true" />
<parameter name="Channel" required="true" >
<para>If this parameter is "all", all channels will be kicked from the conference.</para>
<para>If this parameter is "participants", all non-admin channels will be kicked from the conference.</para>
</parameter>
</syntax>
<description>
</description>
</manager>
<manager name="ConfbridgeLock" language="en_US">
<synopsis>
Lock a Confbridge conference.
</synopsis>
<syntax>
<xi:include xpointer="xpointer(/docs/manager[@name='Login']/syntax/parameter[@name='ActionID'])" />
<parameter name="Conference" required="true" />
</syntax>
<description>
</description>
</manager>
<manager name="ConfbridgeUnlock" language="en_US">
<synopsis>
Unlock a Confbridge conference.
</synopsis>
<syntax>
<xi:include xpointer="xpointer(/docs/manager[@name='Login']/syntax/parameter[@name='ActionID'])" />
<parameter name="Conference" required="true" />
</syntax>
<description>
</description>
</manager>
<manager name="ConfbridgeStartRecord" language="en_US">
<synopsis>
Start recording a Confbridge conference.
</synopsis>
<syntax>
<xi:include xpointer="xpointer(/docs/manager[@name='Login']/syntax/parameter[@name='ActionID'])" />
<parameter name="Conference" required="true" />
<parameter name="RecordFile" required="false" />
</syntax>
<description>
<para>Start recording a conference. If recording is already present an error will be returned. If RecordFile is not provided, the default record file specified in the conference's bridge profile will be used, if that is not present either a file will automatically be generated in the monitor directory.</para>
</description>
</manager>
<manager name="ConfbridgeStopRecord" language="en_US">
<synopsis>
Stop recording a Confbridge conference.
</synopsis>
<syntax>
<xi:include xpointer="xpointer(/docs/manager[@name='Login']/syntax/parameter[@name='ActionID'])" />
<parameter name="Conference" required="true" />
</syntax>
<description>
</description>
</manager>
<manager name="ConfbridgeSetSingleVideoSrc" language="en_US">
<synopsis>
Set a conference user as the single video source distributed to all other participants.
</synopsis>
<syntax>
<xi:include xpointer="xpointer(/docs/manager[@name='Login']/syntax/parameter[@name='ActionID'])" />
<parameter name="Conference" required="true" />
<parameter name="Channel" required="true">
<para>If this parameter is not a complete channel name, the first channel with this prefix will be used.</para>
</parameter>
</syntax>
<description>
</description>
</manager>
***/
/*!
* \par Playing back a file to a channel in a conference
* You might notice in this application that while playing a sound file
* to a channel the actual conference bridge lock is not held. This is done so
* that other channels are not blocked from interacting with the conference bridge.
* Unfortunately because of this it is possible for things to change after the sound file
* is done being played. Data must therefore be checked after reacquiring the conference
* bridge lock if it is important.
*/
static const char app[] = "ConfBridge";
static const char app2[] = "ConfKick";
/*! Number of buckets our conference bridges container can have */
#define CONFERENCE_BRIDGE_BUCKETS 53
/*! Initial recording filename space. */
#define RECORD_FILENAME_INITIAL_SPACE 128
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
/*! \brief Container to hold all conference bridges in progress */
struct ao2_container *conference_bridges;
static void leave_conference(struct confbridge_user *user);
static int play_sound_number(struct confbridge_conference *conference, int say_number);
static int execute_menu_entry(struct confbridge_conference *conference,
struct confbridge_user *user,
struct ast_bridge_channel *bridge_channel,
struct conf_menu_entry *menu_entry,
struct conf_menu *menu);
/*! \brief Hashing function used for conference bridges container */
static int conference_bridge_hash_cb(const void *obj, const int flags)
{
const struct confbridge_conference *conference = obj;
const char *name = obj;
int hash;
switch (flags & (OBJ_POINTER | OBJ_KEY | OBJ_PARTIAL_KEY)) {
default:
case OBJ_POINTER:
name = conference->name;
/* Fall through */
case OBJ_KEY:
hash = ast_str_case_hash(name);
break;
case OBJ_PARTIAL_KEY:
/* Should never happen in hash callback. */
ast_assert(0);
hash = 0;
break;
}
return hash;
}
/*! \brief Comparison function used for conference bridges container */
static int conference_bridge_cmp_cb(void *obj, void *arg, int flags)
{
const struct confbridge_conference *left = obj;
const struct confbridge_conference *right = arg;
const char *right_name = arg;
int cmp;
switch (flags & (OBJ_POINTER | OBJ_KEY | OBJ_PARTIAL_KEY)) {
default:
case OBJ_POINTER:
right_name = right->name;
/* Fall through */
case OBJ_KEY:
cmp = strcasecmp(left->name, right_name);
break;
case OBJ_PARTIAL_KEY:
cmp = strncasecmp(left->name, right_name, strlen(right_name));
break;
}
return cmp ? 0 : CMP_MATCH;
}
const char *conf_get_sound(enum conf_sounds sound, struct bridge_profile_sounds *custom_sounds)
{
switch (sound) {
case CONF_SOUND_HAS_JOINED:
return S_OR(custom_sounds->hasjoin, "conf-hasjoin");
case CONF_SOUND_HAS_LEFT:
return S_OR(custom_sounds->hasleft, "conf-hasleft");
case CONF_SOUND_KICKED:
return S_OR(custom_sounds->kicked, "conf-kicked");
case CONF_SOUND_MUTED:
return S_OR(custom_sounds->muted, "conf-muted");
case CONF_SOUND_UNMUTED:
return S_OR(custom_sounds->unmuted, "conf-unmuted");
case CONF_SOUND_BINAURAL_ON:
return S_OR(custom_sounds->binauralon, "confbridge-binaural-on");
case CONF_SOUND_BINAURAL_OFF:
return S_OR(custom_sounds->binauraloff, "confbridge-binaural-off");
case CONF_SOUND_ONLY_ONE:
return S_OR(custom_sounds->onlyone, "conf-onlyone");
case CONF_SOUND_THERE_ARE:
return S_OR(custom_sounds->thereare, "conf-thereare");
case CONF_SOUND_OTHER_IN_PARTY:
return S_OR(custom_sounds->otherinparty, "conf-otherinparty");
case CONF_SOUND_PLACE_IN_CONF:
return S_OR(custom_sounds->placeintoconf, "conf-placeintoconf");
case CONF_SOUND_WAIT_FOR_LEADER:
return S_OR(custom_sounds->waitforleader, "conf-waitforleader");
case CONF_SOUND_LEADER_HAS_LEFT:
return S_OR(custom_sounds->leaderhasleft, "conf-leaderhasleft");
case CONF_SOUND_GET_PIN:
return S_OR(custom_sounds->getpin, "conf-getpin");
case CONF_SOUND_INVALID_PIN:
return S_OR(custom_sounds->invalidpin, "conf-invalidpin");
case CONF_SOUND_ONLY_PERSON:
return S_OR(custom_sounds->onlyperson, "conf-onlyperson");
case CONF_SOUND_LOCKED:
return S_OR(custom_sounds->locked, "conf-locked");
case CONF_SOUND_LOCKED_NOW:
return S_OR(custom_sounds->lockednow, "conf-lockednow");
case CONF_SOUND_UNLOCKED_NOW:
return S_OR(custom_sounds->unlockednow, "conf-unlockednow");
case CONF_SOUND_ERROR_MENU:
return S_OR(custom_sounds->errormenu, "conf-errormenu");
case CONF_SOUND_JOIN:
return S_OR(custom_sounds->join, "confbridge-join");
case CONF_SOUND_LEAVE:
return S_OR(custom_sounds->leave, "confbridge-leave");
case CONF_SOUND_PARTICIPANTS_MUTED:
return S_OR(custom_sounds->participantsmuted, "conf-now-muted");
case CONF_SOUND_PARTICIPANTS_UNMUTED:
return S_OR(custom_sounds->participantsunmuted, "conf-now-unmuted");
case CONF_SOUND_BEGIN:
return S_OR(custom_sounds->begin, "confbridge-conf-begin");
}
return "";
}
app_confbridge: Enable sending events to participants ConfBridge can now send events to participants via in-dialog MESSAGEs. All current Confbridge events are supported, such as ConfbridgeJoin, ConfbridgeLeave, etc. In addition to those events, a new event ConfbridgeWelcome has been added that will send a list of all current participants to a new participant. For all but the ConfbridgeWelcome event, the JSON message contains information about the bridge, such as its id and name, and information about the channel that triggered the event such as channel name, callerid info, mute status, and the MSID labels for their audio and video tracks. You can use the labels to correlate callerid and mute status to specific video elements in a webrtc client. To control this behavior, the following options have been added to confbridge.conf: bridge_profile/enable_events: This must be enabled on any bridge where events are desired. user_profile/send_events: This must be set for a user profile to send events. Different user profiles connected to the same bridge can have different settings. This allows admins to get events but not normal users for instance. user_profile/echo_events: In some cases, you might not want the user triggering the event to get the event sent back to them. To prevent it, set this to false. A change was also made to res_pjsip_sdp_rtp to save the generated msid to the stream so it can be re-used. This allows participant A's video stream to appear as the same label to all other participants. Change-Id: I26420aa9f101f0b2387dc9e2fd10733197f1318e
2018-05-31 21:22:13 +00:00
static void send_conf_stasis(struct confbridge_conference *conference, struct ast_channel *chan,
struct stasis_message_type *type, struct ast_json *extras, int channel_topic)
{
RAII_VAR(struct stasis_message *, msg, NULL, ao2_cleanup);
RAII_VAR(struct ast_json *, json_object, NULL, ast_json_unref);
json_object = ast_json_pack("{s: s}",
"conference", conference->name);
if (!json_object) {
return;
}
if (extras) {
ast_json_object_update(json_object, extras);
}
ast_bridge_lock(conference->bridge);
msg = ast_bridge_blob_create(type,
conference->bridge,
chan,
json_object);
ast_bridge_unlock(conference->bridge);
if (!msg) {
return;
}
if (ast_test_flag(&conference->b_profile, BRIDGE_OPT_ENABLE_EVENTS)) {
conf_send_event_to_participants(conference, chan, msg);
}
if (channel_topic) {
stasis_publish(ast_channel_topic(chan), msg);
} else {
stasis_publish(ast_bridge_topic(conference->bridge), msg);
}
}
static void send_conf_stasis_snapshots(struct confbridge_conference *conference,
struct ast_channel_snapshot *chan_snapshot, struct stasis_message_type *type,
struct ast_json *extras)
{
RAII_VAR(struct stasis_message *, msg, NULL, ao2_cleanup);
RAII_VAR(struct ast_json *, json_object, NULL, ast_json_unref);
RAII_VAR(struct ast_bridge_snapshot *, bridge_snapshot, NULL, ao2_cleanup);
json_object = ast_json_pack("{s: s}",
"conference", conference->name);
if (!json_object) {
return;
}
if (extras) {
ast_json_object_update(json_object, extras);
}
ast_bridge_lock(conference->bridge);
bridge_snapshot = ast_bridge_snapshot_create(conference->bridge);
ast_bridge_unlock(conference->bridge);
if (!bridge_snapshot) {
return;
}
msg = ast_bridge_blob_create_from_snapshots(type,
bridge_snapshot,
chan_snapshot,
json_object);
if (!msg) {
return;
}
stasis_publish(ast_bridge_topic(conference->bridge), msg);
}
static void send_conf_start_event(struct confbridge_conference *conference)
{
send_conf_stasis(conference, NULL, confbridge_start_type(), NULL, 0);
}
static void send_conf_end_event(struct confbridge_conference *conference)
{
send_conf_stasis(conference, NULL, confbridge_end_type(), NULL, 0);
}
static void send_join_event(struct confbridge_user *user, struct confbridge_conference *conference)
{
struct ast_json *json_object;
json_object = ast_json_pack("{s: b, s: b}",
"admin", ast_test_flag(&user->u_profile, USER_OPT_ADMIN),
"muted", user->muted);
if (!json_object) {
return;
}
send_conf_stasis(conference, user->chan, confbridge_join_type(), json_object, 0);
ast_json_unref(json_object);
}
static void send_leave_event(struct confbridge_user *user, struct confbridge_conference *conference)
{
struct ast_json *json_object;
json_object = ast_json_pack("{s: b}",
"admin", ast_test_flag(&user->u_profile, USER_OPT_ADMIN)
);
if (!json_object) {
return;
}
send_conf_stasis(conference, user->chan, confbridge_leave_type(), json_object, 0);
ast_json_unref(json_object);
}
static void send_start_record_event(struct confbridge_conference *conference)
{
send_conf_stasis(conference, NULL, confbridge_start_record_type(), NULL, 0);
}
static void send_stop_record_event(struct confbridge_conference *conference)
{
send_conf_stasis(conference, NULL, confbridge_stop_record_type(), NULL, 0);
}
static void send_mute_event(struct confbridge_user *user, struct confbridge_conference *conference)
{
struct ast_json *json_object;
json_object = ast_json_pack("{s: b}",
"admin", ast_test_flag(&user->u_profile, USER_OPT_ADMIN)
);
if (!json_object) {
return;
}
send_conf_stasis(conference, user->chan, confbridge_mute_type(), json_object, 1);
ast_json_unref(json_object);
}
static void send_unmute_event(struct confbridge_user *user, struct confbridge_conference *conference)
{
struct ast_json *json_object;
json_object = ast_json_pack("{s: b}",
"admin", ast_test_flag(&user->u_profile, USER_OPT_ADMIN)
);
if (!json_object) {
return;
}
send_conf_stasis(conference, user->chan, confbridge_unmute_type(), json_object, 1);
ast_json_unref(json_object);
}
static void set_rec_filename(struct confbridge_conference *conference, struct ast_str **filename, int is_new)
{
char *rec_file = conference->b_profile.rec_file;
char *ext;
time_t now;
if (ast_str_strlen(*filename)
&& ast_test_flag(&conference->b_profile, BRIDGE_OPT_RECORD_FILE_APPEND)
&& !is_new) {
return;
}
time(&now);
ast_str_reset(*filename);
if (ast_strlen_zero(rec_file)) {
ast_str_set(filename, 0, "confbridge-%s-%u.wav", conference->name,
(unsigned int) now);
} else if (ast_test_flag(&conference->b_profile, BRIDGE_OPT_RECORD_FILE_TIMESTAMP)) {
/* insert time before file extension */
ext = strrchr(rec_file, '.');
if (ext) {
ast_str_set_substr(filename, 0, rec_file, ext - rec_file);
ast_str_append(filename, 0, "-%u%s", (unsigned int) now, ext);
} else {
ast_str_set(filename, 0, "%s-%u", rec_file, (unsigned int) now);
}
} else {
ast_str_set(filename, 0, "%s", rec_file);
}
ast_str_append(filename, 0, ",%s%s,%s",
ast_test_flag(&conference->b_profile, BRIDGE_OPT_RECORD_FILE_APPEND) ? "a" : "",
conference->b_profile.rec_options,
conference->b_profile.rec_command);
}
static int is_new_rec_file(const char *rec_file, struct ast_str **orig_rec_file)
{
if (!ast_strlen_zero(rec_file)) {
if (!*orig_rec_file) {
*orig_rec_file = ast_str_create(RECORD_FILENAME_INITIAL_SPACE);
}
if (*orig_rec_file
&& strcmp(ast_str_buffer(*orig_rec_file), rec_file)) {
ast_str_set(orig_rec_file, 0, "%s", rec_file);
return 1;
}
}
return 0;
}
struct confbridge_conference *conf_find_bridge(const char *conference_name)
{
return ao2_find(conference_bridges, conference_name, OBJ_KEY);
}
/*!
* \internal
* \brief Returns whether or not conference is being recorded.
*
* \param conference The bridge to check for recording
*
* \note Must be called with the conference locked
*
* \retval 1 conference is recording.
* \retval 0 conference is NOT recording.
*/
static int conf_is_recording(struct confbridge_conference *conference)
{
return conference->record_chan != NULL;
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
}
/*!
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
* \internal
* \brief Stop recording a conference bridge
*
* \param conference The conference bridge on which to stop the recording
*
* \note Must be called with the conference locked
*
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
* \retval -1 Failure
* \retval 0 Success
*/
static int conf_stop_record(struct confbridge_conference *conference)
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
{
struct ast_channel *chan;
struct ast_frame f = { AST_FRAME_CONTROL, .subclass.integer = AST_CONTROL_HANGUP };
if (!conf_is_recording(conference)) {
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
return -1;
}
/* Remove the recording channel from the conference bridge. */
chan = conference->record_chan;
conference->record_chan = NULL;
ast_queue_frame(chan, &f);
ast_channel_unref(chan);
ast_test_suite_event_notify("CONF_STOP_RECORD", "Message: stopped conference recording channel\r\nConference: %s", conference->b_profile.name);
send_stop_record_event(conference);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
return 0;
}
/*!
* \internal
* \brief Start recording the conference
*
* \param conference The conference bridge to start recording
*
* \note Must be called with the conference locked
*
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
* \retval 0 success
* \retval non-zero failure
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
*/
static int conf_start_record(struct confbridge_conference *conference)
{
struct ast_app *mixmonapp;
struct ast_channel *chan;
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
struct ast_format_cap *cap;
struct ast_bridge_features *features;
if (conf_is_recording(conference)) {
return -1;
}
mixmonapp = pbx_findapp("MixMonitor");
if (!mixmonapp) {
ast_log(LOG_WARNING, "Cannot record ConfBridge, MixMonitor app is not installed\n");
return -1;
}
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
features = ast_bridge_features_new();
if (!features) {
return -1;
}
ast_set_flag(&features->feature_flags, AST_BRIDGE_CHANNEL_FLAG_IMMOVABLE);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
media formats: re-architect handling of media for performance improvements In the old times media formats were represented using a bit field. This was fast but had a few limitations. 1. Asterisk was limited in how many formats it could handle. 2. Formats, being a bit field, could not include any attribute information. A format was strictly its type, e.g., "this is ulaw". This was changed in Asterisk 10 (see https://wiki.asterisk.org/wiki/display/AST/Media+Architecture+Proposal for notes on that work) which led to the creation of the ast_format structure. This structure allowed Asterisk to handle attributes and bundle information with a format. Additionally, ast_format_cap was created to act as a container for multiple formats that, together, formed the capability of some entity. Another mechanism was added to allow logic to be registered which performed format attribute negotiation. Everywhere throughout the codebase Asterisk was changed to use this strategy. Unfortunately, in software, there is no free lunch. These new capabilities came at a cost. Performance analysis and profiling showed that we spend an inordinate amount of time comparing, copying, and generally manipulating formats and their related structures. Basic prototyping has shown that a reasonably large performance improvement could be made in this area. This patch is the result of that project, which overhauled the media format architecture and its usage in Asterisk to improve performance. Generally, the new philosophy for handling formats is as follows: * The ast_format structure is reference counted. This removed a large amount of the memory allocations and copying that was done in prior versions. * In order to prevent race conditions while keeping things performant, the ast_format structure is immutable by convention and lock-free. Violate this tenet at your peril! * Because formats are reference counted, codecs are also reference counted. The Asterisk core generally provides built-in codecs and caches the ast_format structures created to represent them. Generally, to prevent inordinate amounts of module reference bumping, codecs and formats can be added at run-time but cannot be removed. * All compatibility with the bit field representation of codecs/formats has been moved to a compatibility API. The primary user of this representation is chan_iax2, which must continue to maintain its bit-field usage of formats for interoperability concerns. * When a format is negotiated with attributes, or when a format cannot be represented by one of the cached formats, a new format object is created or cloned from an existing format. That format may have the same codec underlying it, but is a different format than a version of the format with different attributes or without attributes. * While formats are reference counted objects, the reference count maintained on the format should be manipulated with care. Formats are generally cached and will persist for the lifetime of Asterisk and do not explicitly need to have their lifetime modified. An exception to this is when the user of a format does not know where the format came from *and* the user may outlive the provider of the format. This occurs, for example, when a format is read from a channel: the channel may have a format with attributes (hence, non-cached) and the user of the format may last longer than the channel (if the reference to the channel is released prior to the format's reference). For more information on this work, see the API design notes: https://wiki.asterisk.org/wiki/display/AST/Media+Format+Rewrite Finally, this work was the culmination of a large number of developer's efforts. Extra thanks goes to Corey Farrell, who took on a large amount of the work in the Asterisk core, chan_sip, and was an invaluable resource in peer reviews throughout this project. There were a substantial number of patches contributed during this work; the following issues/patch names simply reflect some of the work (and will cause the release scripts to give attribution to the individuals who work on them). Reviews: https://reviewboard.asterisk.org/r/3814 https://reviewboard.asterisk.org/r/3808 https://reviewboard.asterisk.org/r/3805 https://reviewboard.asterisk.org/r/3803 https://reviewboard.asterisk.org/r/3801 https://reviewboard.asterisk.org/r/3798 https://reviewboard.asterisk.org/r/3800 https://reviewboard.asterisk.org/r/3794 https://reviewboard.asterisk.org/r/3793 https://reviewboard.asterisk.org/r/3792 https://reviewboard.asterisk.org/r/3791 https://reviewboard.asterisk.org/r/3790 https://reviewboard.asterisk.org/r/3789 https://reviewboard.asterisk.org/r/3788 https://reviewboard.asterisk.org/r/3787 https://reviewboard.asterisk.org/r/3786 https://reviewboard.asterisk.org/r/3784 https://reviewboard.asterisk.org/r/3783 https://reviewboard.asterisk.org/r/3778 https://reviewboard.asterisk.org/r/3774 https://reviewboard.asterisk.org/r/3775 https://reviewboard.asterisk.org/r/3772 https://reviewboard.asterisk.org/r/3761 https://reviewboard.asterisk.org/r/3754 https://reviewboard.asterisk.org/r/3753 https://reviewboard.asterisk.org/r/3751 https://reviewboard.asterisk.org/r/3750 https://reviewboard.asterisk.org/r/3748 https://reviewboard.asterisk.org/r/3747 https://reviewboard.asterisk.org/r/3746 https://reviewboard.asterisk.org/r/3742 https://reviewboard.asterisk.org/r/3740 https://reviewboard.asterisk.org/r/3739 https://reviewboard.asterisk.org/r/3738 https://reviewboard.asterisk.org/r/3737 https://reviewboard.asterisk.org/r/3736 https://reviewboard.asterisk.org/r/3734 https://reviewboard.asterisk.org/r/3722 https://reviewboard.asterisk.org/r/3713 https://reviewboard.asterisk.org/r/3703 https://reviewboard.asterisk.org/r/3689 https://reviewboard.asterisk.org/r/3687 https://reviewboard.asterisk.org/r/3674 https://reviewboard.asterisk.org/r/3671 https://reviewboard.asterisk.org/r/3667 https://reviewboard.asterisk.org/r/3665 https://reviewboard.asterisk.org/r/3625 https://reviewboard.asterisk.org/r/3602 https://reviewboard.asterisk.org/r/3519 https://reviewboard.asterisk.org/r/3518 https://reviewboard.asterisk.org/r/3516 https://reviewboard.asterisk.org/r/3515 https://reviewboard.asterisk.org/r/3512 https://reviewboard.asterisk.org/r/3506 https://reviewboard.asterisk.org/r/3413 https://reviewboard.asterisk.org/r/3410 https://reviewboard.asterisk.org/r/3387 https://reviewboard.asterisk.org/r/3388 https://reviewboard.asterisk.org/r/3389 https://reviewboard.asterisk.org/r/3390 https://reviewboard.asterisk.org/r/3321 https://reviewboard.asterisk.org/r/3320 https://reviewboard.asterisk.org/r/3319 https://reviewboard.asterisk.org/r/3318 https://reviewboard.asterisk.org/r/3266 https://reviewboard.asterisk.org/r/3265 https://reviewboard.asterisk.org/r/3234 https://reviewboard.asterisk.org/r/3178 ASTERISK-23114 #close Reported by: mjordan media_formats_translation_core.diff uploaded by kharwell (License 6464) rb3506.diff uploaded by mjordan (License 6283) media_format_app_file.diff uploaded by kharwell (License 6464) misc-2.diff uploaded by file (License 5000) chan_mild-3.diff uploaded by file (License 5000) chan_obscure.diff uploaded by file (License 5000) jingle.diff uploaded by file (License 5000) funcs.diff uploaded by file (License 5000) formats.diff uploaded by file (License 5000) core.diff uploaded by file (License 5000) bridges.diff uploaded by file (License 5000) mf-codecs-2.diff uploaded by file (License 5000) mf-app_fax.diff uploaded by file (License 5000) mf-apps-3.diff uploaded by file (License 5000) media-formats-3.diff uploaded by file (License 5000) ASTERISK-23715 rb3713.patch uploaded by coreyfarrell (License 5909) rb3689.patch uploaded by mjordan (License 6283) ASTERISK-23957 rb3722.patch uploaded by mjordan (License 6283) mf-attributes-3.diff uploaded by file (License 5000) ASTERISK-23958 Tested by: jrose rb3822.patch uploaded by coreyfarrell (License 5909) rb3800.patch uploaded by jrose (License 6182) chan_sip.diff uploaded by mjordan (License 6283) rb3747.patch uploaded by jrose (License 6182) ASTERISK-23959 #close Tested by: sgriepentrog, mjordan, coreyfarrell sip_cleanup.diff uploaded by opticron (License 6273) chan_sip_caps.diff uploaded by mjordan (License 6283) rb3751.patch uploaded by coreyfarrell (License 5909) chan_sip-3.diff uploaded by file (License 5000) ASTERISK-23960 #close Tested by: opticron direct_media.diff uploaded by opticron (License 6273) pjsip-direct-media.diff uploaded by file (License 5000) format_cap_remove.diff uploaded by opticron (License 6273) media_format_fixes.diff uploaded by opticron (License 6273) chan_pjsip-2.diff uploaded by file (License 5000) ASTERISK-23966 #close Tested by: rmudgett rb3803.patch uploaded by rmudgetti (License 5621) chan_dahdi.diff uploaded by file (License 5000) ASTERISK-24064 #close Tested by: coreyfarrell, mjordan, opticron, file, rmudgett, sgriepentrog, jrose rb3814.patch uploaded by rmudgett (License 5621) moh_cleanup.diff uploaded by opticron (License 6273) bridge_leak.diff uploaded by opticron (License 6273) translate.diff uploaded by file (License 5000) rb3795.patch uploaded by rmudgett (License 5621) tls_fix.diff uploaded by mjordan (License 6283) fax-mf-fix-2.diff uploaded by file (License 5000) rtp_transfer_stuff uploaded by mjordan (License 6283) rb3787.patch uploaded by rmudgett (License 5621) media-formats-explicit-translate-format-3.diff uploaded by file (License 5000) format_cache_case_fix.diff uploaded by opticron (License 6273) rb3774.patch uploaded by rmudgett (License 5621) rb3775.patch uploaded by rmudgett (License 5621) rtp_engine_fix.diff uploaded by opticron (License 6273) rtp_crash_fix.diff uploaded by opticron (License 6273) rb3753.patch uploaded by mjordan (License 6283) rb3750.patch uploaded by mjordan (License 6283) rb3748.patch uploaded by rmudgett (License 5621) media_format_fixes.diff uploaded by opticron (License 6273) rb3740.patch uploaded by mjordan (License 6283) rb3739.patch uploaded by mjordan (License 6283) rb3734.patch uploaded by mjordan (License 6283) rb3689.patch uploaded by mjordan (License 6283) rb3674.patch uploaded by coreyfarrell (License 5909) rb3671.patch uploaded by coreyfarrell (License 5909) rb3667.patch uploaded by coreyfarrell (License 5909) rb3665.patch uploaded by mjordan (License 6283) rb3625.patch uploaded by coreyfarrell (License 5909) rb3602.patch uploaded by coreyfarrell (License 5909) format_compatibility-2.diff uploaded by file (License 5000) core.diff uploaded by file (License 5000) git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@419044 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2014-07-20 22:06:33 +00:00
