Kevin P. Fleming [Mon, 23 Jul 2012 14:51:45 +0000 (14:51 +0000)]
Blocked revisions 370361
........
Free any datastores attached to dummy channels.
Revision 370205 added the use of a datastore attached to a dummy channel to
resolve a memory leak, but ast_dummy_channel_destructor() in this branch did
not free datastores, resulting in a continued (but slightly smaller) memory
leak. This patch backports the change to free said datastores from the Asterisk
trunk.
(related to issue AST-916)
........
Merged revisions 370360 from http://svn.asterisk.org/svn/asterisk/branches/1.8
Terry Wilson [Sat, 21 Jul 2012 13:25:26 +0000 (13:25 +0000)]
Fix segfault introduced by conversion to ACO API
The value "none" is specified in the config file as a valid value for
the "video_mode" option. The code prior to the ACO conversion did not
check for "none", but just ignored it and relied on the default zero
value. The parsing with ACO is more strict, so without handling
"none" specifically, parsing would fail.
When parsing failed, but the module loaded anyway, the config info
would never be stored, and one place in the code did not check for
this case and would segfault. It was also possible that the
aco_info struct's internals would be destroyed and used as well.
This patch keeps the module from loading after parse failures, adds
the "none" option to "video_mode", registers CLI functions only
after parsing has completed, checks the config data for NULL before
accessing it, and returns -1 on some allocation failures when
initializing.
(closes issue ASTERISK-20159)
Reported by: Birger "WIMPy" Harzenetter
Tested by: Birger "WIMPy" Harzenetter
Patches:
confbridge_fix3.txt uploaded by Terry Wilson
Matthew Jordan [Fri, 20 Jul 2012 19:08:47 +0000 (19:08 +0000)]
Clean up ManagerEvent Dial documentation
The paragraph describing the SubEvent belongs with the SubEvent parameter
itself, and not with its enum values. The order of parsing was placing
the description after the last enum, which isn't correct.
Kinsey Moore [Fri, 20 Jul 2012 15:48:55 +0000 (15:48 +0000)]
Add hangupcause translation support
The HANGUPCAUSE hash (trunk only) meant to replace SIP_CAUSE has now
been replaced with the HANGUPCAUSE and HANGUPCAUSE_KEYS dialplan
functions to better facilitate access to the AST_CAUSE translations
for technology-specific cause codes. The HangupCauseClear application
has also been added to remove this data from the channel.
Matthew Jordan [Thu, 19 Jul 2012 22:17:13 +0000 (22:17 +0000)]
Add the ability to specify technology specific documentation
A number of applications/AMI commands in Asterisk have specific behavioral
differences depending on the resource or channel technology those
applications are executed on. For example, the MessageSend application/
command is technology agnostic, but how the channel drivers that support
that functionality behave is dependant on the protocols and channel
driver implementation. Prior to this patch, those details were either
documented in the application/command documentation itself, or were left
undocumented.
This patch adds a new element to the documentation schema, <info/>. An info
node is essentially a piece of technology specific reference information that
can be included by any top level XML documentation node. For example, the
MessageSend application can now include XMPP/SIP specific information, where
that technology specific information can be defined in chan_motif/res_xmpp/
chan_sip. Likewise, that information can also be included in the MessageSend
AMI command.
Matthew Jordan [Thu, 19 Jul 2012 22:08:20 +0000 (22:08 +0000)]
Fix compilation error when MALLOC_DEBUG is enabled
To fix a memory leak in CEL, a channel datastore was introduced whose
destruction function pointer was pointed to the ast_free macro. Without
MALLOC_DEBUG enabled this compiles as fine, as ast_free is defined as free.
With MALLOC_DEBUG enabled, however, ast_free takes on a definition from a
different place then utils.h, and became undefined. This patch resolves this
by using a reference to ast_free_ptr. When MALLOC_DEBUG is enabled, this
calls ast_free; when MALLOC_DEBUG is not enabled, this is defined to be
ast_free, which is defined to be free.
(issue AST-916)
Reported by: Thomas Arimont
........
Merged revisions 370273 from http://svn.asterisk.org/svn/asterisk/branches/1.8
........
Merged revisions 370274 from http://svn.asterisk.org/svn/asterisk/branches/10
Matthew Jordan [Thu, 19 Jul 2012 21:45:20 +0000 (21:45 +0000)]
Handle extremely out of order RFC 2833 DTMF
The current implementation of RFC 2833 DTMF handling in res_rtp_asterisk will,
if a packet arrives out of order, drop the packet. This is to prevent
duplicate ton generation in the Asterisk core. Since the RTP layer does not
buffer data itself, this is the only option the RTP layer currently has for
handling packets that arrive out of order.