cap = ast_format_cap_alloc(AST_FORMAT_CAP_FLAG_DEFAULT);
if (!cap) {
ast_bridge_features_destroy(features);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
return -1;
}
media formats: re-architect handling of media for performance improvements In the old times media formats were represented using a bit field. This was fast but had a few limitations. 1. Asterisk was limited in how many formats it could handle. 2. Formats, being a bit field, could not include any attribute information. A format was strictly its type, e.g., "this is ulaw". This was changed in Asterisk 10 (see https://wiki.asterisk.org/wiki/display/AST/Media+Architecture+Proposal for notes on that work) which led to the creation of the ast_format structure. This structure allowed Asterisk to handle attributes and bundle information with a format. Additionally, ast_format_cap was created to act as a container for multiple formats that, together, formed the capability of some entity. Another mechanism was added to allow logic to be registered which performed format attribute negotiation. Everywhere throughout the codebase Asterisk was changed to use this strategy. Unfortunately, in software, there is no free lunch. These new capabilities came at a cost. Performance analysis and profiling showed that we spend an inordinate amount of time comparing, copying, and generally manipulating formats and their related structures. Basic prototyping has shown that a reasonably large performance improvement could be made in this area. This patch is the result of that project, which overhauled the media format architecture and its usage in Asterisk to improve performance. Generally, the new philosophy for handling formats is as follows: * The ast_format structure is reference counted. This removed a large amount of the memory allocations and copying that was done in prior versions. * In order to prevent race conditions while keeping things performant, the ast_format structure is immutable by convention and lock-free. Violate this tenet at your peril! * Because formats are reference counted, codecs are also reference counted. The Asterisk core generally provides built-in codecs and caches the ast_format structures created to represent them. Generally, to prevent inordinate amounts of module reference bumping, codecs and formats can be added at run-time but cannot be removed. * All compatibility with the bit field representation of codecs/formats has been moved to a compatibility API. The primary user of this representation is chan_iax2, which must continue to maintain its bit-field usage of formats for interoperability concerns. * When a format is negotiated with attributes, or when a format cannot be represented by one of the cached formats, a new format object is created or cloned from an existing format. That format may have the same codec underlying it, but is a different format than a version of the format with different attributes or without attributes. * While formats are reference counted objects, the reference count maintained on the format should be manipulated with care. Formats are generally cached and will persist for the lifetime of Asterisk and do not explicitly need to have their lifetime modified. An exception to this is when the user of a format does not know where the format came from *and* the user may outlive the provider of the format. This occurs, for example, when a format is read from a channel: the channel may have a format with attributes (hence, non-cached) and the user of the format may last longer than the channel (if the reference to the channel is released prior to the format's reference). For more information on this work, see the API design notes: https://wiki.asterisk.org/wiki/display/AST/Media+Format+Rewrite Finally, this work was the culmination of a large number of developer's efforts. Extra thanks goes to Corey Farrell, who took on a large amount of the work in the Asterisk core, chan_sip, and was an invaluable resource in peer reviews throughout this project. There were a substantial number of patches contributed during this work; the following issues/patch names simply reflect some of the work (and will cause the release scripts to give attribution to the individuals who work on them). Reviews: https://reviewboard.asterisk.org/r/3814 https://reviewboard.asterisk.org/r/3808 https://reviewboard.asterisk.org/r/3805 https://reviewboard.asterisk.org/r/3803 https://reviewboard.asterisk.org/r/3801 https://reviewboard.asterisk.org/r/3798 https://reviewboard.asterisk.org/r/3800 https://reviewboard.asterisk.org/r/3794 https://reviewboard.asterisk.org/r/3793 https://reviewboard.asterisk.org/r/3792 https://reviewboard.asterisk.org/r/3791 https://reviewboard.asterisk.org/r/3790 https://reviewboard.asterisk.org/r/3789 https://reviewboard.asterisk.org/r/3788 https://reviewboard.asterisk.org/r/3787 https://reviewboard.asterisk.org/r/3786 https://reviewboard.asterisk.org/r/3784 https://reviewboard.asterisk.org/r/3783 https://reviewboard.asterisk.org/r/3778 https://reviewboard.asterisk.org/r/3774 https://reviewboard.asterisk.org/r/3775 https://reviewboard.asterisk.org/r/3772 https://reviewboard.asterisk.org/r/3761 https://reviewboard.asterisk.org/r/3754 https://reviewboard.asterisk.org/r/3753 https://reviewboard.asterisk.org/r/3751 https://reviewboard.asterisk.org/r/3750 https://reviewboard.asterisk.org/r/3748 https://reviewboard.asterisk.org/r/3747 https://reviewboard.asterisk.org/r/3746 https://reviewboard.asterisk.org/r/3742 https://reviewboard.asterisk.org/r/3740 https://reviewboard.asterisk.org/r/3739 https://reviewboard.asterisk.org/r/3738 https://reviewboard.asterisk.org/r/3737 https://reviewboard.asterisk.org/r/3736 https://reviewboard.asterisk.org/r/3734 https://reviewboard.asterisk.org/r/3722 https://reviewboard.asterisk.org/r/3713 https://reviewboard.asterisk.org/r/3703 https://reviewboard.asterisk.org/r/3689 https://reviewboard.asterisk.org/r/3687 https://reviewboard.asterisk.org/r/3674 https://reviewboard.asterisk.org/r/3671 https://reviewboard.asterisk.org/r/3667 https://reviewboard.asterisk.org/r/3665 https://reviewboard.asterisk.org/r/3625 https://reviewboard.asterisk.org/r/3602 https://reviewboard.asterisk.org/r/3519 https://reviewboard.asterisk.org/r/3518 https://reviewboard.asterisk.org/r/3516 https://reviewboard.asterisk.org/r/3515 https://reviewboard.asterisk.org/r/3512 https://reviewboard.asterisk.org/r/3506 https://reviewboard.asterisk.org/r/3413 https://reviewboard.asterisk.org/r/3410 https://reviewboard.asterisk.org/r/3387 https://reviewboard.asterisk.org/r/3388 https://reviewboard.asterisk.org/r/3389 https://reviewboard.asterisk.org/r/3390 https://reviewboard.asterisk.org/r/3321 https://reviewboard.asterisk.org/r/3320 https://reviewboard.asterisk.org/r/3319 https://reviewboard.asterisk.org/r/3318 https://reviewboard.asterisk.org/r/3266 https://reviewboard.asterisk.org/r/3265 https://reviewboard.asterisk.org/r/3234 https://reviewboard.asterisk.org/r/3178 ASTERISK-23114 #close Reported by: mjordan media_formats_translation_core.diff uploaded by kharwell (License 6464) rb3506.diff uploaded by mjordan (License 6283) media_format_app_file.diff uploaded by kharwell (License 6464) misc-2.diff uploaded by file (License 5000) chan_mild-3.diff uploaded by file (License 5000) chan_obscure.diff uploaded by file (License 5000) jingle.diff uploaded by file (License 5000) funcs.diff uploaded by file (License 5000) formats.diff uploaded by file (License 5000) core.diff uploaded by file (License 5000) bridges.diff uploaded by file (License 5000) mf-codecs-2.diff uploaded by file (License 5000) mf-app_fax.diff uploaded by file (License 5000) mf-apps-3.diff uploaded by file (License 5000) media-formats-3.diff uploaded by file (License 5000) ASTERISK-23715 rb3713.patch uploaded by coreyfarrell (License 5909) rb3689.patch uploaded by mjordan (License 6283) ASTERISK-23957 rb3722.patch uploaded by mjordan (License 6283) mf-attributes-3.diff uploaded by file (License 5000) ASTERISK-23958 Tested by: jrose rb3822.patch uploaded by coreyfarrell (License 5909) rb3800.patch uploaded by jrose (License 6182) chan_sip.diff uploaded by mjordan (License 6283) rb3747.patch uploaded by jrose (License 6182) ASTERISK-23959 #close Tested by: sgriepentrog, mjordan, coreyfarrell sip_cleanup.diff uploaded by opticron (License 6273) chan_sip_caps.diff uploaded by mjordan (License 6283) rb3751.patch uploaded by coreyfarrell (License 5909) chan_sip-3.diff uploaded by file (License 5000) ASTERISK-23960 #close Tested by: opticron direct_media.diff uploaded by opticron (License 6273) pjsip-direct-media.diff uploaded by file (License 5000) format_cap_remove.diff uploaded by opticron (License 6273) media_format_fixes.diff uploaded by opticron (License 6273) chan_pjsip-2.diff uploaded by file (License 5000) ASTERISK-23966 #close Tested by: rmudgett rb3803.patch uploaded by rmudgetti (License 5621) chan_dahdi.diff uploaded by file (License 5000) ASTERISK-24064 #close Tested by: coreyfarrell, mjordan, opticron, file, rmudgett, sgriepentrog, jrose rb3814.patch uploaded by rmudgett (License 5621) moh_cleanup.diff uploaded by opticron (License 6273) bridge_leak.diff uploaded by opticron (License 6273) translate.diff uploaded by file (License 5000) rb3795.patch uploaded by rmudgett (License 5621) tls_fix.diff uploaded by mjordan (License 6283) fax-mf-fix-2.diff uploaded by file (License 5000) rtp_transfer_stuff uploaded by mjordan (License 6283) rb3787.patch uploaded by rmudgett (License 5621) media-formats-explicit-translate-format-3.diff uploaded by file (License 5000) format_cache_case_fix.diff uploaded by opticron (License 6273) rb3774.patch uploaded by rmudgett (License 5621) rb3775.patch uploaded by rmudgett (License 5621) rtp_engine_fix.diff uploaded by opticron (License 6273) rtp_crash_fix.diff uploaded by opticron (License 6273) rb3753.patch uploaded by mjordan (License 6283) rb3750.patch uploaded by mjordan (License 6283) rb3748.patch uploaded by rmudgett (License 5621) media_format_fixes.diff uploaded by opticron (License 6273) rb3740.patch uploaded by mjordan (License 6283) rb3739.patch uploaded by mjordan (License 6283) rb3734.patch uploaded by mjordan (License 6283) rb3689.patch uploaded by mjordan (License 6283) rb3674.patch uploaded by coreyfarrell (License 5909) rb3671.patch uploaded by coreyfarrell (License 5909) rb3667.patch uploaded by coreyfarrell (License 5909) rb3665.patch uploaded by mjordan (License 6283) rb3625.patch uploaded by coreyfarrell (License 5909) rb3602.patch uploaded by coreyfarrell (License 5909) format_compatibility-2.diff uploaded by file (License 5000) core.diff uploaded by file (License 5000) git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@419044 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2014-07-20 22:06:33 +00:00
ast_format_cap_append(cap, ast_format_slin, 0);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
/* Create the recording channel. */
chan = ast_request("CBRec", cap, NULL, NULL, conference->name, NULL);
media formats: re-architect handling of media for performance improvements In the old times media formats were represented using a bit field. This was fast but had a few limitations. 1. Asterisk was limited in how many formats it could handle. 2. Formats, being a bit field, could not include any attribute information. A format was strictly its type, e.g., "this is ulaw". This was changed in Asterisk 10 (see https://wiki.asterisk.org/wiki/display/AST/Media+Architecture+Proposal for notes on that work) which led to the creation of the ast_format structure. This structure allowed Asterisk to handle attributes and bundle information with a format. Additionally, ast_format_cap was created to act as a container for multiple formats that, together, formed the capability of some entity. Another mechanism was added to allow logic to be registered which performed format attribute negotiation. Everywhere throughout the codebase Asterisk was changed to use this strategy. Unfortunately, in software, there is no free lunch. These new capabilities came at a cost. Performance analysis and profiling showed that we spend an inordinate amount of time comparing, copying, and generally manipulating formats and their related structures. Basic prototyping has shown that a reasonably large performance improvement could be made in this area. This patch is the result of that project, which overhauled the media format architecture and its usage in Asterisk to improve performance. Generally, the new philosophy for handling formats is as follows: * The ast_format structure is reference counted. This removed a large amount of the memory allocations and copying that was done in prior versions. * In order to prevent race conditions while keeping things performant, the ast_format structure is immutable by convention and lock-free. Violate this tenet at your peril! * Because formats are reference counted, codecs are also reference counted. The Asterisk core generally provides built-in codecs and caches the ast_format structures created to represent them. Generally, to prevent inordinate amounts of module reference bumping, codecs and formats can be added at run-time but cannot be removed. * All compatibility with the bit field representation of codecs/formats has been moved to a compatibility API. The primary user of this representation is chan_iax2, which must continue to maintain its bit-field usage of formats for interoperability concerns. * When a format is negotiated with attributes, or when a format cannot be represented by one of the cached formats, a new format object is created or cloned from an existing format. That format may have the same codec underlying it, but is a different format than a version of the format with different attributes or without attributes. * While formats are reference counted objects, the reference count maintained on the format should be manipulated with care. Formats are generally cached and will persist for the lifetime of Asterisk and do not explicitly need to have their lifetime modified. An exception to this is when the user of a format does not know where the format came from *and* the user may outlive the provider of the format. This occurs, for example, when a format is read from a channel: the channel may have a format with attributes (hence, non-cached) and the user of the format may last longer than the channel (if the reference to the channel is released prior to the format's reference). For more information on this work, see the API design notes: https://wiki.asterisk.org/wiki/display/AST/Media+Format+Rewrite Finally, this work was the culmination of a large number of developer's efforts. Extra thanks goes to Corey Farrell, who took on a large amount of the work in the Asterisk core, chan_sip, and was an invaluable resource in peer reviews throughout this project. There were a substantial number of patches contributed during this work; the following issues/patch names simply reflect some of the work (and will cause the release scripts to give attribution to the individuals who work on them). Reviews: https://reviewboard.asterisk.org/r/3814 https://reviewboard.asterisk.org/r/3808 https://reviewboard.asterisk.org/r/3805 https://reviewboard.asterisk.org/r/3803 https://reviewboard.asterisk.org/r/3801 https://reviewboard.asterisk.org/r/3798 https://reviewboard.asterisk.org/r/3800 https://reviewboard.asterisk.org/r/3794 https://reviewboard.asterisk.org/r/3793 https://reviewboard.asterisk.org/r/3792 https://reviewboard.asterisk.org/r/3791 https://reviewboard.asterisk.org/r/3790 https://reviewboard.asterisk.org/r/3789 https://reviewboard.asterisk.org/r/3788 https://reviewboard.asterisk.org/r/3787 https://reviewboard.asterisk.org/r/3786 https://reviewboard.asterisk.org/r/3784 https://reviewboard.asterisk.org/r/3783 https://reviewboard.asterisk.org/r/3778 https://reviewboard.asterisk.org/r/3774 https://reviewboard.asterisk.org/r/3775 https://reviewboard.asterisk.org/r/3772 https://reviewboard.asterisk.org/r/3761 https://reviewboard.asterisk.org/r/3754 https://reviewboard.asterisk.org/r/3753 https://reviewboard.asterisk.org/r/3751 https://reviewboard.asterisk.org/r/3750 https://reviewboard.asterisk.org/r/3748 https://reviewboard.asterisk.org/r/3747 https://reviewboard.asterisk.org/r/3746 https://reviewboard.asterisk.org/r/3742 https://reviewboard.asterisk.org/r/3740 https://reviewboard.asterisk.org/r/3739 https://reviewboard.asterisk.org/r/3738 https://reviewboard.asterisk.org/r/3737 https://reviewboard.asterisk.org/r/3736 https://reviewboard.asterisk.org/r/3734 https://reviewboard.asterisk.org/r/3722 https://reviewboard.asterisk.org/r/3713 https://reviewboard.asterisk.org/r/3703 https://reviewboard.asterisk.org/r/3689 https://reviewboard.asterisk.org/r/3687 https://reviewboard.asterisk.org/r/3674 https://reviewboard.asterisk.org/r/3671 https://reviewboard.asterisk.org/r/3667 https://reviewboard.asterisk.org/r/3665 https://reviewboard.asterisk.org/r/3625 https://reviewboard.asterisk.org/r/3602 https://reviewboard.asterisk.org/r/3519 https://reviewboard.asterisk.org/r/3518 https://reviewboard.asterisk.org/r/3516 https://reviewboard.asterisk.org/r/3515 https://reviewboard.asterisk.org/r/3512 https://reviewboard.asterisk.org/r/3506 https://reviewboard.asterisk.org/r/3413 https://reviewboard.asterisk.org/r/3410 https://reviewboard.asterisk.org/r/3387 https://reviewboard.asterisk.org/r/3388 https://reviewboard.asterisk.org/r/3389 https://reviewboard.asterisk.org/r/3390 https://reviewboard.asterisk.org/r/3321 https://reviewboard.asterisk.org/r/3320 https://reviewboard.asterisk.org/r/3319 https://reviewboard.asterisk.org/r/3318 https://reviewboard.asterisk.org/r/3266 https://reviewboard.asterisk.org/r/3265 https://reviewboard.asterisk.org/r/3234 https://reviewboard.asterisk.org/r/3178 ASTERISK-23114 #close Reported by: mjordan media_formats_translation_core.diff uploaded by kharwell (License 6464) rb3506.diff uploaded by mjordan (License 6283) media_format_app_file.diff uploaded by kharwell (License 6464) misc-2.diff uploaded by file (License 5000) chan_mild-3.diff uploaded by file (License 5000) chan_obscure.diff uploaded by file (License 5000) jingle.diff uploaded by file (License 5000) funcs.diff uploaded by file (License 5000) formats.diff uploaded by file (License 5000) core.diff uploaded by file (License 5000) bridges.diff uploaded by file (License 5000) mf-codecs-2.diff uploaded by file (License 5000) mf-app_fax.diff uploaded by file (License 5000) mf-apps-3.diff uploaded by file (License 5000) media-formats-3.diff uploaded by file (License 5000) ASTERISK-23715 rb3713.patch uploaded by coreyfarrell (License 5909) rb3689.patch uploaded by mjordan (License 6283) ASTERISK-23957 rb3722.patch uploaded by mjordan (License 6283) mf-attributes-3.diff uploaded by file (License 5000) ASTERISK-23958 Tested by: jrose rb3822.patch uploaded by coreyfarrell (License 5909) rb3800.patch uploaded by jrose (License 6182) chan_sip.diff uploaded by mjordan (License 6283) rb3747.patch uploaded by jrose (License 6182) ASTERISK-23959 #close Tested by: sgriepentrog, mjordan, coreyfarrell sip_cleanup.diff uploaded by opticron (License 6273) chan_sip_caps.diff uploaded by mjordan (License 6283) rb3751.patch uploaded by coreyfarrell (License 5909) chan_sip-3.diff uploaded by file (License 5000) ASTERISK-23960 #close Tested by: opticron direct_media.diff uploaded by opticron (License 6273) pjsip-direct-media.diff uploaded by file (License 5000) format_cap_remove.diff uploaded by opticron (License 6273) media_format_fixes.diff uploaded by opticron (License 6273) chan_pjsip-2.diff uploaded by file (License 5000) ASTERISK-23966 #close Tested by: rmudgett rb3803.patch uploaded by rmudgetti (License 5621) chan_dahdi.diff uploaded by file (License 5000) ASTERISK-24064 #close Tested by: coreyfarrell, mjordan, opticron, file, rmudgett, sgriepentrog, jrose rb3814.patch uploaded by rmudgett (License 5621) moh_cleanup.diff uploaded by opticron (License 6273) bridge_leak.diff uploaded by opticron (License 6273) translate.diff uploaded by file (License 5000) rb3795.patch uploaded by rmudgett (License 5621) tls_fix.diff uploaded by mjordan (License 6283) fax-mf-fix-2.diff uploaded by file (License 5000) rtp_transfer_stuff uploaded by mjordan (License 6283) rb3787.patch uploaded by rmudgett (License 5621) media-formats-explicit-translate-format-3.diff uploaded by file (License 5000) format_cache_case_fix.diff uploaded by opticron (License 6273) rb3774.patch uploaded by rmudgett (License 5621) rb3775.patch uploaded by rmudgett (License 5621) rtp_engine_fix.diff uploaded by opticron (License 6273) rtp_crash_fix.diff uploaded by opticron (License 6273) rb3753.patch uploaded by mjordan (License 6283) rb3750.patch uploaded by mjordan (License 6283) rb3748.patch uploaded by rmudgett (License 5621) media_format_fixes.diff uploaded by opticron (License 6273) rb3740.patch uploaded by mjordan (License 6283) rb3739.patch uploaded by mjordan (License 6283) rb3734.patch uploaded by mjordan (License 6283) rb3689.patch uploaded by mjordan (License 6283) rb3674.patch uploaded by coreyfarrell (License 5909) rb3671.patch uploaded by coreyfarrell (License 5909) rb3667.patch uploaded by coreyfarrell (License 5909) rb3665.patch uploaded by mjordan (License 6283) rb3625.patch uploaded by coreyfarrell (License 5909) rb3602.patch uploaded by coreyfarrell (License 5909) format_compatibility-2.diff uploaded by file (License 5000) core.diff uploaded by file (License 5000) git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@419044 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2014-07-20 22:06:33 +00:00
ao2_ref(cap, -1);
if (!chan) {
ast_bridge_features_destroy(features);
return -1;
}
/* Start recording. */
set_rec_filename(conference, &conference->record_filename,
is_new_rec_file(conference->b_profile.rec_file, &conference->orig_rec_file));
ast_answer(chan);
pbx_exec(chan, mixmonapp, ast_str_buffer(conference->record_filename));
/* Put the channel into the conference bridge. */
ast_channel_ref(chan);
conference->record_chan = chan;
if (ast_bridge_impart(conference->bridge, chan, NULL, features,
AST_BRIDGE_IMPART_CHAN_INDEPENDENT)) {
ast_hangup(chan);
ast_channel_unref(chan);
conference->record_chan = NULL;
return -1;
}
ast_test_suite_event_notify("CONF_START_RECORD", "Message: started conference recording channel\r\nConference: %s", conference->b_profile.name);
send_start_record_event(conference);
return 0;
}
/*! \brief Playback the given filename and monitor for any dtmf interrupts.
*
* This function is used to playback sound files on a given channel and optionally
* allow dtmf interrupts to occur.
*
* If the optional bridge_channel parameter is given then sound file playback
* is played on that channel and dtmf interruptions are allowed. However, if
* bridge_channel is not set then the channel parameter is expected to be set
* instead and non interruptible playback is played on that channel.
*
* \param bridge_channel Bridge channel to play file on
* \param channel Optional channel to play file on if bridge_channel not given
* \param filename The file name to playback
*
* \retval -1 failure during playback.
* \retval 0 on file was fully played.
* \retval 1 on dtmf interrupt.
*/
static int play_file(struct ast_bridge_channel *bridge_channel, struct ast_channel *channel,
const char *filename)
{
struct ast_channel *chan;
const char *stop_digits;
int digit;
if (bridge_channel) {
chan = bridge_channel->chan;
stop_digits = AST_DIGIT_ANY;
} else {
chan = channel;
stop_digits = AST_DIGIT_NONE;
}
digit = ast_stream_and_wait(chan, filename, stop_digits);
if (digit < 0) {
ast_log(LOG_WARNING, "Failed to playback file '%s' to channel\n", filename);
return -1;
}
if (digit > 0) {
ast_stopstream(bridge_channel->chan);
ast_bridge_channel_feature_digit_add(bridge_channel, digit);
return 1;
}
return 0;
}
/*!
* \internal
* \brief Complain if the given sound file does not exist.
*
* \param filename Sound file to check if exists.
*
* \retval non-zero if the file exists.
*/
static int sound_file_exists(const char *filename)
{
if (ast_fileexists(filename, NULL, NULL)) {
return -1;
}
ast_log(LOG_WARNING, "File %s does not exist in any format\n", filename);
return 0;
}
/*!
* \brief Announce number of users in the conference bridge to the caller
*
* \param conference Conference bridge to peek at
* \param user Optional Caller
* \param bridge_channel The bridged channel involved
*
* \note if caller is NULL, the announcment will be sent to all participants in the conference.
* \retval 0 on success.
* \retval -1 if the user hung up.
*/
static int announce_user_count(struct confbridge_conference *conference, struct confbridge_user *user,
struct ast_bridge_channel *bridge_channel)
{
const char *other_in_party = conf_get_sound(CONF_SOUND_OTHER_IN_PARTY, conference->b_profile.sounds);
const char *only_one = conf_get_sound(CONF_SOUND_ONLY_ONE, conference->b_profile.sounds);
const char *there_are = conf_get_sound(CONF_SOUND_THERE_ARE, conference->b_profile.sounds);
if (conference->activeusers <= 1) {
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
/* Awww we are the only person in the conference bridge OR we only have waitmarked users */
return 0;
} else if (conference->activeusers == 2) {
if (user) {
/* Eep, there is one other person */
if (play_file(bridge_channel, user->chan, only_one) < 0) {
return -1;
}
} else {
play_sound_file(conference, only_one);
}
} else {
/* Alas multiple others in here */
if (user) {
if (ast_stream_and_wait(user->chan,
there_are,
"")) {
return -1;
}
if (ast_say_number(user->chan, conference->activeusers - 1, "", ast_channel_language(user->chan), NULL)) {
return -1;
}
if (play_file(bridge_channel, user->chan, other_in_party) < 0) {
return -1;
}
} else if (sound_file_exists(there_are) && sound_file_exists(other_in_party)) {
play_sound_file(conference, there_are);
play_sound_number(conference, conference->activeusers - 1);
play_sound_file(conference, other_in_party);
}
}
return 0;
}
/*!
* \brief Play back an audio file to a channel
*
* \param user User to play audio prompt to
* \param filename Prompt to play
*
* \retval 0 on success.
* \retval -1 if the user hung up.
*
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
* \note Generally this should be called when the conference is unlocked to avoid blocking
* the entire conference while the sound is played. But don't unlock the conference bridge
* in the middle of a state transition.
*/
static int play_prompt_to_user(struct confbridge_user *user, const char *filename)
{
return ast_stream_and_wait(user->chan, filename, "");
}
static void handle_video_on_join(struct confbridge_conference *conference, struct ast_channel *chan, int marked)
{
/* Right now, only marked users are automatically set as the single src of video.*/
if (!marked) {
return;
}
if (ast_test_flag(&conference->b_profile, BRIDGE_OPT_VIDEO_SRC_FIRST_MARKED)) {
int set = 1;
struct confbridge_user *user = NULL;
ao2_lock(conference);
/* see if anyone is already the video src */
AST_LIST_TRAVERSE(&conference->active_list, user, list) {
if (user->chan == chan) {
continue;
}
if (ast_bridge_is_video_src(conference->bridge, user->chan)) {
set = 0;
break;
}
}
ao2_unlock(conference);
if (set) {
ast_bridge_set_single_src_video_mode(conference->bridge, chan);
}
} else if (ast_test_flag(&conference->b_profile, BRIDGE_OPT_VIDEO_SRC_LAST_MARKED)) {
/* we joined and are video capable, we override anyone else that may have already been the video feed */
ast_bridge_set_single_src_video_mode(conference->bridge, chan);
}
}
static void handle_video_on_exit(struct confbridge_conference *conference, struct ast_channel *chan)
{
struct confbridge_user *user = NULL;
/* if this isn't a video source, nothing to update */
if (!ast_bridge_is_video_src(conference->bridge, chan)) {
return;
}
ast_bridge_remove_video_src(conference->bridge, chan);
/* If in follow talker mode, make sure to restore this mode on the
* bridge when a source is removed. It is possible this channel was
* only set temporarily as a video source by an AMI or DTMF action. */
if (ast_test_flag(&conference->b_profile, BRIDGE_OPT_VIDEO_SRC_FOLLOW_TALKER)) {
ast_bridge_set_talker_src_video_mode(conference->bridge);
}
/* if the video_mode isn't set to automatically pick the video source, do nothing on exit. */
if (!ast_test_flag(&conference->b_profile, BRIDGE_OPT_VIDEO_SRC_FIRST_MARKED) &&
!ast_test_flag(&conference->b_profile, BRIDGE_OPT_VIDEO_SRC_LAST_MARKED)) {
return;
}
/* Make the next available marked user the video src. */
ao2_lock(conference);
AST_LIST_TRAVERSE(&conference->active_list, user, list) {
if (user->chan == chan) {
continue;
}
if (ast_test_flag(&user->u_profile, USER_OPT_MARKEDUSER)) {
ast_bridge_set_single_src_video_mode(conference->bridge, user->chan);
break;
}
}
ao2_unlock(conference);
}
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
struct hangup_data
{
struct confbridge_conference *conference;
ast_mutex_t lock;
ast_cond_t cond;
int hungup;
};
/*!
* \brief Hang up the announcer channel
*
* This hangs up the announcer channel in the conference. This
* runs in the playback queue taskprocessor since we do not want
* to hang up the channel while it's trying to play an announcement.
*
* This task is performed synchronously, so there is no need to
* perform any cleanup on the passed-in data.
*
* \param data A hangup_data structure
* \return 0
*/
static int hangup_playback(void *data)
{
struct hangup_data *hangup = data;
ast_autoservice_stop(hangup->conference->playback_chan);
ast_hangup(hangup->conference->playback_chan);
hangup->conference->playback_chan = NULL;
ast_mutex_lock(&hangup->lock);
hangup->hungup = 1;
ast_cond_signal(&hangup->cond);
ast_mutex_unlock(&hangup->lock);
return 0;
}
static void hangup_data_init(struct hangup_data *hangup, struct confbridge_conference *conference)
{
ast_mutex_init(&hangup->lock);
ast_cond_init(&hangup->cond, NULL);
hangup->conference = conference;
hangup->hungup = 0;
}
static void hangup_data_destroy(struct hangup_data *hangup)
{
ast_mutex_destroy(&hangup->lock);
ast_cond_destroy(&hangup->cond);
}
/*!
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
* \brief Destroy a conference bridge
*
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
* \param obj The conference bridge object
*/
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
static void destroy_conference_bridge(void *obj)
{
struct confbridge_conference *conference = obj;
ast_debug(1, "Destroying conference bridge '%s'\n", conference->name);
if (conference->playback_chan) {
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
if (conference->playback_queue) {
struct hangup_data hangup;
hangup_data_init(&hangup, conference);
if (!ast_taskprocessor_push(conference->playback_queue, hangup_playback, &hangup)) {
ast_mutex_lock(&hangup.lock);
while (!hangup.hungup) {
ast_cond_wait(&hangup.cond, &hangup.lock);
}
ast_mutex_unlock(&hangup.lock);
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
}
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
hangup_data_destroy(&hangup);
} else {
/* Playback queue is not yet allocated. Just hang up the channel straight */
ast_hangup(conference->playback_chan);
conference->playback_chan = NULL;
}
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
}
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
/* Destroying a conference bridge is simple, all we have to do is destroy the bridging object */
if (conference->bridge) {
ast_bridge_destroy(conference->bridge, 0);
conference->bridge = NULL;
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
}
ast_channel_cleanup(conference->record_chan);
ast_free(conference->orig_rec_file);
ast_free(conference->record_filename);
conf_bridge_profile_destroy(&conference->b_profile);
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
ast_taskprocessor_unreference(conference->playback_queue);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
}
/*! \brief Call the proper join event handler for the user for the conference bridge's current state
* \internal
* \param user The conference bridge user that is joining
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
* \retval 0 success
* \retval -1 failure
*/
static int handle_conf_user_join(struct confbridge_user *user)
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
{
conference_event_fn handler;
if (ast_test_flag(&user->u_profile, USER_OPT_MARKEDUSER)) {
handler = user->conference->state->join_marked;
} else if (ast_test_flag(&user->u_profile, USER_OPT_WAITMARKED)) {
handler = user->conference->state->join_waitmarked;
} else {
handler = user->conference->state->join_unmarked;
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
}
ast_assert(handler != NULL);
if (!handler) {
conf_invalid_event_fn(user);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
return -1;
}
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
handler(user);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
return 0;
}
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
/*! \brief Call the proper leave event handler for the user for the conference bridge's current state
* \internal
* \param user The conference bridge user that is leaving
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
* \retval 0 success
* \retval -1 failure
*/
static int handle_conf_user_leave(struct confbridge_user *user)
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
{
conference_event_fn handler;
if (ast_test_flag(&user->u_profile, USER_OPT_MARKEDUSER)) {
handler = user->conference->state->leave_marked;
} else if (ast_test_flag(&user->u_profile, USER_OPT_WAITMARKED)) {
handler = user->conference->state->leave_waitmarked;
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
} else {
handler = user->conference->state->leave_unmarked;
}
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
ast_assert(handler != NULL);
if (!handler) {
/* This should never happen. If it does, though, it is bad. The user will not have been removed
* from the appropriate list, so counts will be off and stuff. The conference won't be torn down, etc.
* Shouldn't happen, though. */
conf_invalid_event_fn(user);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
return -1;
}
handler(user);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
return 0;
}
void conf_update_user_mute(struct confbridge_user *user)
{
int mute_user;
int mute_system;
int mute_effective;
/* User level mute request. */
mute_user = user->muted;
/* System level mute request. */
mute_system = user->playing_moh
/*
* Do not allow waitmarked users to talk to anyone unless there
* is a marked user present.
*/
|| (!user->conference->markedusers
&& ast_test_flag(&user->u_profile, USER_OPT_WAITMARKED));
mute_effective = mute_user || mute_system;
ast_debug(1, "User %s is %s: user:%d system:%d.\n",
ast_channel_name(user->chan), mute_effective ? "muted" : "unmuted",
mute_user, mute_system);
user->features.mute = mute_effective;
ast_test_suite_event_notify("CONF_MUTE_UPDATE",
"Mode: %s\r\n"
"Conference: %s\r\n"
"Channel: %s",
mute_effective ? "muted" : "unmuted",
user->conference->b_profile.name,
ast_channel_name(user->chan));
}
/*!
* \internal
* \brief Mute/unmute a single user.
*/
static void generic_mute_unmute_user(struct confbridge_conference *conference, struct confbridge_user *user, int mute)
{
/* Set user level mute request. */
user->muted = mute ? 1 : 0;
conf_update_user_mute(user);
ast_test_suite_event_notify("CONF_MUTE",
"Message: participant %s %s\r\n"
"Conference: %s\r\n"
"Channel: %s",
ast_channel_name(user->chan),
mute ? "muted" : "unmuted",
conference->b_profile.name,
ast_channel_name(user->chan));
if (mute) {
send_mute_event(user, conference);
} else {
send_unmute_event(user, conference);
}
}
void conf_moh_stop(struct confbridge_user *user)
{
user->playing_moh = 0;
if (!user->suspended_moh) {
int in_bridge;
/*
* Locking the ast_bridge here is the only way to hold off the
* call to ast_bridge_join() in confbridge_exec() from
* interfering with the bridge and MOH operations here.
*/
ast_bridge_lock(user->conference->bridge);
/*
* Temporarily suspend the user from the bridge so we have
* control to stop MOH if needed.
*/
in_bridge = !ast_bridge_suspend(user->conference->bridge, user->chan);
ast_moh_stop(user->chan);
if (in_bridge) {
ast_bridge_unsuspend(user->conference->bridge, user->chan);
}
ast_bridge_unlock(user->conference->bridge);
}
}
void conf_moh_start(struct confbridge_user *user)
{
user->playing_moh = 1;
if (!user->suspended_moh) {
int in_bridge;
/*
* Locking the ast_bridge here is the only way to hold off the
* call to ast_bridge_join() in confbridge_exec() from
* interfering with the bridge and MOH operations here.
*/
ast_bridge_lock(user->conference->bridge);
/*
* Temporarily suspend the user from the bridge so we have
* control to start MOH if needed.
*/
in_bridge = !ast_bridge_suspend(user->conference->bridge, user->chan);
ast_moh_start(user->chan, user->u_profile.moh_class, NULL);
if (in_bridge) {
ast_bridge_unsuspend(user->conference->bridge, user->chan);
}
ast_bridge_unlock(user->conference->bridge);
}
}
/*!
* \internal
* \brief Unsuspend MOH for the conference user.
*
* \param user Conference user to unsuspend MOH on.
*/
static void conf_moh_unsuspend(struct confbridge_user *user)
{
ao2_lock(user->conference);
if (--user->suspended_moh == 0 && user->playing_moh) {
ast_moh_start(user->chan, user->u_profile.moh_class, NULL);
}
ao2_unlock(user->conference);
}
/*!
* \internal
* \brief Suspend MOH for the conference user.
*
* \param user Conference user to suspend MOH on.
*/
static void conf_moh_suspend(struct confbridge_user *user)
{
ao2_lock(user->conference);
if (user->suspended_moh++ == 0 && user->playing_moh) {
ast_moh_stop(user->chan);
}
ao2_unlock(user->conference);
}
int conf_handle_inactive_waitmarked(struct confbridge_user *user)
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
{
/* If we have not been quieted play back that they are waiting for the leader */
if (!ast_test_flag(&user->u_profile, USER_OPT_QUIET) && play_prompt_to_user(user,
conf_get_sound(CONF_SOUND_WAIT_FOR_LEADER, user->conference->b_profile.sounds))) {
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
/* user hungup while the sound was playing */
return -1;
}
return 0;
}
int conf_handle_only_person(struct confbridge_user *user)
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
{
/* If audio prompts have not been quieted or this prompt quieted play it on out */
if (!ast_test_flag(&user->u_profile, USER_OPT_QUIET | USER_OPT_NOONLYPERSON)) {
if (play_prompt_to_user(user,
conf_get_sound(CONF_SOUND_ONLY_PERSON, user->conference->b_profile.sounds))) {
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
/* user hungup while the sound was playing */
return -1;
}
}
return 0;
}
int conf_add_post_join_action(struct confbridge_user *user, int (*func)(struct confbridge_user *user))
{
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
struct post_join_action *action;
if (!(action = ast_calloc(1, sizeof(*action)))) {
return -1;
}
action->func = func;
AST_LIST_INSERT_TAIL(&user->post_join_list, action, list);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
return 0;
}
void conf_handle_first_join(struct confbridge_conference *conference)
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
{
ast_devstate_changed(AST_DEVICE_INUSE, AST_DEVSTATE_CACHABLE, "confbridge:%s", conference->name);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
}
void conf_handle_second_active(struct confbridge_conference *conference)
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
{
/* If we are the second participant we may need to stop music on hold on the first */
struct confbridge_user *first_user = AST_LIST_FIRST(&conference->active_list);
if (ast_test_flag(&first_user->u_profile, USER_OPT_MUSICONHOLD)) {
conf_moh_stop(first_user);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
}
conf_update_user_mute(first_user);
}
void conf_ended(struct confbridge_conference *conference)
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
{
struct pbx_find_info q = { .stacklen = 0 };
/* Called with a reference to conference */
ao2_unlink(conference_bridges, conference);
send_conf_end_event(conference);
if (!ast_strlen_zero(conference->b_profile.regcontext) &&
pbx_find_extension(NULL, NULL, &q, conference->b_profile.regcontext,
conference->name, 1, NULL, "", E_MATCH)) {
ast_context_remove_extension(conference->b_profile.regcontext,
conference->name, 1, NULL);
}
ao2_lock(conference);
conf_stop_record(conference);
ao2_unlock(conference);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
}
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
/*!
* \internal
* \brief Allocate playback channel for a conference.
* \pre expects conference to be locked before calling this function
*/
static int alloc_playback_chan(struct confbridge_conference *conference)
{
struct ast_format_cap *cap;
char taskprocessor_name[AST_TASKPROCESSOR_MAX_NAME + 1];
cap = ast_format_cap_alloc(AST_FORMAT_CAP_FLAG_DEFAULT);
if (!cap) {
return -1;
}
ast_format_cap_append(cap, ast_format_slin, 0);
conference->playback_chan = ast_request("CBAnn", cap, NULL, NULL,
conference->name, NULL);
ao2_ref(cap, -1);
if (!conference->playback_chan) {
return -1;
}
/* To make sure playback_chan has the same language as the bridge */
ast_channel_lock(conference->playback_chan);
ast_channel_language_set(conference->playback_chan, conference->b_profile.language);
ast_channel_unlock(conference->playback_chan);
ast_debug(1, "Created announcer channel '%s' to conference bridge '%s'\n",
ast_channel_name(conference->playback_chan), conference->name);
ast_taskprocessor_build_name(taskprocessor_name, sizeof(taskprocessor_name),
"Confbridge/%s", conference->name);
conference->playback_queue = ast_taskprocessor_get(taskprocessor_name, TPS_REF_DEFAULT);
if (!conference->playback_queue) {
ast_hangup(conference->playback_chan);
conference->playback_chan = NULL;
return -1;
}
return 0;
}
/*!
* \brief Push the announcer channel into the bridge
*
* \param conference Conference bridge to push the announcer to
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
* \retval 0 Success
* \retval -1 Failed to push the channel to the bridge
*/
static int push_announcer(struct confbridge_conference *conference)
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
{
if (conf_announce_channel_push(conference->playback_chan)) {
ast_hangup(conference->playback_chan);
conference->playback_chan = NULL;
return -1;
}
ast_autoservice_start(conference->playback_chan);
return 0;
}
static void confbridge_unlock_and_unref(void *obj)
{
struct confbridge_conference *conference = obj;
if (!obj) {
return;
}
ao2_unlock(conference);
ao2_ref(conference, -1);
}
void confbridge_handle_atxfer(struct ast_attended_transfer_message *msg)
{
struct ast_channel_snapshot *old_snapshot;
struct ast_channel_snapshot *new_snapshot;
char *confbr_name = NULL;
char *comma;
RAII_VAR(struct confbridge_conference *, conference, NULL, confbridge_unlock_and_unref);
struct confbridge_user *user = NULL;
int found_user = 0;
struct ast_json *json_object;
if (msg->to_transferee.channel_snapshot
&& strcmp(msg->to_transferee.channel_snapshot->dialplan->appl, "ConfBridge") == 0
&& msg->target) {
/* We're transferring a bridge to an extension */
old_snapshot = msg->to_transferee.channel_snapshot;
new_snapshot = msg->target;
} else if (msg->to_transfer_target.channel_snapshot
&& strcmp(msg->to_transfer_target.channel_snapshot->dialplan->appl, "ConfBridge") == 0
&& msg->transferee) {
/* We're transferring a call to a bridge */
old_snapshot = msg->to_transfer_target.channel_snapshot;
new_snapshot = msg->transferee;
} else {
ast_log(LOG_ERROR, "Could not determine proper channels\n");
return;
}
/*
* old_snapshot->data should have the original parameters passed to
* the ConfBridge app:
* conference[,bridge_profile[,user_profile[,menu]]]
* We'll use "conference" to look up the bridge.
*
* We _could_ use old_snapshot->bridgeid to get the bridge but
* that would involve locking the conference_bridges container
* and iterating over it looking for a matching bridge.
*/
if (ast_strlen_zero(old_snapshot->dialplan->data)) {
ast_log(LOG_ERROR, "Channel '%s' didn't have app data set\n", old_snapshot->base->name);
return;
}
confbr_name = ast_strdupa(old_snapshot->dialplan->data);
comma = strchr(confbr_name, ',');
if (comma) {
*comma = '\0';
}
ast_debug(1, "Confbr: %s Leaving: %s Joining: %s\n", confbr_name, old_snapshot->base->name, new_snapshot->base->name);
conference = ao2_find(conference_bridges, confbr_name, OBJ_SEARCH_KEY);
if (!conference) {
ast_log(LOG_ERROR, "Conference bridge '%s' not found\n", confbr_name);
return;
}
ao2_lock(conference);
/*
* We need to grab the user profile for the departing user in order to
* properly format the join/leave messages.
*/
AST_LIST_TRAVERSE(&conference->active_list, user, list) {
if (strcasecmp(ast_channel_name(user->chan), old_snapshot->base->name) == 0) {
found_user = 1;
break;
}
}
/*
* If we didn't find the user in the active list, try the waiting list.
*/
if (!found_user && conference->waitingusers) {
AST_LIST_TRAVERSE(&conference->waiting_list, user, list) {
if (strcasecmp(ast_channel_name(user->chan), old_snapshot->base->name) == 0) {
found_user = 1;
break;
}
}
}
if (!found_user) {
ast_log(LOG_ERROR, "Unable to find user profile for channel '%s' in bridge '%s'\n",
old_snapshot->base->name, confbr_name);
return;
}
/*
* We're going to use the existing user profile to create the messages.
*/
json_object = ast_json_pack("{s: b}",
"admin", ast_test_flag(&user->u_profile, USER_OPT_ADMIN)
);
if (!json_object) {
return;
}
send_conf_stasis_snapshots(conference, old_snapshot, confbridge_leave_type(), json_object);
ast_json_unref(json_object);
json_object = ast_json_pack("{s: b, s: b}",
"admin", ast_test_flag(&user->u_profile, USER_OPT_ADMIN),
"muted", user->muted);
if (!json_object) {
return;
}
send_conf_stasis_snapshots(conference, new_snapshot, confbridge_join_type(), json_object);
ast_json_unref(json_object);
}
/*!
* \brief Join a conference bridge
*
* \param conference_name The conference name
* \param user Conference bridge user structure
*
* \return A pointer to the conference bridge struct, or NULL if the conference room wasn't found.
*/
static struct confbridge_conference *join_conference_bridge(const char *conference_name, struct confbridge_user *user)
{
struct confbridge_conference *conference;
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
struct post_join_action *action;
int max_members_reached = 0;
/* We explicitly lock the conference bridges container ourselves so that other callers can not create duplicate conferences at the same time */
ao2_lock(conference_bridges);
ast_debug(1, "Trying to find conference bridge '%s'\n", conference_name);
/* Attempt to find an existing conference bridge */
conference = ao2_find(conference_bridges, conference_name, OBJ_KEY);
if (conference && conference->b_profile.max_members) {
max_members_reached = conference->b_profile.max_members > conference->activeusers ? 0 : 1;
}
/* When finding a conference bridge that already exists make sure that it is not locked, and if so that we are not an admin */
if (conference && (max_members_reached || conference->locked) && !ast_test_flag(&user->u_profile, USER_OPT_ADMIN)) {
ao2_unlock(conference_bridges);
ast_debug(1, "Conference '%s' is locked and caller is not an admin\n", conference_name);
ast_stream_and_wait(user->chan,
conf_get_sound(CONF_SOUND_LOCKED, conference->b_profile.sounds),
"");
ao2_ref(conference, -1);
return NULL;
}
/* If no conference bridge was found see if we can create one */
if (!conference) {
/* Try to allocate memory for a new conference bridge, if we fail... this won't end well. */
if (!(conference = ao2_alloc(sizeof(*conference), destroy_conference_bridge))) {
ao2_unlock(conference_bridges);
ast_log(LOG_ERROR, "Conference '%s' could not be created.\n", conference_name);
return NULL;
}
/* Setup for the record channel */
conference->record_filename = ast_str_create(RECORD_FILENAME_INITIAL_SPACE);
if (!conference->record_filename) {
ao2_ref(conference, -1);
ao2_unlock(conference_bridges);
return NULL;
}
/* Setup conference bridge parameters */
ast_copy_string(conference->name, conference_name, sizeof(conference->name));
conf_bridge_profile_copy(&conference->b_profile, &user->b_profile);
/* Create an actual bridge that will do the audio mixing */
conference->bridge = ast_bridge_base_new(AST_BRIDGE_CAPABILITY_MULTIMIX,
AST_BRIDGE_FLAG_MASQUERADE_ONLY | AST_BRIDGE_FLAG_TRANSFER_BRIDGE_ONLY,
app, conference_name, NULL);
if (!conference->bridge) {
ao2_ref(conference, -1);
ao2_unlock(conference_bridges);
ast_log(LOG_ERROR, "Conference '%s' mixing bridge could not be created.\n", conference_name);
return NULL;
}
/* Set the internal sample rate on the bridge from the bridge profile */
ast_bridge_set_internal_sample_rate(conference->bridge, conference->b_profile.internal_sample_rate);
/* Set the maximum sample rate on the bridge from the bridge profile */
ast_bridge_set_maximum_sample_rate(conference->bridge, conference->b_profile.maximum_sample_rate);
/* Set the internal mixing interval on the bridge from the bridge profile */
ast_bridge_set_mixing_interval(conference->bridge, conference->b_profile.mix_interval);
ast_bridge_set_binaural_active(conference->bridge, ast_test_flag(&conference->b_profile, BRIDGE_OPT_BINAURAL_ACTIVE));
if (ast_test_flag(&conference->b_profile, BRIDGE_OPT_VIDEO_SRC_FOLLOW_TALKER)) {
ast_bridge_set_talker_src_video_mode(conference->bridge);
} else if (ast_test_flag(&conference->b_profile, BRIDGE_OPT_VIDEO_SRC_SFU)) {
ast_bridge_set_sfu_video_mode(conference->bridge);
ast_bridge_set_remb_send_interval(conference->bridge, conference->b_profile.remb_send_interval);
if (ast_test_flag(&conference->b_profile, BRIDGE_OPT_REMB_BEHAVIOR_AVERAGE)) {
ast_brige_set_remb_behavior(conference->bridge, AST_BRIDGE_VIDEO_SFU_REMB_AVERAGE);
} else if (ast_test_flag(&conference->b_profile, BRIDGE_OPT_REMB_BEHAVIOR_LOWEST)) {
ast_brige_set_remb_behavior(conference->bridge, AST_BRIDGE_VIDEO_SFU_REMB_LOWEST);
} else if (ast_test_flag(&conference->b_profile, BRIDGE_OPT_REMB_BEHAVIOR_HIGHEST)) {
ast_brige_set_remb_behavior(conference->bridge, AST_BRIDGE_VIDEO_SFU_REMB_HIGHEST);
} else if (ast_test_flag(&conference->b_profile, BRIDGE_OPT_REMB_BEHAVIOR_AVERAGE_ALL)) {
ast_brige_set_remb_behavior(conference->bridge, AST_BRIDGE_VIDEO_SFU_REMB_AVERAGE_ALL);
} else if (ast_test_flag(&conference->b_profile, BRIDGE_OPT_REMB_BEHAVIOR_LOWEST_ALL)) {
ast_brige_set_remb_behavior(conference->bridge, AST_BRIDGE_VIDEO_SFU_REMB_LOWEST_ALL);
} else if (ast_test_flag(&conference->b_profile, BRIDGE_OPT_REMB_BEHAVIOR_HIGHEST_ALL)) {
ast_brige_set_remb_behavior(conference->bridge, AST_BRIDGE_VIDEO_SFU_REMB_HIGHEST_ALL);
} else if (ast_test_flag(&conference->b_profile, BRIDGE_OPT_REMB_BEHAVIOR_FORCE)) {
ast_brige_set_remb_behavior(conference->bridge, AST_BRIDGE_VIDEO_SFU_REMB_FORCE);
ast_bridge_set_remb_estimated_bitrate(conference->bridge, conference->b_profile.remb_estimated_bitrate);
}
}
app_confbridge: Always set minimum video update interval. Currently, if multiple video-enabled ConfBridges are conferenced together, we immediately get into a scenario where an infinite sequence of video updates fills up the taskprocessor queue and causes memory consumption to climb unabated until Asterisk is killed. This is due to the core bridging mechanism that provides video updates (softmix_bridge_write_control in bridge_softmix.c) continously updating all the channels in the bridge with video updates. The logic to do so in the core is that the video updates should be provided if the video_update_discard property for the bridge is 0, or if enough time has elapsed since the last video update. Thus, we already have a safeguard built in to ensure the scenario described above does not happen. Currently, however, this safeguard is not being adequately ensured. In app_confbridge, the video_update_discard property defaults to 2000, which is a healthy value that should completely prevent this issue. However, this value is only set onto the bridge in the SFU video mode. This leaves video modes such as follow_talker completely vulnerable, since video_update_discard will actually be 0, since the default or set value was never applied. As a result, the core bridging mechanism will always try to provide video updates regardless of when the last one was sent. To prevent this issue from happening, we now always set the video_update_discard property on the bridge with the value from the bridge profile. The app_confbridge defaults will thus ensure that infinite video updates no longer happen in any video mode. ASTERISK-29907 #close Change-Id: I4accb2536ac62797950468e9930f12ef7dd486b2
2022-06-18 12:17:50 +00:00
/* Always set the minimum interval between video updates, to avoid infinite video updates. */
ast_bridge_set_video_update_discard(conference->bridge, conference->b_profile.video_update_discard);
if (ast_test_flag(&conference->b_profile, BRIDGE_OPT_ENABLE_EVENTS)) {
ast_bridge_set_send_sdp_label(conference->bridge, 1);
}
/* Link it into the conference bridges container */
if (!ao2_link(conference_bridges, conference)) {
ao2_ref(conference, -1);
ao2_unlock(conference_bridges);
ast_log(LOG_ERROR,
"Conference '%s' could not be added to the conferences list.\n", conference_name);
return NULL;
}
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
/* Set the initial state to EMPTY */
conference->state = CONF_STATE_EMPTY;
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
if (alloc_playback_chan(conference)) {
ao2_unlink(conference_bridges, conference);
ao2_ref(conference, -1);
ao2_unlock(conference_bridges);
ast_log(LOG_ERROR, "Could not allocate announcer channel for conference '%s'\n", conference_name);
return NULL;
}
if (push_announcer(conference)) {
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
ao2_unlink(conference_bridges, conference);
ao2_ref(conference, -1);
ao2_unlock(conference_bridges);
ast_log(LOG_ERROR, "Could not add announcer channel for conference '%s' bridge\n", conference_name);
return NULL;
}
if (ast_test_flag(&conference->b_profile, BRIDGE_OPT_RECORD_CONFERENCE)) {
ao2_lock(conference);
conf_start_record(conference);
ao2_unlock(conference);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
}
send_conf_start_event(conference);
if (!ast_strlen_zero(conference->b_profile.regcontext)) {
if (!ast_exists_extension(NULL, conference->b_profile.regcontext, conference->name, 1, NULL)) {
ast_add_extension(conference->b_profile.regcontext, 1, conference->name, 1, NULL, NULL, "Noop", NULL, NULL, "ConfBridge");
}
}
ast_debug(1, "Created conference '%s' and linked to container.\n", conference_name);
}
ao2_unlock(conference_bridges);
/* Setup conference bridge user parameters */
user->conference = conference;
ao2_lock(conference);
/* Determine if the new user should join the conference muted. */
if (ast_test_flag(&user->u_profile, USER_OPT_STARTMUTED)
|| (!ast_test_flag(&user->u_profile, USER_OPT_ADMIN) && conference->muted)) {
/* Set user level mute request. */
user->muted = 1;
}
/*
* Suspend any MOH until the user actually joins the bridge of
* the conference. This way any pre-join file playback does not
* need to worry about MOH.