For the most part, this doesn't matter. For a particular digit, so long as a
BEGIN packet arrives before the first END packet, the digit will be produced.
If subsequent BEGIN packets arrive interleaved with the ENDs, they will be
dropped; likewise, if the BEGIN or END packets themselves are out of order,
those packets are dropped but sufficient information is conveyed to the
Asterisk core to produce the appropriate digit.
For certain sequences of DTMF packets - most notably when, for a particular
digit, an END packet arrives before any BEGIN packet for that digit - this
is a real problem. When an END arrives before any BEGINs, the END packet is
dropped - but at the same time, it causes subsequent BEGIN packets for that
digit to be ignored. When the next in order END packet arrives, it too is
dropped - Asterisk believes that there was no initial BEGIN.
The solution this patch provides is to trust the END packet to convey the
information needed for the Asterisk core to produce the DTMF digit. If we
receive an END packet, and it:
* Has a timestamp greater then the last timestamp received from an END
packet
* Does not have the same sequence number as the last received sequence
number (and is thus not an END packet retransmission)
Then we send the END frame up to the Asterisk core. It contains enough
DTMF information for Asterisk to produce the digit.
On the other hand, if we receive a BEGIN or continuation packet that occurs
with a timestamp equal to or less then the last END timestamp, then we've
received something out of order - but we already have received enough
information to produce the digit. These packets are dropped.
Much thanks goes to Olle Johansson (oej) for providing the idea for this
solution.
Review: https://reviewboard.asterisk.org/r/2033/
(closes issue ASTERISK-18404)
Reported by: Stephane Chazelas
Tested by: Matt Jordan
........
Merged revisions 370252 from http://svn.asterisk.org/svn/asterisk/branches/1.8
........
Merged revisions 370271 from http://svn.asterisk.org/svn/asterisk/branches/10
Kevin P. Fleming [Wed, 18 Jul 2012 19:18:40 +0000 (19:18 +0000)]
Resolve severe memory leak in CEL logging modules.
A customer reported a significant memory leak using Asterisk 1.8. They
have tracked it down to ast_cel_fabricate_channel_from_event() in
main/cel.c, which is called by both in-tree CEL logging modules
(cel_custom.c and cel_sqlite3_custom.c) for each and every CEL event
that they log.
The cause was an incorrect assumption about how data attached to an
ast_channel would be handled when the channel is destroyed; the data
is now stored in a datastore attached to the channel, which is
destroyed along with the channel at the proper time.
(closes issue AST-916)
Reported by: Thomas Arimont
Review: https://reviewboard.asterisk.org/r/2053/
........
Merged revisions 370205 from http://svn.asterisk.org/svn/asterisk/branches/1.8
........
Merged revisions 370206 from http://svn.asterisk.org/svn/asterisk/branches/10
Kevin P. Fleming [Wed, 18 Jul 2012 17:18:20 +0000 (17:18 +0000)]
Ensure that all ast_datastore_info structures are 'const'.
While addressing a bug, I came across a instance of 'struct ast_datastore_info'
that was not declared 'const'. Since the API already expects them to be
'const', this patch changes the declarations of all existing instances
that were not already declared that way.
........
Merged revisions 370183 from http://svn.asterisk.org/svn/asterisk/branches/1.8
........
Merged revisions 370184 from http://svn.asterisk.org/svn/asterisk/branches/10
Fix a crash in pjnath when starting an ICE connectivity check and immediately destroying the ICE session.
The initial ICE connectivity check is scheduled as a timer item that is to be executed immediately. It is possible for this timer item to start executing while the ICE session it is working on is destroyed. To reduce the chance of this any timer items that need to be immediately executed will be executed within the thread that has started the initial ICE connectivity check.
Fix a crash occurring as a result of excess stack usage.
This fix involves moving the allocation of some temporary codec structures to the heap and also reduces the number of maximum payloads to something more sane for both regular and low memory builds.
Add pubsub unsubscription support so subscriptions do not linger for MWI and device state progatation.
The pubsub code did not attempt to remove subscriptions at all. This has now changed so that if a client is being disconnected it will unsubscribe. It will also unsubscribe at connection time so if it unexpectedly disconnected duplicate subscriptions will not occur.