*/
user->suspended_moh = 1;
if (handle_conf_user_join(user)) {
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
/* Invalid event, nothing was done, so we don't want to process a leave. */
ao2_unlock(conference);
ao2_ref(conference, -1);
user->conference = NULL;
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
return NULL;
}
if (ast_check_hangup(user->chan)) {
ao2_unlock(conference);
leave_conference(user);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
return NULL;
}
ao2_unlock(conference);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
/* If an announcement is to be played play it */
if (!ast_strlen_zero(user->u_profile.announcement)) {
if (play_prompt_to_user(user,
user->u_profile.announcement)) {
leave_conference(user);
return NULL;
}
}
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
/* Announce number of users if need be */
if (ast_test_flag(&user->u_profile, USER_OPT_ANNOUNCEUSERCOUNT)) {
if (announce_user_count(conference, user, NULL)) {
leave_conference(user);
return NULL;
}
}
if (ast_test_flag(&user->u_profile, USER_OPT_ANNOUNCEUSERCOUNTALL) &&
(conference->activeusers > user->u_profile.announce_user_count_all_after)) {
int user_count_res;
/*
* We have to autoservice the new user because he has not quite
* joined the conference yet.
*/
ast_autoservice_start(user->chan);
user_count_res = announce_user_count(conference, NULL, NULL);
ast_autoservice_stop(user->chan);
if (user_count_res) {
leave_conference(user);
return NULL;
}
}
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
/* Handle post-join actions */
while ((action = AST_LIST_REMOVE_HEAD(&user->post_join_list, list))) {
action->func(user);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
ast_free(action);
}
return conference;
}
/*!
* \brief Leave a conference
*
* \param user The conference user
*/
static void leave_conference(struct confbridge_user *user)
{
struct post_join_action *action;
ao2_lock(user->conference);
handle_conf_user_leave(user);
ao2_unlock(user->conference);
/* Discard any post-join actions */
while ((action = AST_LIST_REMOVE_HEAD(&user->post_join_list, list))) {
ast_free(action);
}
/* Done mucking with the conference, huzzah */
ao2_ref(user->conference, -1);
user->conference = NULL;
}
static void playback_common(struct confbridge_conference *conference, const char *filename, int say_number)
{
/* Don't try to play if the playback channel has been hung up */
if (!conference->playback_chan) {
return;
}
ast_autoservice_stop(conference->playback_chan);
/* The channel is all under our control, in goes the prompt */
if (!ast_strlen_zero(filename)) {
ast_stream_and_wait(conference->playback_chan, filename, "");
} else if (say_number >= 0) {
ast_say_number(conference->playback_chan, say_number, "",
ast_channel_language(conference->playback_chan), NULL);
}
ast_autoservice_start(conference->playback_chan);
}
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
struct playback_task_data {
struct confbridge_conference *conference;
const char *filename;
int say_number;
int playback_finished;
ast_mutex_t lock;
ast_cond_t cond;
};
/*!
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
* \brief Play an announcement into a confbridge
*
* This runs in the playback queue taskprocessor. This ensures that
* all playbacks are handled in sequence and do not play over top one
* another.
*
* This task runs synchronously so there is no need for performing any
* sort of cleanup on the input parameter.
*
* \param data A playback_task_data
* \return 0
*/
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
static int playback_task(void *data)
{
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
struct playback_task_data *ptd = data;
playback_common(ptd->conference, ptd->filename, ptd->say_number);
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
ast_mutex_lock(&ptd->lock);
ptd->playback_finished = 1;
ast_cond_signal(&ptd->cond);
ast_mutex_unlock(&ptd->lock);
return 0;
}
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
static void playback_task_data_init(struct playback_task_data *ptd, struct confbridge_conference *conference,
const char *filename, int say_number)
{
ast_mutex_init(&ptd->lock);
ast_cond_init(&ptd->cond, NULL);
ptd->filename = filename;
ptd->say_number = say_number;
ptd->conference = conference;
ptd->playback_finished = 0;
}
static void playback_task_data_destroy(struct playback_task_data *ptd)
{
ast_mutex_destroy(&ptd->lock);
ast_cond_destroy(&ptd->cond);
}
static int play_sound_helper(struct confbridge_conference *conference, const char *filename, int say_number)
{
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
struct playback_task_data ptd;
/* Do not waste resources trying to play files that do not exist */
if (ast_strlen_zero(filename)) {
if (say_number < 0) {
return 0;
}
} else if (!sound_file_exists(filename)) {
return 0;
}
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
playback_task_data_init(&ptd, conference, filename, say_number);
if (ast_taskprocessor_push(conference->playback_queue, playback_task, &ptd)) {
if (!ast_strlen_zero(filename)) {
ast_log(LOG_WARNING, "Unable to play file '%s' to conference %s\n",
filename, conference->name);
} else {
ast_log(LOG_WARNING, "Unable to say number '%d' to conference %s\n",
say_number, conference->name);
}
playback_task_data_destroy(&ptd);
return -1;
}
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
/* Wait for the playback to complete */
ast_mutex_lock(&ptd.lock);
while (!ptd.playback_finished) {
ast_cond_wait(&ptd.cond, &ptd.lock);
}
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
ast_mutex_unlock(&ptd.lock);
ConfBridge: Rework announcer channel methodology NOTE: This patch was submitted earlier and reverted because of a failing test. The test has been patched so that it adjusts for the changes here, so this is being resubmitted for review. One feature that confbridge has is the ability to play sounds to all participants in the conference. Prior to this commit, the algorithm for this was as follows: * Grab the playback lock * Push the conference announcer channel into the bridge * Play back the sound * Pull the conference announcer channel from the bridge * Release the playback lock The issue here is that the act of adding the playback channel to the bridge and removing it for each announcement is expensive. Amongst the expenses: * The announcer channel is imparted into the bridge, meaning a new thread is spun up for each playback. * When the announcer is added or removed from the bridge, it results in the BRIDGEPEER channel variable being set on all channels in the bridge. This requires keeping the bridge locked and locking each individual channel in order to set it. * There's also just the general overhead of adding the channel and removing it from the bridge. The bridge potentially has to reconfigure every single time With this commit, the paradigm for playing back announcements has shifted. * The announcer channel is now added to the bridge when the conference is allocated, and it is hung up when the conference is destroyed. * A taskprocessor is used to queue playbacks onto the announcer channel. This keeps the behavior from before where playbacks do not overlap. * The announcer channel is no longer placed into the bridge as departable. Since we are not constantly removing the channel from the bridge, it is safe to add the channel using an independent thread and simply hang the channel up when it is time for the conference to be destroyed. The use of the taskprocessor for playbacks opens up the interesting possibility of having asynchronous announcements played. In this commit, however, the behavior is still exactly the same as it previously was. ASTERISK-26289 Reported by Mark Michelson Change-Id: Ica9fa4907c2f3728cdd1cf0bc564ef4eb40754a0
2016-08-10 20:14:09 +00:00
playback_task_data_destroy(&ptd);
return 0;
}
int play_sound_file(struct confbridge_conference *conference, const char *filename)
{
return play_sound_helper(conference, filename, -1);
}
struct async_playback_task_data {
struct confbridge_conference *conference;
int say_number;
struct ast_channel *initiator;
char filename[0];
};
struct async_datastore_data {
ast_mutex_t lock;
ast_cond_t cond;
int wait;
};
static void async_datastore_data_destroy(void *data)
{
struct async_datastore_data *add = data;
ast_mutex_destroy(&add->lock);
ast_cond_destroy(&add->cond);
ast_free(add);
}
/*!
* \brief Datastore used for timing of async announcement playback
*
* Announcements that are played to the entire conference can be played
* asynchronously (i.e. The channel that queues the playback does not wait
* for the playback to complete before continuing)
*
* The thing about async announcements is that the channel that queues the
* announcement is either not in the bridge or is in some other way "occupied"
* at the time the announcement is queued. Because of that, the initiator of
* the announcement may enter after the announcement has already started,
* resulting in the sound being "clipped".
*
* This datastore makes it so that the channel that queues the async announcement
* can say "I'm ready now". This way the announcement does not start until the
* initiator of the announcement is ready to hear the sound.
*/
static struct ast_datastore_info async_datastore_info = {
.type = "Confbridge async playback",
.destroy = async_datastore_data_destroy,
};
static struct async_datastore_data *async_datastore_data_alloc(void)
{
struct async_datastore_data *add;
add = ast_malloc(sizeof(*add));
if (!add) {
return NULL;
}
ast_mutex_init(&add->lock);
ast_cond_init(&add->cond, NULL);
add->wait = 1;
return add;
}
/*!
* \brief Prepare the async playback datastore
*
* This is done prior to queuing an async announcement. If the
* datastore has not yet been created, it is allocated and initialized.
* If it already exists, we set it to be in "waiting" mode.
*
* \param initiator The channel that is queuing the async playback
* \retval 0 Success
* \retval -1 Failure :(
*/
static int setup_async_playback_datastore(struct ast_channel *initiator)
{
struct ast_datastore *async_datastore;
async_datastore = ast_channel_datastore_find(initiator, &async_datastore_info, NULL);
if (async_datastore) {
struct async_datastore_data *add;
add = async_datastore->data;
add->wait = 1;
return 0;
}
async_datastore = ast_datastore_alloc(&async_datastore_info, NULL);
if (!async_datastore) {
return -1;
}
async_datastore->data = async_datastore_data_alloc();
if (!async_datastore->data) {
ast_datastore_free(async_datastore);
return -1;
}
ast_channel_datastore_add(initiator, async_datastore);
return 0;
}
static struct async_playback_task_data *async_playback_task_data_alloc(
struct confbridge_conference *conference, const char *filename, int say_number,
struct ast_channel *initiator)
{
struct async_playback_task_data *aptd;
aptd = ast_malloc(sizeof(*aptd) + strlen(filename) + 1);
if (!aptd) {
return NULL;
}
/* Safe */
strcpy(aptd->filename, filename);
aptd->say_number = say_number;
/* You may think that we need to bump the conference refcount since we are pushing
* this task to the taskprocessor.
*
* In this case, that actually causes a problem. The destructor for the conference
* pushes a hangup task into the taskprocessor and waits for it to complete before
* continuing. If the destructor gets called from a taskprocessor task, we're
* deadlocked.
*
* So is there a risk of the conference being freed out from under us? No. Since
* the destructor pushes a task into the taskprocessor and waits for it to complete,
* the destructor cannot free the conference out from under us. No further tasks
* can be queued onto the taskprocessor after the hangup since no channels are referencing
* the conference at that point any more.
*/
aptd->conference = conference;
aptd->initiator = initiator;
if (initiator) {
ast_channel_ref(initiator);
ast_channel_lock(aptd->initiator);
/* We don't really care if this fails. If the datastore fails to get set up
* we'll still play the announcement. It's possible that the sound will be
* clipped for the initiator, but that's not the end of the world.
*/
setup_async_playback_datastore(aptd->initiator);
ast_channel_unlock(aptd->initiator);
}
return aptd;
}
static void async_playback_task_data_destroy(struct async_playback_task_data *aptd)
{
ast_channel_cleanup(aptd->initiator);
ast_free(aptd);
}
/*!
* \brief Wait for the initiator of an async playback to be ready
*
* See the description on the async_datastore_info structure for more
* information about what this is about.
*
* \param initiator The channel that queued the async announcement
*/
static void wait_for_initiator(struct ast_channel *initiator)
{
struct ast_datastore *async_datastore;
struct async_datastore_data *add;
ast_channel_lock(initiator);
async_datastore = ast_channel_datastore_find(initiator, &async_datastore_info, NULL);
ast_channel_unlock(initiator);
if (!async_datastore) {
return;
}
add = async_datastore->data;
ast_mutex_lock(&add->lock);
while (add->wait) {
ast_cond_wait(&add->cond, &add->lock);
}
ast_mutex_unlock(&add->lock);
}
/*!
* \brief Play an announcement into a confbridge asynchronously
*
* This runs in the playback queue taskprocessor. This ensures that
* all playbacks are handled in sequence and do not play over top one
* another.
*
* \param data An async_playback_task_data
* \return 0
*/
static int async_playback_task(void *data)
{
struct async_playback_task_data *aptd = data;
/* Wait for the initiator to get back in the bridge or be hung up */
if (aptd->initiator) {
wait_for_initiator(aptd->initiator);
}
playback_common(aptd->conference, aptd->filename, aptd->say_number);
async_playback_task_data_destroy(aptd);
return 0;
}
static int async_play_sound_helper(struct confbridge_conference *conference,
const char *filename, int say_number, struct ast_channel *initiator)
{
struct async_playback_task_data *aptd;
/* Do not waste resources trying to play files that do not exist */
if (ast_strlen_zero(filename)) {
if (say_number < 0) {
return 0;
}
} else if (!sound_file_exists(filename)) {
return 0;
}
aptd = async_playback_task_data_alloc(conference, filename, say_number, initiator);
if (!aptd) {
return -1;
}
if (ast_taskprocessor_push(conference->playback_queue, async_playback_task, aptd)) {
if (!ast_strlen_zero(filename)) {
ast_log(LOG_WARNING, "Unable to play file '%s' to conference '%s'\n",
filename, conference->name);
} else {
ast_log(LOG_WARNING, "Unable to say number '%d' to conference '%s'\n",
say_number, conference->name);
}
async_playback_task_data_destroy(aptd);
return -1;
}
return 0;
}
int async_play_sound_file(struct confbridge_conference *conference,
const char *filename, struct ast_channel *initiator)
{
return async_play_sound_helper(conference, filename, -1, initiator);
}
void async_play_sound_ready(struct ast_channel *chan)
{
struct ast_datastore *async_datastore;
struct async_datastore_data *add;
ast_channel_lock(chan);
async_datastore = ast_channel_datastore_find(chan, &async_datastore_info, NULL);
ast_channel_unlock(chan);
if (!async_datastore) {
return;
}
add = async_datastore->data;
ast_mutex_lock(&add->lock);
add->wait = 0;
ast_cond_signal(&add->cond);
ast_mutex_unlock(&add->lock);
}
/*!
* \brief Play number into the conference bridge
*
* \param conference The conference bridge to say the number into
* \param say_number number to say
*
* \retval 0 success
* \retval -1 failure
*/
static int play_sound_number(struct confbridge_conference *conference, int say_number)
{
return play_sound_helper(conference, NULL, say_number);
}
static int conf_handle_talker_cb(struct ast_bridge_channel *bridge_channel, void *hook_pvt, int talking)
{
struct confbridge_user *user = hook_pvt;
RAII_VAR(struct confbridge_conference *, conference, NULL, ao2_cleanup);
struct ast_json *talking_extras;
conference = ao2_find(conference_bridges, user->conference->name, OBJ_KEY);
if (!conference) {
/* Remove the hook since the conference does not exist. */
return -1;
}
ao2_lock(conference);
user->talking = talking;
ao2_unlock(conference);
talking_extras = ast_json_pack("{s: s, s: b}",
"talking_status", talking ? "on" : "off",
"admin", ast_test_flag(&user->u_profile, USER_OPT_ADMIN));
if (!talking_extras) {
return 0;
}
send_conf_stasis(conference, bridge_channel->chan, confbridge_talking_type(), talking_extras, 0);
ast_json_unref(talking_extras);
return 0;
}
static int conf_get_pin(struct ast_channel *chan, struct confbridge_user *user)
{
char pin_guess[MAX_PIN+1] = { 0, };
const char *pin = user->u_profile.pin;
char *tmp = pin_guess;
int i, res;
unsigned int len = MAX_PIN;
/*
* NOTE: We have not joined a conference yet so we have to use
* the bridge profile requested by the user.
*/
/* give them three tries to get the pin right */
for (i = 0; i < 3; i++) {
if (ast_app_getdata(chan,
conf_get_sound(CONF_SOUND_GET_PIN, user->b_profile.sounds),
tmp, len, 0) >= 0) {
if (!strcasecmp(pin, pin_guess)) {
return 0;
}
}
ast_streamfile(chan,
conf_get_sound(CONF_SOUND_INVALID_PIN, user->b_profile.sounds),
ast_channel_language(chan));
res = ast_waitstream(chan, AST_DIGIT_ANY);
if (res > 0) {
/* Account for digit already read during ivalid pin playback
* resetting pin buf. */
pin_guess[0] = res;
pin_guess[1] = '\0';
tmp = pin_guess + 1;
len = MAX_PIN - 1;
} else {
/* reset pin buf as empty buffer. */
tmp = pin_guess;
len = MAX_PIN;
}
}
return -1;
}
static int user_timeout(struct ast_bridge_channel *bridge_channel, void *ignore)
{
ast_bridge_channel_leave_bridge(bridge_channel, BRIDGE_CHANNEL_STATE_END, 0);
pbx_builtin_setvar_helper(bridge_channel->chan, "CONFBRIDGE_RESULT", "TIMEOUT");
return -1;
}
static int conf_rec_name(struct confbridge_user *user, const char *conf_name)
{
char destdir[PATH_MAX];
int res;
int duration = 20;
snprintf(destdir, sizeof(destdir), "%s/confbridge", ast_config_AST_SPOOL_DIR);
if (ast_mkdir(destdir, 0777) != 0) {
ast_log(LOG_WARNING, "mkdir '%s' failed: %s\n", destdir, strerror(errno));
return -1;
}
snprintf(user->name_rec_location, sizeof(user->name_rec_location),
"%s/confbridge-name-%s-%s", destdir,
conf_name, ast_channel_uniqueid(user->chan));
if (!(ast_test_flag(&user->u_profile, USER_OPT_ANNOUNCE_JOIN_LEAVE_REVIEW))) {
res = ast_play_and_record(user->chan,
"vm-rec-name",
user->name_rec_location,
10,
"sln",
&duration,
NULL,
ast_dsp_get_threshold_from_settings(THRESHOLD_SILENCE),
0,
NULL);
} else {
res = ast_record_review(user->chan,
"vm-rec-name",
user->name_rec_location,
10,
"sln",
&duration,
NULL);
}
if (res == -1) {
ast_filedelete(user->name_rec_location, NULL);
user->name_rec_location[0] = '\0';
return -1;
}
return 0;
}
struct async_delete_name_rec_task_data {
struct confbridge_conference *conference;
char filename[0];
};
static struct async_delete_name_rec_task_data *async_delete_name_rec_task_data_alloc(
struct confbridge_conference *conference, const char *filename)
{
struct async_delete_name_rec_task_data *atd;
atd = ast_malloc(sizeof(*atd) + strlen(filename) + 1);
if (!atd) {
return NULL;
}
/* Safe */
strcpy(atd->filename, filename);
atd->conference = conference;
return atd;
}
static void async_delete_name_rec_task_data_destroy(struct async_delete_name_rec_task_data *atd)
{
ast_free(atd);
}
/*!
* \brief Delete user's name file asynchronously
*
* This runs in the playback queue taskprocessor. This ensures that
* sound file is removed after playback is finished and not before.
*
* \param data An async_delete_name_rec_task_data
* \return 0
*/
static int async_delete_name_rec_task(void *data)
{
struct async_delete_name_rec_task_data *atd = data;
ast_filedelete(atd->filename, NULL);
ast_log(LOG_DEBUG, "Conference '%s' removed user name file '%s'\n",
atd->conference->name, atd->filename);
async_delete_name_rec_task_data_destroy(atd);
return 0;
}
static int async_delete_name_rec(struct confbridge_conference *conference,
const char *filename)
{
struct async_delete_name_rec_task_data *atd;
if (ast_strlen_zero(filename)) {
return 0;
} else if (!sound_file_exists(filename)) {
return 0;
}
atd = async_delete_name_rec_task_data_alloc(conference, filename);
if (!atd) {
return -1;
}
if (ast_taskprocessor_push(conference->playback_queue, async_delete_name_rec_task, atd)) {
ast_log(LOG_WARNING, "Conference '%s' was unable to remove user name file '%s'\n",
conference->name, filename);
async_delete_name_rec_task_data_destroy(atd);
return -1;
}
return 0;
}
static int join_callback(struct ast_bridge_channel *bridge_channel, void *ignore)
{
async_play_sound_ready(bridge_channel->chan);
return 0;
}
struct confbridge_hook_data {
struct confbridge_conference *conference;
struct confbridge_user *user;
enum ast_bridge_hook_type hook_type;
};
static int send_event_hook_callback(struct ast_bridge_channel *bridge_channel, void *data)
{
struct confbridge_hook_data *hook_data = data;
if (hook_data->hook_type == AST_BRIDGE_HOOK_TYPE_JOIN) {
send_join_event(hook_data->user, hook_data->conference);
} else {
send_leave_event(hook_data->user, hook_data->conference);
}
return 0;
}
/*! \brief The ConfBridge application */
static int confbridge_exec(struct ast_channel *chan, const char *data)
{
int res = 0, volume_adjustments[2];
int quiet = 0;
int async_delete_task_pushed = 0;
char *parse;
const char *b_profile_name = NULL;
const char *u_profile_name = NULL;
const char *menu_profile_name = NULL;
struct confbridge_conference *conference = NULL;
struct confbridge_user user = {
.chan = chan,
.tech_args.talking_threshold = DEFAULT_TALKING_THRESHOLD,
.tech_args.silence_threshold = DEFAULT_SILENCE_THRESHOLD,
.tech_args.drop_silence = 0,
};
struct confbridge_hook_data *join_hook_data;
struct confbridge_hook_data *leave_hook_data;
AST_DECLARE_APP_ARGS(args,
AST_APP_ARG(conf_name);
AST_APP_ARG(b_profile_name);
AST_APP_ARG(u_profile_name);
AST_APP_ARG(menu_profile_name);
);
if (ast_bridge_features_init(&user.features)) {
pbx_builtin_setvar_helper(chan, "CONFBRIDGE_RESULT", "FAILED");
res = -1;
goto confbridge_cleanup;
}
/* We need to make a copy of the input string if we are going to modify it! */
parse = ast_strdupa(data);
AST_STANDARD_APP_ARGS(args, parse);
if (ast_strlen_zero(args.conf_name)) {
pbx_builtin_setvar_helper(chan, "CONFBRIDGE_RESULT", "FAILED");
ast_log(LOG_WARNING, "%s requires an argument (conference name[,options])\n", app);
res = -1;
goto confbridge_cleanup;
}
if (strlen(args.conf_name) >= MAX_CONF_NAME) {
pbx_builtin_setvar_helper(chan, "CONFBRIDGE_RESULT", "FAILED");
ast_log(LOG_WARNING, "%s does not accept conference names longer than %d\n", app, MAX_CONF_NAME - 1);
res = -1;
goto confbridge_cleanup;
}
/* bridge profile name */
if (args.argc > 1 && !ast_strlen_zero(args.b_profile_name)) {
b_profile_name = args.b_profile_name;
}
if (!conf_find_bridge_profile(chan, b_profile_name, &user.b_profile)) {
pbx_builtin_setvar_helper(chan, "CONFBRIDGE_RESULT", "FAILED");
ast_log(LOG_WARNING, "Conference bridge profile %s does not exist\n", b_profile_name ?
b_profile_name : DEFAULT_BRIDGE_PROFILE);
res = -1;
goto confbridge_cleanup;
}
/* user profile name */
if (args.argc > 2 && !ast_strlen_zero(args.u_profile_name)) {
u_profile_name = args.u_profile_name;
}
if (!conf_find_user_profile(chan, u_profile_name, &user.u_profile)) {
pbx_builtin_setvar_helper(chan, "CONFBRIDGE_RESULT", "FAILED");
ast_log(LOG_WARNING, "Conference user profile %s does not exist\n", u_profile_name ?
u_profile_name : DEFAULT_USER_PROFILE);
res = -1;
goto confbridge_cleanup;
}
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
/* If channel hasn't been answered already, answer it, unless we're explicitly not supposed to */
if ((ast_channel_state(chan) != AST_STATE_UP) && (ast_test_flag(&user.u_profile, USER_OPT_ANSWER_CHANNEL))) {
ast_answer(chan);
}
quiet = ast_test_flag(&user.u_profile, USER_OPT_QUIET);
/* ask for a PIN immediately after finding user profile. This has to be
* prompted for requardless of quiet setting. */
if (!ast_strlen_zero(user.u_profile.pin)) {
if (conf_get_pin(chan, &user)) {
pbx_builtin_setvar_helper(chan, "CONFBRIDGE_RESULT", "FAILED");
res = -1; /* invalid PIN */
goto confbridge_cleanup;
}
}
/* See if we need them to record a intro name */
if (!quiet &&
(ast_test_flag(&user.u_profile, USER_OPT_ANNOUNCE_JOIN_LEAVE) ||
(ast_test_flag(&user.u_profile, USER_OPT_ANNOUNCE_JOIN_LEAVE_REVIEW)))) {
if (conf_rec_name(&user, args.conf_name)) {
pbx_builtin_setvar_helper(chan, "CONFBRIDGE_RESULT", "FAILED");
res = -1; /* Hangup during name recording */
goto confbridge_cleanup;
}
}
/* menu name */
if (args.argc > 3 && !ast_strlen_zero(args.menu_profile_name)) {
menu_profile_name = args.menu_profile_name;
}
if (conf_set_menu_to_user(chan, &user, menu_profile_name)) {
pbx_builtin_setvar_helper(chan, "CONFBRIDGE_RESULT", "FAILED");
ast_log(LOG_WARNING, "Conference menu profile %s does not exist\n", menu_profile_name ?
menu_profile_name : DEFAULT_MENU_PROFILE);
res = -1;
goto confbridge_cleanup;
}
/* Set if DTMF should pass through for this user or not */
if (ast_test_flag(&user.u_profile, USER_OPT_DTMF_PASS)) {
user.features.dtmf_passthrough = 1;
} else {
user.features.dtmf_passthrough = 0;
}
/* Set if text messaging is enabled for this user or not */
if (ast_test_flag(&user.u_profile, USER_OPT_TEXT_MESSAGING)) {
user.features.text_messaging = 1;
} else {
user.features.text_messaging = 0;
}
/* Set dsp threshold values if present */
if (user.u_profile.talking_threshold) {
user.tech_args.talking_threshold = user.u_profile.talking_threshold;
}
if (user.u_profile.silence_threshold) {
user.tech_args.silence_threshold = user.u_profile.silence_threshold;
}
/* Set a talker indicate call back if talking detection is requested */
if (ast_test_flag(&user.u_profile, USER_OPT_TALKER_DETECT)) {
if (ast_bridge_talk_detector_hook(&user.features, conf_handle_talker_cb,
&user, NULL, AST_BRIDGE_HOOK_REMOVE_ON_PULL)) {
pbx_builtin_setvar_helper(chan, "CONFBRIDGE_RESULT", "FAILED");
res = -1;
goto confbridge_cleanup;
}
}
/* Look for a conference bridge matching the provided name */
if (!(conference = join_conference_bridge(args.conf_name, &user))) {
pbx_builtin_setvar_helper(chan, "CONFBRIDGE_RESULT", "FAILED");
res = -1;
goto confbridge_cleanup;
}
/* Keep a copy of volume adjustments so we can restore them later if need be */
volume_adjustments[0] = ast_audiohook_volume_get(chan, AST_AUDIOHOOK_DIRECTION_READ);
volume_adjustments[1] = ast_audiohook_volume_get(chan, AST_AUDIOHOOK_DIRECTION_WRITE);
if (ast_test_flag(&user.u_profile, USER_OPT_DROP_SILENCE)) {
user.tech_args.drop_silence = 1;
}
if (ast_test_flag(&user.u_profile, USER_OPT_JITTERBUFFER)) {
ast_func_write(chan, "JITTERBUFFER(adaptive)", "default");
}
if (ast_test_flag(&user.u_profile, USER_OPT_DENOISE)) {
ast_func_write(chan, "DENOISE(rx)", "on");
}
/* if this user has a intro, play it before entering */
if (!ast_strlen_zero(user.name_rec_location)) {
ast_autoservice_start(chan);
play_sound_file(conference, user.name_rec_location);
play_sound_file(conference,
conf_get_sound(CONF_SOUND_HAS_JOINED, conference->b_profile.sounds));
ast_autoservice_stop(chan);
}
if (!quiet) {
const char *join_sound = conf_get_sound(CONF_SOUND_JOIN, conference->b_profile.sounds);
/* if hear_own_join_sound is enabled play the Join sound to everyone */
if (ast_test_flag(&user.u_profile, USER_OPT_HEAR_OWN_JOIN_SOUND) ) {
if (strcmp(conference->b_profile.language, ast_channel_language(chan))) {
ast_stream_and_wait(chan, join_sound, "");
ast_autoservice_start(chan);
play_sound_file(conference, join_sound);
ast_autoservice_stop(chan);
} else {
async_play_sound_file(conference, join_sound, chan);
}
/* if hear_own_join_sound is disabled only play the Join sound to just the conference */
} else {
ast_autoservice_start(chan);
play_sound_file(conference, join_sound);
ast_autoservice_stop(chan);
}
}
if (user.u_profile.timeout) {
ast_bridge_interval_hook(&user.features,
0,
user.u_profile.timeout * 1000,
user_timeout,
NULL,
NULL,
AST_BRIDGE_HOOK_REMOVE_ON_PULL);
}
/* See if we need to automatically set this user as a video source or not */
handle_video_on_join(conference, user.chan, ast_test_flag(&user.u_profile, USER_OPT_MARKEDUSER));
conf_moh_unsuspend(&user);
join_hook_data = ast_malloc(sizeof(*join_hook_data));
if (!join_hook_data) {
res = -1;
goto confbridge_cleanup;
}
join_hook_data->user = &user;
join_hook_data->conference = conference;
join_hook_data->hook_type = AST_BRIDGE_HOOK_TYPE_JOIN;
res = ast_bridge_join_hook(&user.features, send_event_hook_callback,
join_hook_data, ast_free_ptr, 0);
if (res) {
ast_free(join_hook_data);
ast_log(LOG_ERROR, "Couldn't add bridge join hook for channel '%s'\n", ast_channel_name(chan));
goto confbridge_cleanup;
}
leave_hook_data = ast_malloc(sizeof(*leave_hook_data));
if (!leave_hook_data) {
/* join_hook_data is cleaned up by ast_bridge_features_cleanup via the goto */
res = -1;
goto confbridge_cleanup;
}
leave_hook_data->user = &user;
leave_hook_data->conference = conference;
leave_hook_data->hook_type = AST_BRIDGE_HOOK_TYPE_LEAVE;
res = ast_bridge_leave_hook(&user.features, send_event_hook_callback,
leave_hook_data, ast_free_ptr, 0);
if (res) {
/* join_hook_data is cleaned up by ast_bridge_features_cleanup via the goto */
ast_free(leave_hook_data);
ast_log(LOG_ERROR, "Couldn't add bridge leave hook for channel '%s'\n", ast_channel_name(chan));
goto confbridge_cleanup;
}
if (ast_bridge_join_hook(&user.features, join_callback, NULL, NULL, 0)) {
async_play_sound_ready(user.chan);
}
ast_bridge_join(conference->bridge,
chan,
NULL,
&user.features,
&user.tech_args,
0);
/* This is a catch-all in case joining the bridge failed or for some reason
* an async announcement got queued up and hasn't been told to play yet
*/
async_play_sound_ready(chan);
if (!user.kicked && ast_check_hangup(chan)) {
pbx_builtin_setvar_helper(chan, "CONFBRIDGE_RESULT", "HANGUP");
}
/* if we're shutting down, don't attempt to do further processing */
if (ast_shutting_down()) {
HTTP: Stop accepting requests on final system shutdown. There are three CLI commands to stop and restart Asterisk each. 1) core stop/restart now - Hangup all calls and stop or restart Asterisk. New channels are prevented while the shutdown request is pending. 2) core stop/restart gracefully - Stop or restart Asterisk when there are no calls remaining in the system. New channels are prevented while the shutdown request is pending. 3) core stop/restart when convenient - Stop or restart Asterisk when there are no calls in the system. New calls are not prevented while the shutdown request is pending. ARI has made stopping/restarting Asterisk more problematic. While a shutdown request is pending it is desirable to continue to process ARI HTTP requests for current calls. To handle the current calls while a shutdown request is pending, a new committed to shutdown phase is needed so ARI applications can deal with the calls until the system is fully committed to shutdown. * Added a new shutdown committed phase so ARI applications can deal with calls until the final committed to shutdown phase is reached. * Made refuse new HTTP requests when the system has reached the final system shutdown phase. Starting anything while the system is actively releasing resources and unloading modules is not a good thing. * Split the bridging framework shutdown to not cleanup the global bridging containers when shutting down in a hurry. This is similar to how other modules prevent crashes on rapid system shutdown. * Moved ast_begin_shutdown(), ast_cancel_shutdown(), and ast_shutting_down(). You should not have to include channel.h just to access these system functions. ASTERISK-24752 #close Reported by: Matthew Jordan Review: https://reviewboard.asterisk.org/r/4399/ ........ Merged revisions 431692 from http://svn.asterisk.org/svn/asterisk/branches/13 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@431694 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2015-02-11 17:39:13 +00:00
/*
* Not taking any new calls at this time. We cannot create
* the announcer channel if this is the first channel into
* the conference and we certainly cannot create any
* recording channel.
*/
leave_conference(&user);
conference = NULL;
goto confbridge_cleanup;
}
/* If this user was a video source, we need to clean up and possibly pick a new source. */
handle_video_on_exit(conference, user.chan);
/* if this user has a intro, play it when leaving */
if (!quiet && !ast_strlen_zero(user.name_rec_location)) {
async_play_sound_file(conference, user.name_rec_location, NULL);
async_play_sound_file(conference,
conf_get_sound(CONF_SOUND_HAS_LEFT, conference->b_profile.sounds), NULL);
async_delete_name_rec(conference, user.name_rec_location);
async_delete_task_pushed = 1;
}
/* play the leave sound */
if (!quiet) {
const char *leave_sound = conf_get_sound(CONF_SOUND_LEAVE, conference->b_profile.sounds);
async_play_sound_file(conference, leave_sound, NULL);
}
/* If the user was kicked from the conference play back the audio prompt for it */
if (!quiet && user.kicked) {
res = ast_stream_and_wait(chan,
conf_get_sound(CONF_SOUND_KICKED, conference->b_profile.sounds),
"");
}
/* Easy as pie, depart this channel from the conference bridge */
leave_conference(&user);
conference = NULL;
/* Restore volume adjustments to previous values in case they were changed */
if (volume_adjustments[0]) {
ast_audiohook_volume_set(chan, AST_AUDIOHOOK_DIRECTION_READ, volume_adjustments[0]);
}
if (volume_adjustments[1]) {
ast_audiohook_volume_set(chan, AST_AUDIOHOOK_DIRECTION_WRITE, volume_adjustments[1]);
}
confbridge_cleanup:
if (!async_delete_task_pushed && !ast_strlen_zero(user.name_rec_location)) {
ast_filedelete(user.name_rec_location, NULL);
}
ast_bridge_features_cleanup(&user.features);
conf_bridge_profile_destroy(&user.b_profile);
return res;
}
static int action_toggle_mute(struct confbridge_conference *conference,
struct confbridge_user *user,
struct ast_bridge_channel *bridge_channel)
{
int mute;
/* Toggle user level mute request. */
mute = !user->muted;
generic_mute_unmute_user(conference, user, mute);
return play_file(bridge_channel, NULL,
conf_get_sound(mute ? CONF_SOUND_MUTED : CONF_SOUND_UNMUTED,
conference->b_profile.sounds)) < 0;
}
static int action_toggle_binaural(struct confbridge_conference *conference,
struct confbridge_user *user,
struct ast_bridge_channel *bridge_channel)
{
unsigned int binaural;
ast_bridge_channel_lock_bridge(bridge_channel);
binaural = !bridge_channel->binaural_suspended;
bridge_channel->binaural_suspended = binaural;
ast_bridge_unlock(bridge_channel->bridge);
return play_file(bridge_channel, NULL, (binaural ?
conf_get_sound(CONF_SOUND_BINAURAL_OFF, user->b_profile.sounds) :
conf_get_sound(CONF_SOUND_BINAURAL_ON, user->b_profile.sounds))) < 0;
}
static int action_toggle_mute_participants(struct confbridge_conference *conference, struct confbridge_user *user)
{
struct confbridge_user *cur_user = NULL;
const char *sound_to_play;
int mute;
ao2_lock(conference);
/* Toggle bridge level mute request. */
mute = !conference->muted;
conference->muted = mute;
AST_LIST_TRAVERSE(&conference->active_list, cur_user, list) {
if (!ast_test_flag(&cur_user->u_profile, USER_OPT_ADMIN)) {
/* Set user level to bridge level mute request. */
cur_user->muted = mute;
conf_update_user_mute(cur_user);
}
}
ao2_unlock(conference);
sound_to_play = conf_get_sound(
mute ? CONF_SOUND_PARTICIPANTS_MUTED : CONF_SOUND_PARTICIPANTS_UNMUTED,
conference->b_profile.sounds);
if (strcmp(conference->b_profile.language, ast_channel_language(user->chan))) {
/* The host needs to hear it seperately, as they don't get the audio from play_sound_helper */
ast_stream_and_wait(user->chan, sound_to_play, "");
/* Announce to the group that all participants are muted */
ast_autoservice_start(user->chan);
play_sound_file(conference, sound_to_play);
ast_autoservice_stop(user->chan);
} else {
/* Playing the sound asynchronously lets the sound be heard by everyone at once */
async_play_sound_file(conference, sound_to_play, user->chan);
}
return 0;
}
static int action_playback(struct ast_bridge_channel *bridge_channel, const char *playback_file)
{
char *file_copy = ast_strdupa(playback_file);
char *file = NULL;
while ((file = ast_strsep(&file_copy, '&', AST_STRSEP_STRIP | AST_STRSEP_TRIM))) {
if (ast_stream_and_wait(bridge_channel->chan, file, "")) {
ast_log(LOG_WARNING, "Failed to playback file %s to channel\n", file);
return -1;
}
}
return 0;
}
static int action_playback_and_continue(struct confbridge_conference *conference,
struct confbridge_user *user,
struct ast_bridge_channel *bridge_channel,
struct conf_menu *menu,
const char *playback_file,
const char *cur_dtmf,
int *stop_prompts)
{
int i;
int digit = 0;
char dtmf[MAXIMUM_DTMF_FEATURE_STRING];
struct conf_menu_entry new_menu_entry = { { 0, }, };
char *file_copy = ast_strdupa(playback_file);
char *file = NULL;
while ((file = ast_strsep(&file_copy, '&', AST_STRSEP_STRIP | AST_STRSEP_TRIM))) {
if (ast_streamfile(bridge_channel->chan, file, ast_channel_language(bridge_channel->chan))) {
ast_log(LOG_WARNING, "Failed to playback file %s to channel\n", file);
return -1;
}
/* now wait for more digits. */
if (!(digit = ast_waitstream(bridge_channel->chan, AST_DIGIT_ANY))) {
/* streaming finished and no DTMF was entered */
continue;
} else if (digit == -1) {
/* error */
return -1;
} else {
break; /* dtmf was entered */
}
}
if (!digit) {
/* streaming finished on all files and no DTMF was entered */
return -1;
}
ast_stopstream(bridge_channel->chan);
/* If we get here, then DTMF has been entered, This means no
* additional prompts should be played for this menu entry */
*stop_prompts = 1;
/* If a digit was pressed during the payback, update
* the dtmf string and look for a new menu entry in the
* menu structure */
ast_copy_string(dtmf, cur_dtmf, sizeof(dtmf));
for (i = 0; i < (MAXIMUM_DTMF_FEATURE_STRING - 1); i++) {
dtmf[i] = cur_dtmf[i];
if (!dtmf[i]) {
dtmf[i] = (char) digit;
dtmf[i + 1] = '\0';
i = -1;
break;
}
}
/* If i is not -1 then the new dtmf digit was _NOT_ added to the string.
* If this is the case, no new DTMF sequence should be looked for. */
if (i != -1) {
return 0;
}
if (conf_find_menu_entry_by_sequence(dtmf, menu, &new_menu_entry)) {
execute_menu_entry(conference,
user,
bridge_channel,
&new_menu_entry, menu);
conf_menu_entry_destroy(&new_menu_entry);
}
return 0;
}
static int action_kick_last(struct confbridge_conference *conference,
struct ast_bridge_channel *bridge_channel,
struct confbridge_user *user)
{
struct confbridge_user *last_user = NULL;
int isadmin = ast_test_flag(&user->u_profile, USER_OPT_ADMIN);
if (!isadmin) {
play_file(bridge_channel, NULL,
conf_get_sound(CONF_SOUND_ERROR_MENU, conference->b_profile.sounds));
ast_log(LOG_WARNING, "Only admin users can use the kick_last menu action. Channel %s of conf %s is not an admin.\n",
Replace direct access to channel name with accessor functions There are many benefits to making the ast_channel an opaque handle, from increasing maintainability to presenting ways to kill masquerades. This patch kicks things off by taking things a field at a time, renaming the field to '__do_not_use_${fieldname}' and then writing setters/getters and converting the existing code to using them. When all fields are done, we can move ast_channel to a C file from channel.h and lop off the '__do_not_use_'. This patch sets up main/channel_interal_api.c to be the only file that actually accesses the ast_channel's fields directly. The intent would be for any API functions in channel.c to use the accessor functions. No more monkeying around with channel internals. We should use our own APIs. The interesting changes in this patch are the addition of channel_internal_api.c, the moving of the AST_DATA stuff from channel.c to channel_internal_api.c (note: the AST_DATA stuff will have to be reworked to use accessor functions when ast_channel is really opaque), and some re-working of the way channel iterators/callbacks are handled so as to avoid creating fake ast_channels on the stack to pass in matching data by directly accessing fields (since "name" is a stringfield and the fake channel doesn't init the stringfields, you can't use the ast_channel_name_set() function). I went with ast_channel_name(chan) for a getter, and ast_channel_name_set(chan, name) for a setter. The majority of the grunt-work for this change was done by writing a semantic patch using Coccinelle ( http://coccinelle.lip6.fr/ ). Review: https://reviewboard.asterisk.org/r/1655/ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@350223 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-01-09 22:15:50 +00:00
ast_channel_name(bridge_channel->chan),
conference->name);
return -1;
}
ao2_lock(conference);
last_user = AST_LIST_LAST(&conference->active_list);
if (!last_user) {
ao2_unlock(conference);
return 0;
}
if (last_user == user || ast_test_flag(&last_user->u_profile, USER_OPT_ADMIN)) {
ao2_unlock(conference);
play_file(bridge_channel, NULL,
conf_get_sound(CONF_SOUND_ERROR_MENU, conference->b_profile.sounds));
} else if (!last_user->kicked) {
last_user->kicked = 1;
pbx_builtin_setvar_helper(last_user->chan, "CONFBRIDGE_RESULT", "KICKED");
ast_bridge_remove(conference->bridge, last_user->chan);
ao2_unlock(conference);
}
return 0;
}
static int action_dialplan_exec(struct ast_bridge_channel *bridge_channel, struct conf_menu_action *menu_action)
{
struct ast_pbx_args args;
struct ast_pbx *pbx;
char *exten;
char *context;
int priority;
int res;
memset(&args, 0, sizeof(args));
args.no_hangup_chan = 1;
ast_channel_lock(bridge_channel->chan);
/*save off*/
exten = ast_strdupa(ast_channel_exten(bridge_channel->chan));
context = ast_strdupa(ast_channel_context(bridge_channel->chan));
priority = ast_channel_priority(bridge_channel->chan);
pbx = ast_channel_pbx(bridge_channel->chan);
ast_channel_pbx_set(bridge_channel->chan, NULL);
/*set new*/
ast_channel_exten_set(bridge_channel->chan, menu_action->data.dialplan_args.exten);
ast_channel_context_set(bridge_channel->chan, menu_action->data.dialplan_args.context);
ast_channel_priority_set(bridge_channel->chan, menu_action->data.dialplan_args.priority);
ast_channel_unlock(bridge_channel->chan);
/*execute*/
res = ast_pbx_run_args(bridge_channel->chan, &args);
/*restore*/
ast_channel_lock(bridge_channel->chan);
ast_channel_exten_set(bridge_channel->chan, exten);
ast_channel_context_set(bridge_channel->chan, context);
ast_channel_priority_set(bridge_channel->chan, priority);
ast_channel_pbx_set(bridge_channel->chan, pbx);
ast_channel_unlock(bridge_channel->chan);
return res;
}
static int execute_menu_entry(struct confbridge_conference *conference,
struct confbridge_user *user,
struct ast_bridge_channel *bridge_channel,
struct conf_menu_entry *menu_entry,
struct conf_menu *menu)
{
struct conf_menu_action *menu_action;
int isadmin = ast_test_flag(&user->u_profile, USER_OPT_ADMIN);
int stop_prompts = 0;
int res = 0;
AST_LIST_TRAVERSE(&menu_entry->actions, menu_action, action) {
switch (menu_action->id) {
case MENU_ACTION_TOGGLE_MUTE:
res |= action_toggle_mute(conference, user, bridge_channel);
break;
case MENU_ACTION_TOGGLE_BINAURAL:
action_toggle_binaural(conference, user, bridge_channel);
break;
case MENU_ACTION_ADMIN_TOGGLE_MUTE_PARTICIPANTS:
if (!isadmin) {
break;
}
action_toggle_mute_participants(conference, user);
break;
case MENU_ACTION_PARTICIPANT_COUNT:
announce_user_count(conference, user, bridge_channel);
break;
case MENU_ACTION_PLAYBACK:
if (!stop_prompts) {
res |= action_playback(bridge_channel, menu_action->data.playback_file);
ast_test_suite_event_notify("CONF_MENU_PLAYBACK",
"Message: %s\r\nChannel: %s",
menu_action->data.playback_file, ast_channel_name(bridge_channel->chan));
}
break;
case MENU_ACTION_RESET_LISTENING:
ast_audiohook_volume_set(user->chan, AST_AUDIOHOOK_DIRECTION_WRITE, 0);
break;
case MENU_ACTION_RESET_TALKING:
ast_audiohook_volume_set(user->chan, AST_AUDIOHOOK_DIRECTION_READ, 0);
break;
case MENU_ACTION_INCREASE_LISTENING:
ast_audiohook_volume_adjust(user->chan,
AST_AUDIOHOOK_DIRECTION_WRITE, 1);
break;
case MENU_ACTION_DECREASE_LISTENING:
ast_audiohook_volume_adjust(user->chan,
AST_AUDIOHOOK_DIRECTION_WRITE, -1);
break;
case MENU_ACTION_INCREASE_TALKING:
ast_audiohook_volume_adjust(user->chan,
AST_AUDIOHOOK_DIRECTION_READ, 1);
break;
case MENU_ACTION_DECREASE_TALKING:
ast_audiohook_volume_adjust(user->chan,
AST_AUDIOHOOK_DIRECTION_READ, -1);
break;
case MENU_ACTION_PLAYBACK_AND_CONTINUE:
if (!(stop_prompts)) {
res |= action_playback_and_continue(conference,
user,
bridge_channel,
menu,
menu_action->data.playback_file,
menu_entry->dtmf,
&stop_prompts);
}
break;
case MENU_ACTION_DIALPLAN_EXEC:
res |= action_dialplan_exec(bridge_channel, menu_action);
break;
case MENU_ACTION_ADMIN_TOGGLE_LOCK:
if (!isadmin) {
break;
}
conference->locked = (!conference->locked ? 1 : 0);
res |= play_file(bridge_channel, NULL,
conf_get_sound(
conference->locked ? CONF_SOUND_LOCKED_NOW : CONF_SOUND_UNLOCKED_NOW,
conference->b_profile.sounds)) < 0;
break;
case MENU_ACTION_ADMIN_KICK_LAST:
res |= action_kick_last(conference, bridge_channel, user);
break;
case MENU_ACTION_LEAVE:
pbx_builtin_setvar_helper(bridge_channel->chan, "CONFBRIDGE_RESULT", "DTMF");
ao2_lock(conference);
ast_bridge_remove(conference->bridge, bridge_channel->chan);
ast_test_suite_event_notify("CONF_MENU_LEAVE",
"Channel: %s",
ast_channel_name(bridge_channel->chan));
ao2_unlock(conference);
break;
case MENU_ACTION_NOOP:
break;
case MENU_ACTION_SET_SINGLE_VIDEO_SRC:
ao2_lock(conference);
if (!ast_test_flag(&conference->b_profile, BRIDGE_OPT_VIDEO_SRC_SFU)) {
ast_bridge_set_single_src_video_mode(conference->bridge, bridge_channel->chan);
}
ao2_unlock(conference);
break;
case MENU_ACTION_RELEASE_SINGLE_VIDEO_SRC:
handle_video_on_exit(conference, bridge_channel->chan);
break;
}
}
return res;
}
int conf_handle_dtmf(struct ast_bridge_channel *bridge_channel,
struct confbridge_user *user,
struct conf_menu_entry *menu_entry,
struct conf_menu *menu)
{
/* See if music on hold is playing */
conf_moh_suspend(user);
/* execute the list of actions associated with this menu entry */
execute_menu_entry(user->conference, user, bridge_channel, menu_entry, menu);
/* See if music on hold needs to be started back up again */
conf_moh_unsuspend(user);
async_play_sound_ready(bridge_channel->chan);
return 0;
}
static int kick_conference_participant(struct confbridge_conference *conference,
const char *channel)
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
{
int res = -1;
int match;
struct confbridge_user *user = NULL;
int all = !strcasecmp("all", channel);
int participants = !strcasecmp("participants", channel);
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
SCOPED_AO2LOCK(bridge_lock, conference);
AST_LIST_TRAVERSE(&conference->active_list, user, list) {
if (user->kicked) {
continue;
}
match = !strcasecmp(channel, ast_channel_name(user->chan));
if (match || all
|| (participants && !ast_test_flag(&user->u_profile, USER_OPT_ADMIN))) {
user->kicked = 1;
pbx_builtin_setvar_helper(user->chan, "CONFBRIDGE_RESULT", "KICKED");
ast_bridge_remove(conference->bridge, user->chan);
res = 0;
if (match) {
return res;
}
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
}
}
AST_LIST_TRAVERSE(&conference->waiting_list, user, list) {
if (user->kicked) {
continue;
}
match = !strcasecmp(channel, ast_channel_name(user->chan));
if (match || all
|| (participants && !ast_test_flag(&user->u_profile, USER_OPT_ADMIN))) {
user->kicked = 1;
pbx_builtin_setvar_helper(user->chan, "CONFBRIDGE_RESULT", "KICKED");
ast_bridge_remove(conference->bridge, user->chan);
res = 0;
if (match) {
return res;
}
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
}
}
return res;
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
}
static char *complete_confbridge_name(const char *line, const char *word, int pos, int state)
{
int which = 0;
struct confbridge_conference *conference;
char *res = NULL;
int wordlen = strlen(word);
struct ao2_iterator iter;
iter = ao2_iterator_init(conference_bridges, 0);
while ((conference = ao2_iterator_next(&iter))) {
if (!strncasecmp(conference->name, word, wordlen) && ++which > state) {
res = ast_strdup(conference->name);
ao2_ref(conference, -1);
break;
}
ao2_ref(conference, -1);
}
ao2_iterator_destroy(&iter);
return res;
}
static char *complete_confbridge_participant(const char *conference_name, const char *line, const char *word, int pos, int state)
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
{
int which = 0;
RAII_VAR(struct confbridge_conference *, conference, NULL, ao2_cleanup);
struct confbridge_user *user;
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
char *res = NULL;
int wordlen = strlen(word);
conference = ao2_find(conference_bridges, conference_name, OBJ_KEY);
if (!conference) {
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
return NULL;
}
if (!strncasecmp("all", word, wordlen) && ++which > state) {
return ast_strdup("all");
}
if (!strncasecmp("participants", word, wordlen) && ++which > state) {
return ast_strdup("participants");
}
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
{
SCOPED_AO2LOCK(bridge_lock, conference);
AST_LIST_TRAVERSE(&conference->active_list, user, list) {
if (!strncasecmp(ast_channel_name(user->chan), word, wordlen) && ++which > state) {
res = ast_strdup(ast_channel_name(user->chan));
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
return res;
}
}
AST_LIST_TRAVERSE(&conference->waiting_list, user, list) {
if (!strncasecmp(ast_channel_name(user->chan), word, wordlen) && ++which > state) {
res = ast_strdup(ast_channel_name(user->chan));
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
return res;
}
}
}
return NULL;
}
static char *handle_cli_confbridge_kick(struct ast_cli_entry *e, int cmd, struct ast_cli_args *a)
{
struct confbridge_conference *conference;
int not_found;
switch (cmd) {
case CLI_INIT:
e->command = "confbridge kick";
e->usage =
"Usage: confbridge kick <conference> <channel>\n"
" Kicks a channel out of the conference bridge.\n"
" (all to kick everyone, participants to kick non-admins).\n";
return NULL;
case CLI_GENERATE:
if (a->pos == 2) {
return complete_confbridge_name(a->line, a->word, a->pos, a->n);
}
if (a->pos == 3) {
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
return complete_confbridge_participant(a->argv[2], a->line, a->word, a->pos, a->n);
}
return NULL;
}
if (a->argc != 4) {
return CLI_SHOWUSAGE;
}
conference = ao2_find(conference_bridges, a->argv[2], OBJ_KEY);
if (!conference) {
ast_cli(a->fd, "No conference bridge named '%s' found!\n", a->argv[2]);
return CLI_SUCCESS;
}
not_found = kick_conference_participant(conference, a->argv[3]);
ao2_ref(conference, -1);
if (not_found) {
if (!strcasecmp("all", a->argv[3]) || !strcasecmp("participants", a->argv[3])) {
ast_cli(a->fd, "No participants found!\n");
} else {
ast_cli(a->fd, "No participant named '%s' found!\n", a->argv[3]);
}
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
return CLI_SUCCESS;
}
ast_cli(a->fd, "Kicked '%s' out of conference '%s'\n", a->argv[3], a->argv[2]);
return CLI_SUCCESS;
}
static void handle_cli_confbridge_list_item(struct ast_cli_args *a, struct confbridge_user *user, int waiting)
{
char flag_str[6 + 1];/* Max flags + terminator */
int pos = 0;
/* Build flags column string. */
if (ast_test_flag(&user->u_profile, USER_OPT_ADMIN)) {
flag_str[pos++] = 'A';
}
if (ast_test_flag(&user->u_profile, USER_OPT_MARKEDUSER)) {
flag_str[pos++] = 'M';
}
if (ast_test_flag(&user->u_profile, USER_OPT_WAITMARKED)) {
flag_str[pos++] = 'W';
}
if (ast_test_flag(&user->u_profile, USER_OPT_ENDMARKED)) {
flag_str[pos++] = 'E';
}
if (user->muted) {
flag_str[pos++] = 'm';
}
if (waiting) {
flag_str[pos++] = 'w';
}
flag_str[pos] = '\0';
ast_cli(a->fd, "%-30s %-6s %-16s %-16s %-16s %s\n",
ast_channel_name(user->chan),
flag_str,
user->u_profile.name,
user->conference->b_profile.name,
user->menu_name,
S_COR(ast_channel_caller(user->chan)->id.number.valid,
ast_channel_caller(user->chan)->id.number.str, "<unknown>"));
}
static char *handle_cli_confbridge_list(struct ast_cli_entry *e, int cmd, struct ast_cli_args *a)
{
struct confbridge_conference *conference;
switch (cmd) {
case CLI_INIT:
e->command = "confbridge list";
e->usage =
"Usage: confbridge list [<name>]\n"
" Lists all currently active conference bridges or a specific conference bridge.\n"
"\n"
" When a conference bridge name is provided, flags may be shown for users. Below\n"
" are the flags and what they represent.\n"
"\n"
" Flags:\n"
" A - The user is an admin\n"
" M - The user is a marked user\n"
" W - The user must wait for a marked user to join\n"
" E - The user will be kicked after the last marked user leaves the conference\n"
" m - The user is muted\n"
" w - The user is waiting for a marked user to join\n";
return NULL;
case CLI_GENERATE:
if (a->pos == 2) {
return complete_confbridge_name(a->line, a->word, a->pos, a->n);
}
return NULL;
}
if (a->argc == 2) {
struct ao2_iterator iter;
ast_cli(a->fd, "Conference Bridge Name Users Marked Locked Muted\n");
ast_cli(a->fd, "================================ ====== ====== ====== =====\n");
iter = ao2_iterator_init(conference_bridges, 0);
while ((conference = ao2_iterator_next(&iter))) {
ast_cli(a->fd, "%-32s %6u %6u %-6s %s\n",
conference->name,
conference->activeusers + conference->waitingusers,
conference->markedusers,
AST_CLI_YESNO(conference->locked),
AST_CLI_YESNO(conference->muted));
ao2_ref(conference, -1);
}
ao2_iterator_destroy(&iter);
return CLI_SUCCESS;
}
if (a->argc == 3) {
struct confbridge_user *user;
conference = ao2_find(conference_bridges, a->argv[2], OBJ_KEY);
if (!conference) {
ast_cli(a->fd, "No conference bridge named '%s' found!\n", a->argv[2]);
return CLI_SUCCESS;
}
ast_cli(a->fd, "Channel Flags User Profile Bridge Profile Menu CallerID\n");
ast_cli(a->fd, "============================== ====== ================ ================ ================ ================\n");
ao2_lock(conference);
AST_LIST_TRAVERSE(&conference->active_list, user, list) {
handle_cli_confbridge_list_item(a, user, 0);
}
AST_LIST_TRAVERSE(&conference->waiting_list, user, list) {
handle_cli_confbridge_list_item(a, user, 1);
}
ao2_unlock(conference);
ao2_ref(conference, -1);
return CLI_SUCCESS;
}
return CLI_SHOWUSAGE;
}
/*! \internal
* \brief finds a conference by name and locks/unlocks.
*
* \retval 0 success
* \retval -1 conference not found
*/
static int generic_lock_unlock_helper(int lock, const char *conference_name)
{
struct confbridge_conference *conference;
int res = 0;
conference = ao2_find(conference_bridges, conference_name, OBJ_KEY);
if (!conference) {
return -1;
}
ao2_lock(conference);
conference->locked = lock;
ast_test_suite_event_notify("CONF_LOCK", "Message: conference %s\r\nConference: %s", conference->locked ? "locked" : "unlocked", conference->b_profile.name);
ao2_unlock(conference);
ao2_ref(conference, -1);
return res;
}
/*! \internal
* \brief finds a conference user by channel name and mutes/unmutes them.
*
* \retval 0 success
* \retval -1 conference not found
* \retval -2 user not found
*/
static int generic_mute_unmute_helper(int mute, const char *conference_name,
const char *chan_name)
{
RAII_VAR(struct confbridge_conference *, conference, NULL, ao2_cleanup);
struct confbridge_user *user;
int all = !strcasecmp("all", chan_name);
int participants = !strcasecmp("participants", chan_name);
int res = -2;
conference = ao2_find(conference_bridges, conference_name, OBJ_KEY);
if (!conference) {
return -1;
}
{
SCOPED_AO2LOCK(bridge_lock, conference);
AST_LIST_TRAVERSE(&conference->active_list, user, list) {
int match = !strncasecmp(chan_name, ast_channel_name(user->chan),
strlen(chan_name));
if (match || all
|| (participants && !ast_test_flag(&user->u_profile, USER_OPT_ADMIN))) {
generic_mute_unmute_user(conference, user, mute);
res = 0;
if (match) {
return res;
}
}
}
AST_LIST_TRAVERSE(&conference->waiting_list, user, list) {
int match = !strncasecmp(chan_name, ast_channel_name(user->chan),
strlen(chan_name));
if (match || all
|| (participants && !ast_test_flag(&user->u_profile, USER_OPT_ADMIN))) {
generic_mute_unmute_user(conference, user, mute);
res = 0;
if (match) {
return res;
}
}
}
}
return res;
}
static int cli_mute_unmute_helper(int mute, struct ast_cli_args *a)
{
int res = generic_mute_unmute_helper(mute, a->argv[2], a->argv[3]);
if (res == -1) {
ast_cli(a->fd, "No conference bridge named '%s' found!\n", a->argv[2]);
return -1;
} else if (res == -2) {
if (!strcasecmp("all", a->argv[3]) || !strcasecmp("participants", a->argv[3])) {
ast_cli(a->fd, "No participants found in conference %s\n", a->argv[2]);
} else {
ast_cli(a->fd, "No channel named '%s' found in conference %s\n", a->argv[3], a->argv[2]);
}
return -1;
}
ast_cli(a->fd, "%s %s from confbridge %s\n", mute ? "Muting" : "Unmuting", a->argv[3], a->argv[2]);
return 0;
}
static char *handle_cli_confbridge_mute(struct ast_cli_entry *e, int cmd, struct ast_cli_args *a)
{
switch (cmd) {
case CLI_INIT:
e->command = "confbridge mute";
e->usage =
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
"Usage: confbridge mute <conference> <channel>\n"
" Mute a channel in a conference.\n"
" (all to mute everyone, participants to mute non-admins)\n"
" If the specified channel is a prefix,\n"
" the action will be taken on the first\n"
" matching channel.\n";
return NULL;
case CLI_GENERATE:
if (a->pos == 2) {
return complete_confbridge_name(a->line, a->word, a->pos, a->n);
}
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
if (a->pos == 3) {
return complete_confbridge_participant(a->argv[2], a->line, a->word, a->pos, a->n);
}
return NULL;
}
if (a->argc != 4) {
return CLI_SHOWUSAGE;
}
cli_mute_unmute_helper(1, a);
return CLI_SUCCESS;
}
static char *handle_cli_confbridge_unmute(struct ast_cli_entry *e, int cmd, struct ast_cli_args *a)
{
switch (cmd) {
case CLI_INIT:
e->command = "confbridge unmute";
e->usage =
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
"Usage: confbridge unmute <conference> <channel>\n"
" Unmute a channel in a conference.\n"
" (all to unmute everyone, participants to unmute non-admins)\n"
" If the specified channel is a prefix,\n"
" the action will be taken on the first\n"
" matching channel.\n";
return NULL;
case CLI_GENERATE:
if (a->pos == 2) {
return complete_confbridge_name(a->line, a->word, a->pos, a->n);
}
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
if (a->pos == 3) {
return complete_confbridge_participant(a->argv[2], a->line, a->word, a->pos, a->n);
}
return NULL;
}
if (a->argc != 4) {
return CLI_SHOWUSAGE;
}
cli_mute_unmute_helper(0, a);
return CLI_SUCCESS;
}
static char *handle_cli_confbridge_lock(struct ast_cli_entry *e, int cmd, struct ast_cli_args *a)
{
switch (cmd) {
case CLI_INIT:
e->command = "confbridge lock";
e->usage =
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
"Usage: confbridge lock <conference>\n"
" Lock a conference. While locked, no new non-admins\n"
" may join the conference.\n";
return NULL;
case CLI_GENERATE:
if (a->pos == 2) {
return complete_confbridge_name(a->line, a->word, a->pos, a->n);
}
return NULL;
}
if (a->argc != 3) {
return CLI_SHOWUSAGE;
}
if (generic_lock_unlock_helper(1, a->argv[2])) {
ast_cli(a->fd, "Conference %s is not found\n", a->argv[2]);
} else {
ast_cli(a->fd, "Conference %s is locked.\n", a->argv[2]);
}
return CLI_SUCCESS;
}
static char *handle_cli_confbridge_unlock(struct ast_cli_entry *e, int cmd, struct ast_cli_args *a)
{
switch (cmd) {
case CLI_INIT:
e->command = "confbridge unlock";
e->usage =
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
"Usage: confbridge unlock <conference>\n"
" Unlock a previously locked conference.\n";
return NULL;
case CLI_GENERATE:
if (a->pos == 2) {
return complete_confbridge_name(a->line, a->word, a->pos, a->n);
}
return NULL;
}
if (a->argc != 3) {
return CLI_SHOWUSAGE;
}
if (generic_lock_unlock_helper(0, a->argv[2])) {
ast_cli(a->fd, "Conference %s is not found\n", a->argv[2]);
} else {
ast_cli(a->fd, "Conference %s is unlocked.\n", a->argv[2]);
}
return CLI_SUCCESS;
}
static char *handle_cli_confbridge_start_record(struct ast_cli_entry *e, int cmd, struct ast_cli_args *a)
{
const char *rec_file = NULL;
struct confbridge_conference *conference;
switch (cmd) {
case CLI_INIT:
e->command = "confbridge record start";
e->usage =
"Usage: confbridge record start <conference> <file>\n"
" <file> is optional, Otherwise the bridge profile\n"
" record file will be used. If the bridge profile\n"
" has no record file specified, a file will automatically\n"
" be generated in the monitor directory\n";
return NULL;
case CLI_GENERATE:
if (a->pos == 3) {
return complete_confbridge_name(a->line, a->word, a->pos, a->n);
}
return NULL;
}
if (a->argc < 4) {
return CLI_SHOWUSAGE;
}
if (a->argc == 5) {
rec_file = a->argv[4];
}
conference = ao2_find(conference_bridges, a->argv[3], OBJ_KEY);
if (!conference) {
ast_cli(a->fd, "Conference not found.\n");
return CLI_FAILURE;
}
ao2_lock(conference);
if (conf_is_recording(conference)) {
ast_cli(a->fd, "Conference is already being recorded.\n");
ao2_unlock(conference);
ao2_ref(conference, -1);
return CLI_SUCCESS;
}
if (!ast_strlen_zero(rec_file)) {
ast_copy_string(conference->b_profile.rec_file, rec_file, sizeof(conference->b_profile.rec_file));
}
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
if (conf_start_record(conference)) {
ast_cli(a->fd, "Could not start recording due to internal error.\n");
ao2_unlock(conference);
ao2_ref(conference, -1);
return CLI_FAILURE;
}
ao2_unlock(conference);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
ast_cli(a->fd, "Recording started\n");
ao2_ref(conference, -1);
return CLI_SUCCESS;
}
static char *handle_cli_confbridge_stop_record(struct ast_cli_entry *e, int cmd, struct ast_cli_args *a)
{
struct confbridge_conference *conference;
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
int ret;
switch (cmd) {
case CLI_INIT:
e->command = "confbridge record stop";
e->usage =
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
"Usage: confbridge record stop <conference>\n"
" Stop a previously started recording.\n";
return NULL;
case CLI_GENERATE:
if (a->pos == 3) {
return complete_confbridge_name(a->line, a->word, a->pos, a->n);
}
return NULL;
}
if (a->argc != 4) {
return CLI_SHOWUSAGE;
}
conference = ao2_find(conference_bridges, a->argv[3], OBJ_KEY);
if (!conference) {
ast_cli(a->fd, "Conference not found.\n");
return CLI_SUCCESS;
}
ao2_lock(conference);
ret = conf_stop_record(conference);
ao2_unlock(conference);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
ast_cli(a->fd, "Recording %sstopped.\n", ret ? "could not be " : "");
ao2_ref(conference, -1);
return CLI_SUCCESS;
}
static struct ast_cli_entry cli_confbridge[] = {
AST_CLI_DEFINE(handle_cli_confbridge_list, "List conference bridges and participants."),
AST_CLI_DEFINE(handle_cli_confbridge_kick, "Kick participants out of conference bridges."),
AST_CLI_DEFINE(handle_cli_confbridge_mute, "Mute participants."),
AST_CLI_DEFINE(handle_cli_confbridge_unmute, "Unmute participants."),
AST_CLI_DEFINE(handle_cli_confbridge_lock, "Lock a conference."),
AST_CLI_DEFINE(handle_cli_confbridge_unlock, "Unlock a conference."),
AST_CLI_DEFINE(handle_cli_confbridge_start_record, "Start recording a conference"),
AST_CLI_DEFINE(handle_cli_confbridge_stop_record, "Stop recording a conference."),
};
static struct ast_custom_function confbridge_function = {
.name = "CONFBRIDGE",
.write = func_confbridge_helper,
};
static int func_confbridge_info(struct ast_channel *chan, const char *cmd, char *data, char *buf, size_t len);
static struct ast_custom_function confbridge_info_function = {
.name = "CONFBRIDGE_INFO",
.read = func_confbridge_info,
};
static int func_confbridge_channels(struct ast_channel *chan, const char *cmd, char *data, char *buf, size_t len)
{
char *parse, *outbuf;
struct confbridge_conference *conference;
struct confbridge_user *user;
int bytes, count = 0;
size_t outlen;
AST_DECLARE_APP_ARGS(args,
AST_APP_ARG(type);
AST_APP_ARG(confno);
);
/* parse all the required arguments and make sure they exist. */
if (ast_strlen_zero(data)) {
return -1;
}
parse = ast_strdupa(data);
AST_STANDARD_APP_ARGS(args, parse);
if (ast_strlen_zero(args.confno) || ast_strlen_zero(args.type)) {
ast_log(LOG_WARNING, "Usage: %s(category,confno)", cmd);
return -1;
}
conference = ao2_find(conference_bridges, args.confno, OBJ_KEY);
if (!conference) {
ast_debug(1, "No such conference: %s\n", args.confno);
return -1;
}
outbuf = buf;
outlen = len;
ao2_lock(conference);
if (!strcasecmp(args.type, "parties")) {
AST_LIST_TRAVERSE(&conference->active_list, user, list) {
bytes = snprintf(outbuf, outlen, "%s%s", count++ ? "," : "", ast_channel_name(user->chan));
outbuf += bytes;
outlen -= bytes;
}
AST_LIST_TRAVERSE(&conference->waiting_list, user, list) {
bytes = snprintf(outbuf, outlen, "%s%s", count++ ? "," : "", ast_channel_name(user->chan));
outbuf += bytes;
outlen -= bytes;
}
} else if (!strcasecmp(args.type, "active")) {
AST_LIST_TRAVERSE(&conference->active_list, user, list) {
bytes = snprintf(outbuf, outlen, "%s%s", count++ ? "," : "", ast_channel_name(user->chan));
outbuf += bytes;
outlen -= bytes;
}
} else if (!strcasecmp(args.type, "waiting")) {
AST_LIST_TRAVERSE(&conference->waiting_list, user, list) {
bytes = snprintf(outbuf, outlen, "%s%s", count++ ? "," : "", ast_channel_name(user->chan));
outbuf += bytes;
outlen -= bytes;
}
} else if (!strcasecmp(args.type, "admins")) {
AST_LIST_TRAVERSE(&conference->active_list, user, list) {
if (ast_test_flag(&user->u_profile, USER_OPT_ADMIN)) {
bytes = snprintf(outbuf, outlen, "%s%s", count++ ? "," : "", ast_channel_name(user->chan));
outbuf += bytes;
outlen -= bytes;
}
}
} else if (!strcasecmp(args.type, "marked")) {
AST_LIST_TRAVERSE(&conference->active_list, user, list) {
if (ast_test_flag(&user->u_profile, USER_OPT_MARKEDUSER)) {
bytes = snprintf(outbuf, outlen, "%s%s", count++ ? "," : "", ast_channel_name(user->chan));
outbuf += bytes;
outlen -= bytes;
}
}
} else {
ast_log(LOG_ERROR, "Invalid keyword '%s' passed to %s.\n", args.type, cmd);
}
ao2_unlock(conference);
ao2_ref(conference, -1);
return 0;
}
static struct ast_custom_function confbridge_channels_function = {
.name = "CONFBRIDGE_CHANNELS",
.read = func_confbridge_channels,
};
static int action_confbridgelist_item(struct mansession *s, const char *id_text, struct confbridge_conference *conference, struct confbridge_user *user, int waiting)
{
struct ast_channel_snapshot *snapshot;
struct ast_str *snap_str;
snapshot = ast_channel_snapshot_get_latest(ast_channel_uniqueid(user->chan));
if (!snapshot) {
return 0;
}
snap_str = ast_manager_build_channel_state_string(snapshot);
if (!snap_str) {
ao2_ref(snapshot, -1);
return 0;
}
astman_append(s,
"Event: ConfbridgeList\r\n"
"%s"
"Conference: %s\r\n"
"Admin: %s\r\n"
"MarkedUser: %s\r\n"
"WaitMarked: %s\r\n"
"EndMarked: %s\r\n"
"EndMarkedAny: %s\r\n"
"Waiting: %s\r\n"
"Muted: %s\r\n"
"Talking: %s\r\n"
"AnsweredTime: %d\r\n"
"%s"
"\r\n",
id_text,
conference->name,
AST_YESNO(ast_test_flag(&user->u_profile, USER_OPT_ADMIN)),
AST_YESNO(ast_test_flag(&user->u_profile, USER_OPT_MARKEDUSER)),
AST_YESNO(ast_test_flag(&user->u_profile, USER_OPT_WAITMARKED)),
AST_YESNO(ast_test_flag(&user->u_profile, USER_OPT_ENDMARKED)),
AST_YESNO(ast_test_flag(&user->u_profile, USER_OPT_ENDMARKEDANY)),
AST_YESNO(waiting),
AST_YESNO(user->muted),
AST_YESNO(user->talking),
ast_channel_get_up_time(user->chan),
ast_str_buffer(snap_str));
ast_free(snap_str);
ao2_ref(snapshot, -1);
return 1;
}
static int action_confbridgelist(struct mansession *s, const struct message *m)
{
const char *actionid = astman_get_header(m, "ActionID");
const char *conference_name = astman_get_header(m, "Conference");
struct confbridge_user *user;
struct confbridge_conference *conference;
char id_text[80];
int total = 0;
id_text[0] = '\0';
if (!ast_strlen_zero(actionid)) {
snprintf(id_text, sizeof(id_text), "ActionID: %s\r\n", actionid);
}
if (ast_strlen_zero(conference_name)) {
astman_send_error(s, m, "No Conference name provided.");
return 0;
}
if (!ao2_container_count(conference_bridges)) {
astman_send_error(s, m, "No active conferences.");
return 0;
}
conference = ao2_find(conference_bridges, conference_name, OBJ_KEY);
if (!conference) {
astman_send_error(s, m, "No Conference by that name found.");
return 0;
}
astman_send_listack(s, m, "Confbridge user list will follow", "start");
ao2_lock(conference);
AST_LIST_TRAVERSE(&conference->active_list, user, list) {
total += action_confbridgelist_item(s, id_text, conference, user, 0);
}
AST_LIST_TRAVERSE(&conference->waiting_list, user, list) {
total += action_confbridgelist_item(s, id_text, conference, user, 1);
}
ao2_unlock(conference);
ao2_ref(conference, -1);
AMI: Make AMI actions that generate event lists consistent. * Made the following AMI actions use list API calls for consistency: Agents BridgeInfo BridgeList BridgeTechnologyList ConfbridgeLIst ConfbridgeLIstRooms CoreShowChannels DAHDIShowChannels DBGet DeviceStateList ExtensionStateList FAXSessions Hangup IAXpeerlist IAXpeers IAXregistry MeetmeList MeetmeListRooms MWIGet ParkedCalls Parkinglots PJSIPShowEndpoint PJSIPShowEndpoints PJSIPShowRegistrationsInbound PJSIPShowRegistrationsOutbound PJSIPShowResourceLists PJSIPShowSubscriptionsInbound PJSIPShowSubscriptionsOutbound PresenceStateList PRIShowSpans QueueStatus QueueSummary ShowDialPlan SIPpeers SIPpeerstatus SIPshowregistry SKINNYdevices SKINNYlines Status VoicemailUsersList * Incremented the AMI version to 2.7.0. * Changed astman_send_listack() to not use the listflag parameter and always set the value to "Start" so the start capitalization is consistent. i.e., The FAXSessions used "Start" while the rest of the system used "start". The corresponding complete event always used "Complete". * Fixed ami_show_resource_lists() "PJSIPShowResourceLists" to output the AMI ActionID for all of its list events. * Fixed off-nominal AMI protocol error in manager_bridge_info(), manager_parking_status_single_lot(), and manager_parking_status_all_lots(). Use of astman_send_error() after responding to the original AMI action request violates the action response pattern by sending two responses. * Fixed minor protocol error in action_getconfig() when no requested categories are found. Each line needs to be formatted as "Header: text". * Fixed off-nominal memory leak in manager_build_parked_call_string(). * Eliminated unnecessary use of RAII_VAR() in ami_subscription_detail(). ASTERISK-24049 #close Reported by: Jonathan Rose Review: https://reviewboard.asterisk.org/r/4315/ ........ Merged revisions 430434 from http://svn.asterisk.org/svn/asterisk/branches/13 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@430435 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2015-01-09 18:16:54 +00:00
astman_send_list_complete_start(s, m, "ConfbridgeListComplete", total);
astman_send_list_complete_end(s);
return 0;
}
static int action_confbridgelistrooms(struct mansession *s, const struct message *m)
{
const char *actionid = astman_get_header(m, "ActionID");
struct confbridge_conference *conference;
struct ao2_iterator iter;
char id_text[512] = "";
int totalitems = 0;
if (!ast_strlen_zero(actionid)) {
snprintf(id_text, sizeof(id_text), "ActionID: %s\r\n", actionid);
}
if (!ao2_container_count(conference_bridges)) {
astman_send_error(s, m, "No active conferences.");
return 0;
}
astman_send_listack(s, m, "Confbridge conferences will follow", "start");
/* Traverse the conference list */
iter = ao2_iterator_init(conference_bridges, 0);
while ((conference = ao2_iterator_next(&iter))) {
totalitems++;
ao2_lock(conference);
astman_append(s,
"Event: ConfbridgeListRooms\r\n"
"%s"
"Conference: %s\r\n"
"Parties: %u\r\n"
"Marked: %u\r\n"
"Locked: %s\r\n"
"Muted: %s\r\n"
"\r\n",
id_text,
conference->name,
conference->activeusers + conference->waitingusers,
conference->markedusers,
AST_YESNO(conference->locked),
AST_YESNO(conference->muted));
ao2_unlock(conference);
ao2_ref(conference, -1);
}
ao2_iterator_destroy(&iter);
/* Send final confirmation */
AMI: Make AMI actions that generate event lists consistent. * Made the following AMI actions use list API calls for consistency: Agents BridgeInfo BridgeList BridgeTechnologyList ConfbridgeLIst ConfbridgeLIstRooms CoreShowChannels DAHDIShowChannels DBGet DeviceStateList ExtensionStateList FAXSessions Hangup IAXpeerlist IAXpeers IAXregistry MeetmeList MeetmeListRooms MWIGet ParkedCalls Parkinglots PJSIPShowEndpoint PJSIPShowEndpoints PJSIPShowRegistrationsInbound PJSIPShowRegistrationsOutbound PJSIPShowResourceLists PJSIPShowSubscriptionsInbound PJSIPShowSubscriptionsOutbound PresenceStateList PRIShowSpans QueueStatus QueueSummary ShowDialPlan SIPpeers SIPpeerstatus SIPshowregistry SKINNYdevices SKINNYlines Status VoicemailUsersList * Incremented the AMI version to 2.7.0. * Changed astman_send_listack() to not use the listflag parameter and always set the value to "Start" so the start capitalization is consistent. i.e., The FAXSessions used "Start" while the rest of the system used "start". The corresponding complete event always used "Complete". * Fixed ami_show_resource_lists() "PJSIPShowResourceLists" to output the AMI ActionID for all of its list events. * Fixed off-nominal AMI protocol error in manager_bridge_info(), manager_parking_status_single_lot(), and manager_parking_status_all_lots(). Use of astman_send_error() after responding to the original AMI action request violates the action response pattern by sending two responses. * Fixed minor protocol error in action_getconfig() when no requested categories are found. Each line needs to be formatted as "Header: text". * Fixed off-nominal memory leak in manager_build_parked_call_string(). * Eliminated unnecessary use of RAII_VAR() in ami_subscription_detail(). ASTERISK-24049 #close Reported by: Jonathan Rose Review: https://reviewboard.asterisk.org/r/4315/ ........ Merged revisions 430434 from http://svn.asterisk.org/svn/asterisk/branches/13 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@430435 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2015-01-09 18:16:54 +00:00
astman_send_list_complete_start(s, m, "ConfbridgeListRoomsComplete", totalitems);
astman_send_list_complete_end(s);
return 0;
}
static int action_mute_unmute_helper(struct mansession *s, const struct message *m, int mute)
{
const char *conference_name = astman_get_header(m, "Conference");
const char *channel_name = astman_get_header(m, "Channel");
int res = 0;
if (ast_strlen_zero(conference_name)) {
astman_send_error(s, m, "No Conference name provided.");
return 0;
}
if (ast_strlen_zero(channel_name)) {
astman_send_error(s, m, "No channel name provided.");
return 0;
}
if (!ao2_container_count(conference_bridges)) {
astman_send_error(s, m, "No active conferences.");
return 0;
}
res = generic_mute_unmute_helper(mute, conference_name, channel_name);
if (res == -1) {
astman_send_error(s, m, "No Conference by that name found.");
return 0;
} else if (res == -2) {
astman_send_error(s, m, "No Channel by that name found in Conference.");
return 0;
}
astman_send_ack(s, m, mute ? "User muted" : "User unmuted");
return 0;
}
static int action_confbridgeunmute(struct mansession *s, const struct message *m)
{
return action_mute_unmute_helper(s, m, 0);
}
static int action_confbridgemute(struct mansession *s, const struct message *m)
{
return action_mute_unmute_helper(s, m, 1);
}
static int action_lock_unlock_helper(struct mansession *s, const struct message *m, int lock)
{
const char *conference_name = astman_get_header(m, "Conference");
int res = 0;
if (ast_strlen_zero(conference_name)) {
astman_send_error(s, m, "No Conference name provided.");
return 0;
}
if (!ao2_container_count(conference_bridges)) {
astman_send_error(s, m, "No active conferences.");
return 0;
}
if ((res = generic_lock_unlock_helper(lock, conference_name))) {
astman_send_error(s, m, "No Conference by that name found.");
return 0;
}
astman_send_ack(s, m, lock ? "Conference locked" : "Conference unlocked");
return 0;
}
static int action_confbridgeunlock(struct mansession *s, const struct message *m)
{
return action_lock_unlock_helper(s, m, 0);
}
static int action_confbridgelock(struct mansession *s, const struct message *m)
{
return action_lock_unlock_helper(s, m, 1);
}
static int action_confbridgekick(struct mansession *s, const struct message *m)
{
const char *conference_name = astman_get_header(m, "Conference");
const char *channel = astman_get_header(m, "Channel");
struct confbridge_conference *conference;
int found;
if (ast_strlen_zero(conference_name)) {
astman_send_error(s, m, "No Conference name provided.");
return 0;
}
if (!ao2_container_count(conference_bridges)) {
astman_send_error(s, m, "No active conferences.");
return 0;
}
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
conference = ao2_find(conference_bridges, conference_name, OBJ_KEY);
if (!conference) {
astman_send_error(s, m, "No Conference by that name found.");
return 0;
}
found = !kick_conference_participant(conference, channel);
ao2_ref(conference, -1);
if (found) {
astman_send_ack(s, m, !strcmp("all", channel) ? "All participants kicked" : "User kicked");
} else {
astman_send_error(s, m, "No Channel by that name found in Conference.");
}
return 0;
}
static int action_confbridgestartrecord(struct mansession *s, const struct message *m)
{
const char *conference_name = astman_get_header(m, "Conference");
const char *recordfile = astman_get_header(m, "RecordFile");
struct confbridge_conference *conference;
if (ast_strlen_zero(conference_name)) {
astman_send_error(s, m, "No Conference name provided.");
return 0;
}
if (!ao2_container_count(conference_bridges)) {
astman_send_error(s, m, "No active conferences.");
return 0;
}
conference = ao2_find(conference_bridges, conference_name, OBJ_KEY);
if (!conference) {
astman_send_error(s, m, "No Conference by that name found.");
return 0;
}
ao2_lock(conference);
if (conf_is_recording(conference)) {
astman_send_error(s, m, "Conference is already being recorded.");
ao2_unlock(conference);
ao2_ref(conference, -1);
return 0;
}
if (!ast_strlen_zero(recordfile)) {
ast_copy_string(conference->b_profile.rec_file, recordfile, sizeof(conference->b_profile.rec_file));
}
if (conf_start_record(conference)) {
astman_send_error(s, m, "Internal error starting conference recording.");
ao2_unlock(conference);
ao2_ref(conference, -1);
return 0;
}
ao2_unlock(conference);
ao2_ref(conference, -1);
astman_send_ack(s, m, "Conference Recording Started.");
return 0;
}
static int action_confbridgestoprecord(struct mansession *s, const struct message *m)
{
const char *conference_name = astman_get_header(m, "Conference");
struct confbridge_conference *conference;
if (ast_strlen_zero(conference_name)) {
astman_send_error(s, m, "No Conference name provided.");
return 0;
}
if (!ao2_container_count(conference_bridges)) {
astman_send_error(s, m, "No active conferences.");
return 0;
}
conference = ao2_find(conference_bridges, conference_name, OBJ_KEY);
if (!conference) {
astman_send_error(s, m, "No Conference by that name found.");
return 0;
}
ao2_lock(conference);
if (conf_stop_record(conference)) {
ao2_unlock(conference);
astman_send_error(s, m, "Internal error while stopping recording.");
ao2_ref(conference, -1);
return 0;
}
ao2_unlock(conference);
ao2_ref(conference, -1);
astman_send_ack(s, m, "Conference Recording Stopped.");
return 0;
}
static int action_confbridgesetsinglevideosrc(struct mansession *s, const struct message *m)
{
const char *conference_name = astman_get_header(m, "Conference");
const char *channel = astman_get_header(m, "Channel");
struct confbridge_user *user;
struct confbridge_conference *conference;
if (ast_strlen_zero(conference_name)) {
astman_send_error(s, m, "No Conference name provided.");
return 0;
}
if (ast_strlen_zero(channel)) {
astman_send_error(s, m, "No channel name provided.");
return 0;
}
if (!ao2_container_count(conference_bridges)) {
astman_send_error(s, m, "No active conferences.");
return 0;
}
conference = ao2_find(conference_bridges, conference_name, OBJ_KEY);
if (!conference) {
astman_send_error(s, m, "No Conference by that name found.");
return 0;
}
/* find channel and set as video src. */
ao2_lock(conference);
AST_LIST_TRAVERSE(&conference->active_list, user, list) {
if (!strncmp(channel, ast_channel_name(user->chan), strlen(channel))) {
ast_bridge_set_single_src_video_mode(conference->bridge, user->chan);
break;
}
}
ao2_unlock(conference);
ao2_ref(conference, -1);
/* do not access user after conference unlock. We are just
* using this check to see if it was found or not */
if (!user) {
astman_send_error(s, m, "No channel by that name found in conference.");
return 0;
}
astman_send_ack(s, m, "Conference single video source set.");
return 0;
}
static int func_confbridge_info(struct ast_channel *chan, const char *cmd, char *data, char *buf, size_t len)
{
char *parse;
struct confbridge_conference *conference;
struct confbridge_user *user;
int count = 0;
AST_DECLARE_APP_ARGS(args,
AST_APP_ARG(type);
AST_APP_ARG(confno);
);
/* parse all the required arguments and make sure they exist. */
if (ast_strlen_zero(data)) {
return -1;
}
parse = ast_strdupa(data);
AST_STANDARD_APP_ARGS(args, parse);
if (ast_strlen_zero(args.confno) || ast_strlen_zero(args.type)) {
return -1;
}
conference = ao2_find(conference_bridges, args.confno, OBJ_KEY);
if (!conference) {
snprintf(buf, len, "0");
return 0;
}
/* get the correct count for the type requested */
ao2_lock(conference);
if (!strcasecmp(args.type, "parties")) {
AST_LIST_TRAVERSE(&conference->active_list, user, list) {
count++;
}
AST_LIST_TRAVERSE(&conference->waiting_list, user, list) {
Clean up ConfBridge commands to account for wait_marked users When ConfBridge was refactored to better handle the concept of marked, wait_marked, and normal users co-existing in a conference (thereby implementing a state machine for the conference), the wait_marked users were put into their own list of conference participants, separate from the active users. This list is used for wait_marked users when they are waiting in a conference but no marked user has joined; normal users may have joined at this point however. There are several AMI/CLI commands that affect conference users that were not checking the wait_marked users list: * CLI/AMI commands that mute/unmute a participant. In this case, wait_marked users have to remain in their particular state and should not be affected - however, the commands would return "Channel not found" as opposed to the appropriate error condition. * CLI/AMI commands that kick a participant. An admin should always be able to kick a participant out of the conference. This patch fixes both sets of commands, and cleans up the CLI commands slightly by allowing them to complete a participant name (this was supposed to have been added, but the function call was commented out and wasn't implemented). Review: https://reviewboard.asterisk.org/r/2346/ (closes issue AST-1114) Reported by: John Bigelow Tested by: John Bigelow ........ Merged revisions 382068 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@382070 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2013-02-26 15:52:02 +00:00
count++;
}
} else if (!strcasecmp(args.type, "admins")) {
AST_LIST_TRAVERSE(&conference->active_list, user, list) {
if (ast_test_flag(&user->u_profile, USER_OPT_ADMIN)) {
count++;
}
}
} else if (!strcasecmp(args.type, "marked")) {
AST_LIST_TRAVERSE(&conference->active_list, user, list) {
if (ast_test_flag(&user->u_profile, USER_OPT_MARKEDUSER)) {
count++;
}
}
} else if (!strcasecmp(args.type, "locked")) {
count = conference->locked;
} else if (!strcasecmp(args.type, "muted")) {
count = conference->muted;
} else {
ast_log(LOG_ERROR, "Invalid keyword '%s' passed to CONFBRIDGE_INFO.\n", args.type);
}
snprintf(buf, len, "%d", count);
ao2_unlock(conference);
ao2_ref(conference, -1);
return 0;
}
static int confkick_exec(struct ast_channel *chan, const char *data)
{
char *parse;
struct confbridge_conference *conference;
int not_found;
AST_DECLARE_APP_ARGS(args,
AST_APP_ARG(confbridge);
AST_APP_ARG(channel);
);
if (ast_strlen_zero(data)) {
ast_log(LOG_WARNING, "No conference bridge specified.\n");
pbx_builtin_setvar_helper(chan, "CONFKICKSTATUS", "FAILURE");
return 0;
}
parse = ast_strdupa(data);
AST_STANDARD_APP_ARGS(args, parse);
conference = ao2_find(conference_bridges, args.confbridge, OBJ_KEY);
if (!conference) {
ast_log(LOG_WARNING, "No conference bridge named '%s' found!\n", args.confbridge);
pbx_builtin_setvar_helper(chan, "CONFKICKSTATUS", "FAILURE");
return 0;
}
if (ast_strlen_zero(args.channel)) {
not_found = kick_conference_participant(conference, "all");
} else {
not_found = kick_conference_participant(conference, args.channel);
}
ao2_ref(conference, -1);
if (not_found) {
if (ast_strlen_zero(args.channel) || !strcasecmp("all", args.channel) || !strcasecmp("participants", args.channel)) {
ast_log(LOG_WARNING, "No participants found in conference bridge '%s'!\n", args.confbridge);
} else {
ast_log(LOG_WARNING, "No participant named '%s' found in conference bridge '%s'!\n", args.channel, args.confbridge);
}
pbx_builtin_setvar_helper(chan, "CONFKICKSTATUS", "FAILURE");
return 0;
}
ast_debug(1, "Kicked '%s' out of conference '%s'\n", args.channel, args.confbridge);
pbx_builtin_setvar_helper(chan, "CONFKICKSTATUS", "SUCCESS");
return 0;
}
void conf_add_user_active(struct confbridge_conference *conference, struct confbridge_user *user)
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
{
AST_LIST_INSERT_TAIL(&conference->active_list, user, list);
conference->activeusers++;
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
}
void conf_add_user_marked(struct confbridge_conference *conference, struct confbridge_user *user)
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
{
AST_LIST_INSERT_TAIL(&conference->active_list, user, list);
conference->activeusers++;
conference->markedusers++;
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
}
void conf_add_user_waiting(struct confbridge_conference *conference, struct confbridge_user *user)
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
{
AST_LIST_INSERT_TAIL(&conference->waiting_list, user, list);
conference->waitingusers++;
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
}
void conf_remove_user_active(struct confbridge_conference *conference, struct confbridge_user *user)
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
{
AST_LIST_REMOVE(&conference->active_list, user, list);
conference->activeusers--;
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
}
void conf_remove_user_marked(struct confbridge_conference *conference, struct confbridge_user *user)
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
{
AST_LIST_REMOVE(&conference->active_list, user, list);
conference->activeusers--;
conference->markedusers--;
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
}
void conf_mute_only_active(struct confbridge_conference *conference)
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
{
struct confbridge_user *only_user = AST_LIST_FIRST(&conference->active_list);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
/* Turn on MOH if the single participant is set up for it */
if (ast_test_flag(&only_user->u_profile, USER_OPT_MUSICONHOLD)) {
conf_moh_start(only_user);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
}
conf_update_user_mute(only_user);
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
}
void conf_remove_user_waiting(struct confbridge_conference *conference, struct confbridge_user *user)
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
{
AST_LIST_REMOVE(&conference->waiting_list, user, list);
conference->waitingusers--;
Resolve issues in ConfBridge regarding marked, waitmarked, and unmarked users Thank's to Neil Tallim (flan)'s tireless testing, issue reporting, and patches it became clear that app_confbridge had some complex logic in how it handled interactions between marked, waitmarked, and unmarked users. In particular, there were some areas in which the interactions between the users resulted in inconsistent behavior, and app_confbridge was missing logic in how to handle some corner cases. Some areas included: * Poor handling of mixing unmarked and waitmarked users * Inconsistencies in how MOH and muting was applied to various users * Handling of various announcements for different user profile options flan's patches seem to fix the various issues, but highlighted how hard the code could be to maintain. In an attempt to make things easier to maintain and to more fully enumerate the various cases that exist, this patch breaks up the logic into a state machine-like setup. Please note that the various state transitioned are documented on the Asterisk wiki: https://wiki.asterisk.org/wiki/display/AST/Confbridge+state+changes Review: //https://reviewboard.asterisk.org/r/2072/ Note that for the following issues, mjordan uploaded the patch, although it was written by twilson. Any contributor license discrepency is due to that. (closes issue ASTERISK-19562) Reported by: flan Tested by: flan, mjordan, jrose patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-19726) Reported by: flan Tested by: flan patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) (closes issue ASTERISK-20181) Reported by: Jonathan White Tested by: Jonathan White patches: bugASTERISK-19562_ASTERISK-19726_ASTERISK-20181.patch uploaded by twilson (license 6283) ........ Merged revisions 374652 from http://svn.asterisk.org/svn/asterisk/branches/10 ........ Merged revisions 374657 from http://svn.asterisk.org/svn/asterisk/branches/11 git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@374658 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2012-10-08 18:48:34 +00:00
}
/*!
* \internal
* \brief Unregister a ConfBridge channel technology.
* \since 12.0.0
*
* \param tech What to unregister.
*/
static void unregister_channel_tech(struct ast_channel_tech *tech)
{
ast_channel_unregister(tech);
media formats: re-architect handling of media for performance improvements In the old times media formats were represented using a bit field. This was fast but had a few limitations. 1. Asterisk was limited in how many formats it could handle. 2. Formats, being a bit field, could not include any attribute information. A format was strictly its type, e.g., "this is ulaw". This was changed in Asterisk 10 (see https://wiki.asterisk.org/wiki/display/AST/Media+Architecture+Proposal for notes on that work) which led to the creation of the ast_format structure. This structure allowed Asterisk to handle attributes and bundle information with a format. Additionally, ast_format_cap was created to act as a container for multiple formats that, together, formed the capability of some entity. Another mechanism was added to allow logic to be registered which performed format attribute negotiation. Everywhere throughout the codebase Asterisk was changed to use this strategy. Unfortunately, in software, there is no free lunch. These new capabilities came at a cost. Performance analysis and profiling showed that we spend an inordinate amount of time comparing, copying, and generally manipulating formats and their related structures. Basic prototyping has shown that a reasonably large performance improvement could be made in this area. This patch is the result of that project, which overhauled the media format architecture and its usage in Asterisk to improve performance. Generally, the new philosophy for handling formats is as follows: * The ast_format structure is reference counted. This removed a large amount of the memory allocations and copying that was done in prior versions. * In order to prevent race conditions while keeping things performant, the ast_format structure is immutable by convention and lock-free. Violate this tenet at your peril! * Because formats are reference counted, codecs are also reference counted. The Asterisk core generally provides built-in codecs and caches the ast_format structures created to represent them. Generally, to prevent inordinate amounts of module reference bumping, codecs and formats can be added at run-time but cannot be removed. * All compatibility with the bit field representation of codecs/formats has been moved to a compatibility API. The primary user of this representation is chan_iax2, which must continue to maintain its bit-field usage of formats for interoperability concerns. * When a format is negotiated with attributes, or when a format cannot be represented by one of the cached formats, a new format object is created or cloned from an existing format. That format may have the same codec underlying it, but is a different format than a version of the format with different attributes or without attributes. * While formats are reference counted objects, the reference count maintained on the format should be manipulated with care. Formats are generally cached and will persist for the lifetime of Asterisk and do not explicitly need to have their lifetime modified. An exception to this is when the user of a format does not know where the format came from *and* the user may outlive the provider of the format. This occurs, for example, when a format is read from a channel: the channel may have a format with attributes (hence, non-cached) and the user of the format may last longer than the channel (if the reference to the channel is released prior to the format's reference). For more information on this work, see the API design notes: https://wiki.asterisk.org/wiki/display/AST/Media+Format+Rewrite Finally, this work was the culmination of a large number of developer's efforts. Extra thanks goes to Corey Farrell, who took on a large amount of the work in the Asterisk core, chan_sip, and was an invaluable resource in peer reviews throughout this project. There were a substantial number of patches contributed during this work; the following issues/patch names simply reflect some of the work (and will cause the release scripts to give attribution to the individuals who work on them). Reviews: https://reviewboard.asterisk.org/r/3814 https://reviewboard.asterisk.org/r/3808 https://reviewboard.asterisk.org/r/3805 https://reviewboard.asterisk.org/r/3803 https://reviewboard.asterisk.org/r/3801 https://reviewboard.asterisk.org/r/3798 https://reviewboard.asterisk.org/r/3800 https://reviewboard.asterisk.org/r/3794 https://reviewboard.asterisk.org/r/3793 https://reviewboard.asterisk.org/r/3792 https://reviewboard.asterisk.org/r/3791 https://reviewboard.asterisk.org/r/3790 https://reviewboard.asterisk.org/r/3789 https://reviewboard.asterisk.org/r/3788 https://reviewboard.asterisk.org/r/3787 https://reviewboard.asterisk.org/r/3786 https://reviewboard.asterisk.org/r/3784 https://reviewboard.asterisk.org/r/3783 https://reviewboard.asterisk.org/r/3778 https://reviewboard.asterisk.org/r/3774 https://reviewboard.asterisk.org/r/3775 https://reviewboard.asterisk.org/r/3772 https://reviewboard.asterisk.org/r/3761 https://reviewboard.asterisk.org/r/3754 https://reviewboard.asterisk.org/r/3753 https://reviewboard.asterisk.org/r/3751 https://reviewboard.asterisk.org/r/3750 https://reviewboard.asterisk.org/r/3748 https://reviewboard.asterisk.org/r/3747 https://reviewboard.asterisk.org/r/3746 https://reviewboard.asterisk.org/r/3742 https://reviewboard.asterisk.org/r/3740 https://reviewboard.asterisk.org/r/3739 https://reviewboard.asterisk.org/r/3738 https://reviewboard.asterisk.org/r/3737 https://reviewboard.asterisk.org/r/3736 https://reviewboard.asterisk.org/r/3734 https://reviewboard.asterisk.org/r/3722 https://reviewboard.asterisk.org/r/3713 https://reviewboard.asterisk.org/r/3703 https://reviewboard.asterisk.org/r/3689 https://reviewboard.asterisk.org/r/3687 https://reviewboard.asterisk.org/r/3674 https://reviewboard.asterisk.org/r/3671 https://reviewboard.asterisk.org/r/3667 https://reviewboard.asterisk.org/r/3665 https://reviewboard.asterisk.org/r/3625 https://reviewboard.asterisk.org/r/3602 https://reviewboard.asterisk.org/r/3519 https://reviewboard.asterisk.org/r/3518 https://reviewboard.asterisk.org/r/3516 https://reviewboard.asterisk.org/r/3515 https://reviewboard.asterisk.org/r/3512 https://reviewboard.asterisk.org/r/3506 https://reviewboard.asterisk.org/r/3413 https://reviewboard.asterisk.org/r/3410 https://reviewboard.asterisk.org/r/3387 https://reviewboard.asterisk.org/r/3388 https://reviewboard.asterisk.org/r/3389 https://reviewboard.asterisk.org/r/3390 https://reviewboard.asterisk.org/r/3321 https://reviewboard.asterisk.org/r/3320 https://reviewboard.asterisk.org/r/3319 https://reviewboard.asterisk.org/r/3318 https://reviewboard.asterisk.org/r/3266 https://reviewboard.asterisk.org/r/3265 https://reviewboard.asterisk.org/r/3234 https://reviewboard.asterisk.org/r/3178 ASTERISK-23114 #close Reported by: mjordan media_formats_translation_core.diff uploaded by kharwell (License 6464) rb3506.diff uploaded by mjordan (License 6283) media_format_app_file.diff uploaded by kharwell (License 6464) misc-2.diff uploaded by file (License 5000) chan_mild-3.diff uploaded by file (License 5000) chan_obscure.diff uploaded by file (License 5000) jingle.diff uploaded by file (License 5000) funcs.diff uploaded by file (License 5000) formats.diff uploaded by file (License 5000) core.diff uploaded by file (License 5000) bridges.diff uploaded by file (License 5000) mf-codecs-2.diff uploaded by file (License 5000) mf-app_fax.diff uploaded by file (License 5000) mf-apps-3.diff uploaded by file (License 5000) media-formats-3.diff uploaded by file (License 5000) ASTERISK-23715 rb3713.patch uploaded by coreyfarrell (License 5909) rb3689.patch uploaded by mjordan (License 6283) ASTERISK-23957 rb3722.patch uploaded by mjordan (License 6283) mf-attributes-3.diff uploaded by file (License 5000) ASTERISK-23958 Tested by: jrose rb3822.patch uploaded by coreyfarrell (License 5909) rb3800.patch uploaded by jrose (License 6182) chan_sip.diff uploaded by mjordan (License 6283) rb3747.patch uploaded by jrose (License 6182) ASTERISK-23959 #close Tested by: sgriepentrog, mjordan, coreyfarrell sip_cleanup.diff uploaded by opticron (License 6273) chan_sip_caps.diff uploaded by mjordan (License 6283) rb3751.patch uploaded by coreyfarrell (License 5909) chan_sip-3.diff uploaded by file (License 5000) ASTERISK-23960 #close Tested by: opticron direct_media.diff uploaded by opticron (License 6273) pjsip-direct-media.diff uploaded by file (License 5000) format_cap_remove.diff uploaded by opticron (License 6273) media_format_fixes.diff uploaded by opticron (License 6273) chan_pjsip-2.diff uploaded by file (License 5000) ASTERISK-23966 #close Tested by: rmudgett rb3803.patch uploaded by rmudgetti (License 5621) chan_dahdi.diff uploaded by file (License 5000) ASTERISK-24064 #close Tested by: coreyfarrell, mjordan, opticron, file, rmudgett, sgriepentrog, jrose rb3814.patch uploaded by rmudgett (License 5621) moh_cleanup.diff uploaded by opticron (License 6273) bridge_leak.diff uploaded by opticron (License 6273) translate.diff uploaded by file (License 5000) rb3795.patch uploaded by rmudgett (License 5621) tls_fix.diff uploaded by mjordan (License 6283) fax-mf-fix-2.diff uploaded by file (License 5000) rtp_transfer_stuff uploaded by mjordan (License 6283) rb3787.patch uploaded by rmudgett (License 5621) media-formats-explicit-translate-format-3.diff uploaded by file (License 5000) format_cache_case_fix.diff uploaded by opticron (License 6273) rb3774.patch uploaded by rmudgett (License 5621) rb3775.patch uploaded by rmudgett (License 5621) rtp_engine_fix.diff uploaded by opticron (License 6273) rtp_crash_fix.diff uploaded by opticron (License 6273) rb3753.patch uploaded by mjordan (License 6283) rb3750.patch uploaded by mjordan (License 6283) rb3748.patch uploaded by rmudgett (License 5621) media_format_fixes.diff uploaded by opticron (License 6273) rb3740.patch uploaded by mjordan (License 6283) rb3739.patch uploaded by mjordan (License 6283) rb3734.patch uploaded by mjordan (License 6283) rb3689.patch uploaded by mjordan (License 6283) rb3674.patch uploaded by coreyfarrell (License 5909) rb3671.patch uploaded by coreyfarrell (License 5909) rb3667.patch uploaded by coreyfarrell (License 5909) rb3665.patch uploaded by mjordan (License 6283) rb3625.patch uploaded by coreyfarrell (License 5909) rb3602.patch uploaded by coreyfarrell (License 5909) format_compatibility-2.diff uploaded by file (License 5000) core.diff uploaded by file (License 5000) git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@419044 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2014-07-20 22:06:33 +00:00
ao2_cleanup(tech->capabilities);
}
/*!
* \internal
* \brief Register a ConfBridge channel technology.
* \since 12.0.0
*
* \param tech What to register.
*
* \retval 0 on success.
* \retval -1 on error.
*/
static int register_channel_tech(struct ast_channel_tech *tech)
{
media formats: re-architect handling of media for performance improvements In the old times media formats were represented using a bit field. This was fast but had a few limitations. 1. Asterisk was limited in how many formats it could handle. 2. Formats, being a bit field, could not include any attribute information. A format was strictly its type, e.g., "this is ulaw". This was changed in Asterisk 10 (see https://wiki.asterisk.org/wiki/display/AST/Media+Architecture+Proposal for notes on that work) which led to the creation of the ast_format structure. This structure allowed Asterisk to handle attributes and bundle information with a format. Additionally, ast_format_cap was created to act as a container for multiple formats that, together, formed the capability of some entity. Another mechanism was added to allow logic to be registered which performed format attribute negotiation. Everywhere throughout the codebase Asterisk was changed to use this strategy. Unfortunately, in software, there is no free lunch. These new capabilities came at a cost. Performance analysis and profiling showed that we spend an inordinate amount of time comparing, copying, and generally manipulating formats and their related structures. Basic prototyping has shown that a reasonably large performance improvement could be made in this area. This patch is the result of that project, which overhauled the media format architecture and its usage in Asterisk to improve performance. Generally, the new philosophy for handling formats is as follows: * The ast_format structure is reference counted. This removed a large amount of the memory allocations and copying that was done in prior versions. * In order to prevent race conditions while keeping things performant, the ast_format structure is immutable by convention and lock-free. Violate this tenet at your peril! * Because formats are reference counted, codecs are also reference counted. The Asterisk core generally provides built-in codecs and caches the ast_format structures created to represent them. Generally, to prevent inordinate amounts of module reference bumping, codecs and formats can be added at run-time but cannot be removed. * All compatibility with the bit field representation of codecs/formats has been moved to a compatibility API. The primary user of this representation is chan_iax2, which must continue to maintain its bit-field usage of formats for interoperability concerns. * When a format is negotiated with attributes, or when a format cannot be represented by one of the cached formats, a new format object is created or cloned from an existing format. That format may have the same codec underlying it, but is a different format than a version of the format with different attributes or without attributes. * While formats are reference counted objects, the reference count maintained on the format should be manipulated with care. Formats are generally cached and will persist for the lifetime of Asterisk and do not explicitly need to have their lifetime modified. An exception to this is when the user of a format does not know where the format came from *and* the user may outlive the provider of the format. This occurs, for example, when a format is read from a channel: the channel may have a format with attributes (hence, non-cached) and the user of the format may last longer than the channel (if the reference to the channel is released prior to the format's reference). For more information on this work, see the API design notes: https://wiki.asterisk.org/wiki/display/AST/Media+Format+Rewrite Finally, this work was the culmination of a large number of developer's efforts. Extra thanks goes to Corey Farrell, who took on a large amount of the work in the Asterisk core, chan_sip, and was an invaluable resource in peer reviews throughout this project. There were a substantial number of patches contributed during this work; the following issues/patch names simply reflect some of the work (and will cause the release scripts to give attribution to the individuals who work on them). Reviews: https://reviewboard.asterisk.org/r/3814 https://reviewboard.asterisk.org/r/3808 https://reviewboard.asterisk.org/r/3805 https://reviewboard.asterisk.org/r/3803 https://reviewboard.asterisk.org/r/3801 https://reviewboard.asterisk.org/r/3798 https://reviewboard.asterisk.org/r/3800 https://reviewboard.asterisk.org/r/3794 https://reviewboard.asterisk.org/r/3793 https://reviewboard.asterisk.org/r/3792 https://reviewboard.asterisk.org/r/3791 https://reviewboard.asterisk.org/r/3790 https://reviewboard.asterisk.org/r/3789 https://reviewboard.asterisk.org/r/3788 https://reviewboard.asterisk.org/r/3787 https://reviewboard.asterisk.org/r/3786 https://reviewboard.asterisk.org/r/3784 https://reviewboard.asterisk.org/r/3783 https://reviewboard.asterisk.org/r/3778 https://reviewboard.asterisk.org/r/3774 https://reviewboard.asterisk.org/r/3775 https://reviewboard.asterisk.org/r/3772 https://reviewboard.asterisk.org/r/3761 https://reviewboard.asterisk.org/r/3754 https://reviewboard.asterisk.org/r/3753 https://reviewboard.asterisk.org/r/3751 https://reviewboard.asterisk.org/r/3750 https://reviewboard.asterisk.org/r/3748 https://reviewboard.asterisk.org/r/3747 https://reviewboard.asterisk.org/r/3746 https://reviewboard.asterisk.org/r/3742 https://reviewboard.asterisk.org/r/3740 https://reviewboard.asterisk.org/r/3739 https://reviewboard.asterisk.org/r/3738 https://reviewboard.asterisk.org/r/3737 https://reviewboard.asterisk.org/r/3736 https://reviewboard.asterisk.org/r/3734 https://reviewboard.asterisk.org/r/3722 https://reviewboard.asterisk.org/r/3713 https://reviewboard.asterisk.org/r/3703 https://reviewboard.asterisk.org/r/3689 https://reviewboard.asterisk.org/r/3687 https://reviewboard.asterisk.org/r/3674 https://reviewboard.asterisk.org/r/3671 https://reviewboard.asterisk.org/r/3667 https://reviewboard.asterisk.org/r/3665 https://reviewboard.asterisk.org/r/3625 https://reviewboard.asterisk.org/r/3602 https://reviewboard.asterisk.org/r/3519 https://reviewboard.asterisk.org/r/3518 https://reviewboard.asterisk.org/r/3516 https://reviewboard.asterisk.org/r/3515 https://reviewboard.asterisk.org/r/3512 https://reviewboard.asterisk.org/r/3506 https://reviewboard.asterisk.org/r/3413 https://reviewboard.asterisk.org/r/3410 https://reviewboard.asterisk.org/r/3387 https://reviewboard.asterisk.org/r/3388 https://reviewboard.asterisk.org/r/3389 https://reviewboard.asterisk.org/r/3390 https://reviewboard.asterisk.org/r/3321 https://reviewboard.asterisk.org/r/3320 https://reviewboard.asterisk.org/r/3319 https://reviewboard.asterisk.org/r/3318 https://reviewboard.asterisk.org/r/3266 https://reviewboard.asterisk.org/r/3265 https://reviewboard.asterisk.org/r/3234 https://reviewboard.asterisk.org/r/3178 ASTERISK-23114 #close Reported by: mjordan media_formats_translation_core.diff uploaded by kharwell (License 6464) rb3506.diff uploaded by mjordan (License 6283) media_format_app_file.diff uploaded by kharwell (License 6464) misc-2.diff uploaded by file (License 5000) chan_mild-3.diff uploaded by file (License 5000) chan_obscure.diff uploaded by file (License 5000) jingle.diff uploaded by file (License 5000) funcs.diff uploaded by file (License 5000) formats.diff uploaded by file (License 5000) core.diff uploaded by file (License 5000) bridges.diff uploaded by file (License 5000) mf-codecs-2.diff uploaded by file (License 5000) mf-app_fax.diff uploaded by file (License 5000) mf-apps-3.diff uploaded by file (License 5000) media-formats-3.diff uploaded by file (License 5000) ASTERISK-23715 rb3713.patch uploaded by coreyfarrell (License 5909) rb3689.patch uploaded by mjordan (License 6283) ASTERISK-23957 rb3722.patch uploaded by mjordan (License 6283) mf-attributes-3.diff uploaded by file (License 5000) ASTERISK-23958 Tested by: jrose rb3822.patch uploaded by coreyfarrell (License 5909) rb3800.patch uploaded by jrose (License 6182) chan_sip.diff uploaded by mjordan (License 6283) rb3747.patch uploaded by jrose (License 6182) ASTERISK-23959 #close Tested by: sgriepentrog, mjordan, coreyfarrell sip_cleanup.diff uploaded by opticron (License 6273) chan_sip_caps.diff uploaded by mjordan (License 6283) rb3751.patch uploaded by coreyfarrell (License 5909) chan_sip-3.diff uploaded by file (License 5000) ASTERISK-23960 #close Tested by: opticron direct_media.diff uploaded by opticron (License 6273) pjsip-direct-media.diff uploaded by file (License 5000) format_cap_remove.diff uploaded by opticron (License 6273) media_format_fixes.diff uploaded by opticron (License 6273) chan_pjsip-2.diff uploaded by file (License 5000) ASTERISK-23966 #close Tested by: rmudgett rb3803.patch uploaded by rmudgetti (License 5621) chan_dahdi.diff uploaded by file (License 5000) ASTERISK-24064 #close Tested by: coreyfarrell, mjordan, opticron, file, rmudgett, sgriepentrog, jrose rb3814.patch uploaded by rmudgett (License 5621) moh_cleanup.diff uploaded by opticron (License 6273) bridge_leak.diff uploaded by opticron (License 6273) translate.diff uploaded by file (License 5000) rb3795.patch uploaded by rmudgett (License 5621) tls_fix.diff uploaded by mjordan (License 6283) fax-mf-fix-2.diff uploaded by file (License 5000) rtp_transfer_stuff uploaded by mjordan (License 6283) rb3787.patch uploaded by rmudgett (License 5621) media-formats-explicit-translate-format-3.diff uploaded by file (License 5000) format_cache_case_fix.diff uploaded by opticron (License 6273) rb3774.patch uploaded by rmudgett (License 5621) rb3775.patch uploaded by rmudgett (License 5621) rtp_engine_fix.diff uploaded by opticron (License 6273) rtp_crash_fix.diff uploaded by opticron (License 6273) rb3753.patch uploaded by mjordan (License 6283) rb3750.patch uploaded by mjordan (License 6283) rb3748.patch uploaded by rmudgett (License 5621) media_format_fixes.diff uploaded by opticron (License 6273) rb3740.patch uploaded by mjordan (License 6283) rb3739.patch uploaded by mjordan (License 6283) rb3734.patch uploaded by mjordan (License 6283) rb3689.patch uploaded by mjordan (License 6283) rb3674.patch uploaded by coreyfarrell (License 5909) rb3671.patch uploaded by coreyfarrell (License 5909) rb3667.patch uploaded by coreyfarrell (License 5909) rb3665.patch uploaded by mjordan (License 6283) rb3625.patch uploaded by coreyfarrell (License 5909) rb3602.patch uploaded by coreyfarrell (License 5909) format_compatibility-2.diff uploaded by file (License 5000) core.diff uploaded by file (License 5000) git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@419044 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2014-07-20 22:06:33 +00:00
tech->capabilities = ast_format_cap_alloc(AST_FORMAT_CAP_FLAG_DEFAULT);
if (!tech->capabilities) {
return -1;
}
media formats: re-architect handling of media for performance improvements In the old times media formats were represented using a bit field. This was fast but had a few limitations. 1. Asterisk was limited in how many formats it could handle. 2. Formats, being a bit field, could not include any attribute information. A format was strictly its type, e.g., "this is ulaw". This was changed in Asterisk 10 (see https://wiki.asterisk.org/wiki/display/AST/Media+Architecture+Proposal for notes on that work) which led to the creation of the ast_format structure. This structure allowed Asterisk to handle attributes and bundle information with a format. Additionally, ast_format_cap was created to act as a container for multiple formats that, together, formed the capability of some entity. Another mechanism was added to allow logic to be registered which performed format attribute negotiation. Everywhere throughout the codebase Asterisk was changed to use this strategy. Unfortunately, in software, there is no free lunch. These new capabilities came at a cost. Performance analysis and profiling showed that we spend an inordinate amount of time comparing, copying, and generally manipulating formats and their related structures. Basic prototyping has shown that a reasonably large performance improvement could be made in this area. This patch is the result of that project, which overhauled the media format architecture and its usage in Asterisk to improve performance. Generally, the new philosophy for handling formats is as follows: * The ast_format structure is reference counted. This removed a large amount of the memory allocations and copying that was done in prior versions. * In order to prevent race conditions while keeping things performant, the ast_format structure is immutable by convention and lock-free. Violate this tenet at your peril! * Because formats are reference counted, codecs are also reference counted. The Asterisk core generally provides built-in codecs and caches the ast_format structures created to represent them. Generally, to prevent inordinate amounts of module reference bumping, codecs and formats can be added at run-time but cannot be removed. * All compatibility with the bit field representation of codecs/formats has been moved to a compatibility API. The primary user of this representation is chan_iax2, which must continue to maintain its bit-field usage of formats for interoperability concerns. * When a format is negotiated with attributes, or when a format cannot be represented by one of the cached formats, a new format object is created or cloned from an existing format. That format may have the same codec underlying it, but is a different format than a version of the format with different attributes or without attributes. * While formats are reference counted objects, the reference count maintained on the format should be manipulated with care. Formats are generally cached and will persist for the lifetime of Asterisk and do not explicitly need to have their lifetime modified. An exception to this is when the user of a format does not know where the format came from *and* the user may outlive the provider of the format. This occurs, for example, when a format is read from a channel: the channel may have a format with attributes (hence, non-cached) and the user of the format may last longer than the channel (if the reference to the channel is released prior to the format's reference). For more information on this work, see the API design notes: https://wiki.asterisk.org/wiki/display/AST/Media+Format+Rewrite Finally, this work was the culmination of a large number of developer's efforts. Extra thanks goes to Corey Farrell, who took on a large amount of the work in the Asterisk core, chan_sip, and was an invaluable resource in peer reviews throughout this project. There were a substantial number of patches contributed during this work; the following issues/patch names simply reflect some of the work (and will cause the release scripts to give attribution to the individuals who work on them). Reviews: https://reviewboard.asterisk.org/r/3814 https://reviewboard.asterisk.org/r/3808 https://reviewboard.asterisk.org/r/3805 https://reviewboard.asterisk.org/r/3803 https://reviewboard.asterisk.org/r/3801 https://reviewboard.asterisk.org/r/3798 https://reviewboard.asterisk.org/r/3800 https://reviewboard.asterisk.org/r/3794 https://reviewboard.asterisk.org/r/3793 https://reviewboard.asterisk.org/r/3792 https://reviewboard.asterisk.org/r/3791 https://reviewboard.asterisk.org/r/3790 https://reviewboard.asterisk.org/r/3789 https://reviewboard.asterisk.org/r/3788 https://reviewboard.asterisk.org/r/3787 https://reviewboard.asterisk.org/r/3786 https://reviewboard.asterisk.org/r/3784 https://reviewboard.asterisk.org/r/3783 https://reviewboard.asterisk.org/r/3778 https://reviewboard.asterisk.org/r/3774 https://reviewboard.asterisk.org/r/3775 https://reviewboard.asterisk.org/r/3772 https://reviewboard.asterisk.org/r/3761 https://reviewboard.asterisk.org/r/3754 https://reviewboard.asterisk.org/r/3753 https://reviewboard.asterisk.org/r/3751 https://reviewboard.asterisk.org/r/3750 https://reviewboard.asterisk.org/r/3748 https://reviewboard.asterisk.org/r/3747 https://reviewboard.asterisk.org/r/3746 https://reviewboard.asterisk.org/r/3742 https://reviewboard.asterisk.org/r/3740 https://reviewboard.asterisk.org/r/3739 https://reviewboard.asterisk.org/r/3738 https://reviewboard.asterisk.org/r/3737 https://reviewboard.asterisk.org/r/3736 https://reviewboard.asterisk.org/r/3734 https://reviewboard.asterisk.org/r/3722 https://reviewboard.asterisk.org/r/3713 https://reviewboard.asterisk.org/r/3703 https://reviewboard.asterisk.org/r/3689 https://reviewboard.asterisk.org/r/3687 https://reviewboard.asterisk.org/r/3674 https://reviewboard.asterisk.org/r/3671 https://reviewboard.asterisk.org/r/3667 https://reviewboard.asterisk.org/r/3665 https://reviewboard.asterisk.org/r/3625 https://reviewboard.asterisk.org/r/3602 https://reviewboard.asterisk.org/r/3519 https://reviewboard.asterisk.org/r/3518 https://reviewboard.asterisk.org/r/3516 https://reviewboard.asterisk.org/r/3515 https://reviewboard.asterisk.org/r/3512 https://reviewboard.asterisk.org/r/3506 https://reviewboard.asterisk.org/r/3413 https://reviewboard.asterisk.org/r/3410 https://reviewboard.asterisk.org/r/3387 https://reviewboard.asterisk.org/r/3388 https://reviewboard.asterisk.org/r/3389 https://reviewboard.asterisk.org/r/3390 https://reviewboard.asterisk.org/r/3321 https://reviewboard.asterisk.org/r/3320 https://reviewboard.asterisk.org/r/3319 https://reviewboard.asterisk.org/r/3318 https://reviewboard.asterisk.org/r/3266 https://reviewboard.asterisk.org/r/3265 https://reviewboard.asterisk.org/r/3234 https://reviewboard.asterisk.org/r/3178 ASTERISK-23114 #close Reported by: mjordan media_formats_translation_core.diff uploaded by kharwell (License 6464) rb3506.diff uploaded by mjordan (License 6283) media_format_app_file.diff uploaded by kharwell (License 6464) misc-2.diff uploaded by file (License 5000) chan_mild-3.diff uploaded by file (License 5000) chan_obscure.diff uploaded by file (License 5000) jingle.diff uploaded by file (License 5000) funcs.diff uploaded by file (License 5000) formats.diff uploaded by file (License 5000) core.diff uploaded by file (License 5000) bridges.diff uploaded by file (License 5000) mf-codecs-2.diff uploaded by file (License 5000) mf-app_fax.diff uploaded by file (License 5000) mf-apps-3.diff uploaded by file (License 5000) media-formats-3.diff uploaded by file (License 5000) ASTERISK-23715 rb3713.patch uploaded by coreyfarrell (License 5909) rb3689.patch uploaded by mjordan (License 6283) ASTERISK-23957 rb3722.patch uploaded by mjordan (License 6283) mf-attributes-3.diff uploaded by file (License 5000) ASTERISK-23958 Tested by: jrose rb3822.patch uploaded by coreyfarrell (License 5909) rb3800.patch uploaded by jrose (License 6182) chan_sip.diff uploaded by mjordan (License 6283) rb3747.patch uploaded by jrose (License 6182) ASTERISK-23959 #close Tested by: sgriepentrog, mjordan, coreyfarrell sip_cleanup.diff uploaded by opticron (License 6273) chan_sip_caps.diff uploaded by mjordan (License 6283) rb3751.patch uploaded by coreyfarrell (License 5909) chan_sip-3.diff uploaded by file (License 5000) ASTERISK-23960 #close Tested by: opticron direct_media.diff uploaded by opticron (License 6273) pjsip-direct-media.diff uploaded by file (License 5000) format_cap_remove.diff uploaded by opticron (License 6273) media_format_fixes.diff uploaded by opticron (License 6273) chan_pjsip-2.diff uploaded by file (License 5000) ASTERISK-23966 #close Tested by: rmudgett rb3803.patch uploaded by rmudgetti (License 5621) chan_dahdi.diff uploaded by file (License 5000) ASTERISK-24064 #close Tested by: coreyfarrell, mjordan, opticron, file, rmudgett, sgriepentrog, jrose rb3814.patch uploaded by rmudgett (License 5621) moh_cleanup.diff uploaded by opticron (License 6273) bridge_leak.diff uploaded by opticron (License 6273) translate.diff uploaded by file (License 5000) rb3795.patch uploaded by rmudgett (License 5621) tls_fix.diff uploaded by mjordan (License 6283) fax-mf-fix-2.diff uploaded by file (License 5000) rtp_transfer_stuff uploaded by mjordan (License 6283) rb3787.patch uploaded by rmudgett (License 5621) media-formats-explicit-translate-format-3.diff uploaded by file (License 5000) format_cache_case_fix.diff uploaded by opticron (License 6273) rb3774.patch uploaded by rmudgett (License 5621) rb3775.patch uploaded by rmudgett (License 5621) rtp_engine_fix.diff uploaded by opticron (License 6273) rtp_crash_fix.diff uploaded by opticron (License 6273) rb3753.patch uploaded by mjordan (License 6283) rb3750.patch uploaded by mjordan (License 6283) rb3748.patch uploaded by rmudgett (License 5621) media_format_fixes.diff uploaded by opticron (License 6273) rb3740.patch uploaded by mjordan (License 6283) rb3739.patch uploaded by mjordan (License 6283) rb3734.patch uploaded by mjordan (License 6283) rb3689.patch uploaded by mjordan (License 6283) rb3674.patch uploaded by coreyfarrell (License 5909) rb3671.patch uploaded by coreyfarrell (License 5909) rb3667.patch uploaded by coreyfarrell (License 5909) rb3665.patch uploaded by mjordan (License 6283) rb3625.patch uploaded by coreyfarrell (License 5909) rb3602.patch uploaded by coreyfarrell (License 5909) format_compatibility-2.diff uploaded by file (License 5000) core.diff uploaded by file (License 5000) git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@419044 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2014-07-20 22:06:33 +00:00
ast_format_cap_append_by_type(tech->capabilities, AST_MEDIA_TYPE_UNKNOWN);
if (ast_channel_register(tech)) {
ast_log(LOG_ERROR, "Unable to register channel technology %s(%s).\n",
tech->type, tech->description);
return -1;
}
return 0;
}
/*! \brief Called when module is being unloaded */
static int unload_module(void)
{
ast_unregister_application(app);
ast_unregister_application(app2);
ast_custom_function_unregister(&confbridge_function);
ast_custom_function_unregister(&confbridge_info_function);
ast_custom_function_unregister(&confbridge_channels_function);
ast_cli_unregister_multiple(cli_confbridge, ARRAY_LEN(cli_confbridge));
ast_manager_unregister("ConfbridgeList");
ast_manager_unregister("ConfbridgeListRooms");
ast_manager_unregister("ConfbridgeMute");
ast_manager_unregister("ConfbridgeUnmute");
ast_manager_unregister("ConfbridgeKick");
ast_manager_unregister("ConfbridgeUnlock");
ast_manager_unregister("ConfbridgeLock");
ast_manager_unregister("ConfbridgeStartRecord");
ast_manager_unregister("ConfbridgeStopRecord");
ast_manager_unregister("ConfbridgeSetSingleVideoSrc");
/* Unsubscribe from stasis confbridge message type and clean it up. */
manager_confbridge_shutdown();
/* Get rid of the conference bridges container. Since we only allow dynamic ones none will be active. */
ao2_cleanup(conference_bridges);
conference_bridges = NULL;
conf_destroy_config();
unregister_channel_tech(conf_announce_get_tech());
unregister_channel_tech(conf_record_get_tech());
return 0;
}
/*!
* \brief Load the module
*
* Module loading including tests for configuration or dependencies.
* This function can return AST_MODULE_LOAD_FAILURE, AST_MODULE_LOAD_DECLINE,
* or AST_MODULE_LOAD_SUCCESS. If a dependency or environment variable fails
* tests return AST_MODULE_LOAD_FAILURE. If the module can not load the
* configuration file or other non-critical problem return
* AST_MODULE_LOAD_DECLINE. On success return AST_MODULE_LOAD_SUCCESS.
*/
static int load_module(void)
{
int res = 0;
if (conf_load_config()) {
ast_log(LOG_ERROR, "Unable to load config. Not loading module.\n");
return AST_MODULE_LOAD_DECLINE;
}
if (register_channel_tech(conf_record_get_tech())
|| register_channel_tech(conf_announce_get_tech())) {
unload_module();
return AST_MODULE_LOAD_DECLINE;
}
/* Create a container to hold the conference bridges */
conference_bridges = ao2_container_alloc_hash(AO2_ALLOC_OPT_LOCK_MUTEX, 0,
CONFERENCE_BRIDGE_BUCKETS,
conference_bridge_hash_cb, NULL, conference_bridge_cmp_cb);
if (!conference_bridges) {
unload_module();
return AST_MODULE_LOAD_DECLINE;
}
/* Setup manager stasis subscriptions */
res |= manager_confbridge_init();
res |= ast_register_application_xml(app, confbridge_exec);
res |= ast_register_application_xml(app2, confkick_exec);
res |= ast_custom_function_register_escalating(&confbridge_function, AST_CFE_WRITE);
res |= ast_custom_function_register(&confbridge_info_function);
res |= ast_custom_function_register(&confbridge_channels_function);
res |= ast_cli_register_multiple(cli_confbridge, ARRAY_LEN(cli_confbridge));
res |= ast_manager_register_xml("ConfbridgeList", EVENT_FLAG_REPORTING, action_confbridgelist);
res |= ast_manager_register_xml("ConfbridgeListRooms", EVENT_FLAG_REPORTING, action_confbridgelistrooms);
res |= ast_manager_register_xml("ConfbridgeMute", EVENT_FLAG_CALL, action_confbridgemute);
res |= ast_manager_register_xml("ConfbridgeUnmute", EVENT_FLAG_CALL, action_confbridgeunmute);
res |= ast_manager_register_xml("ConfbridgeKick", EVENT_FLAG_CALL, action_confbridgekick);
res |= ast_manager_register_xml("ConfbridgeUnlock", EVENT_FLAG_CALL, action_confbridgeunlock);
res |= ast_manager_register_xml("ConfbridgeLock", EVENT_FLAG_CALL, action_confbridgelock);
res |= ast_manager_register_xml("ConfbridgeStartRecord", EVENT_FLAG_SYSTEM, action_confbridgestartrecord);
res |= ast_manager_register_xml("ConfbridgeStopRecord", EVENT_FLAG_SYSTEM, action_confbridgestoprecord);
res |= ast_manager_register_xml("ConfbridgeSetSingleVideoSrc", EVENT_FLAG_CALL, action_confbridgesetsinglevideosrc);
if (res) {
unload_module();
return AST_MODULE_LOAD_DECLINE;
}
return AST_MODULE_LOAD_SUCCESS;
}
static int reload(void)
{
return conf_reload_config();
}
AST_MODULE_INFO(ASTERISK_GPL_KEY, AST_MODFLAG_LOAD_ORDER, "Conference Bridge Application",
.support_level = AST_MODULE_SUPPORT_CORE,
.load = load_module,
.unload = unload_module,
.reload = reload,
.load_pri = AST_MODPRI_DEVSTATE_PROVIDER,
.optional_modules = "codec_speex,func_jitterbuffer",
);