Fix a crash as a result of propagating MWI or device state over XMPP when the client is disconnected.
The MWI and device state propagation code wrongly assumes that an XMPP client connection will remain established at all times. This fix corrects that by making the lifetime of the subscription the same as the lifetime of the connection itself. As the connection is established and disconnected the subscription itself is created and destroyed.
Fix an issue where a service discovery request could crash Asterisk.
A server sending a service discovery request to us may or may not put a from attribute in the message. If the from attribute is present use it in the to attribute for the result. If the from attribute is not present do not add a to attribute.
Add support for parsing SDP attributes, generating SDP attributes, and passing it through.
This support includes codecs such as H.263, H.264, SILK, and CELT. You are able to set up a call and have attribute information pass. This should help considerably with video calls.
contrib/scripts/live_ast currently assumes that it is being run from the
top-level directory of the source tree. It creates a script that will
also set the working directory.
This fix avoids the need to set the working directory if the caller sets
LIVE_AST_BASE_DIR instead.
It relies on realpath for that. If realpath is not available, it will
fall back to the original behaviour.
Kinsey Moore [Thu, 12 Jul 2012 19:05:11 +0000 (19:05 +0000)]
Prevent double uri_escaping in chan_sip when pedantic is enabled
If pedantic mode is enabled, outbound invites will have double-escaped
contacts. This avoids setting an already-escaped string into a field
where it is expected to be unescaped.
(closes issue ASTERISK-20023)
Reported by: Walter Doekes
........
Merged revisions 369993 from http://svn.asterisk.org/svn/asterisk/branches/1.8
........
Merged revisions 369994 from http://svn.asterisk.org/svn/asterisk/branches/10
Michael L. Young [Thu, 12 Jul 2012 14:38:44 +0000 (14:38 +0000)]
Correct Documentation For DEC Function
The documentation for DEC in func_math.c was incorrect. Looks like a copy and
paste error.
(Closes issue ASTERISK-20095)
Reported by: Billy Chia
Tested by: Michael L. Young
Patches:
func_math.patch uploaded by Billy Chia (license 6381)
........
Merged revisions 369970 from http://svn.asterisk.org/svn/asterisk/branches/1.8
........
Merged revisions 369971 from http://svn.asterisk.org/svn/asterisk/branches/10
Michael L. Young [Thu, 12 Jul 2012 14:27:56 +0000 (14:27 +0000)]
Correct Documentation For DEC Function
The documentation for DEC in func_math.c was incorrect. Looks like a copy and
paste error.
(Closes issue ASTERISK-20095)
Reported by: Billy Chia
Tested by: Michael L. Young
Patches:
func_math.patch uploaded by Billy Chia (license 6381)
........
Merged revisions 369970 from http://svn.asterisk.org/svn/asterisk/branches/1.8
Jonathan Rose [Wed, 11 Jul 2012 18:33:36 +0000 (18:33 +0000)]
Named ACLs: Introduces a system for creating and sharing ACLs
This patch adds Named ACL functionality to Asterisk. This allows system
administrators to define an ACL and refer to it by a unique name. Configurable
items can then refer to that name when specifying access control lists.
It also includes updates to all core supported consumers of ACLs. That includes
manager, chan_sip, and chan_iax2. This feature is based on the deluxepine-trunk
by Olle E. Johansson and provides a subset of the Named ACL functionality
implemented in that branch. For more information on this feature, see acl.conf
and/or the Asterisk wiki.
Allow the REALTIME() function to report errors back to the caller.
Also, do more error checking on the arguments specified to the REALTIME()
function and clarify the documentation. While I was editing the file, a
few coding guidelines fixups, as well.
Matthew Jordan [Wed, 11 Jul 2012 17:14:45 +0000 (17:14 +0000)]
Don't perform an XInclude to a document node that may not always be present
Because some of the manager events are defined in the top of the source, due
to the macro calls not containing all necessary information to have the
documentation colocated with the call itself, several include statements were
failing when built with 'make'. While this did not cause any problems in
compilation or validation, it did result in a number of warnings being dumped
to stderr.
This patch changes those references such that they always resolve, regardless
of the documentation build options.
Do not consider failure to read the configuration file in chan_motif to be a show stopper for loading Asterisk by returning decline instead of failure.
(closes issue ASTERISK-20103)
Reported by: Terry Wilson
Matthew Jordan [Wed, 11 Jul 2012 02:06:05 +0000 (02:06 +0000)]
Fix validation errors when producing documentation using default build script
The awk script parses out the first instance of the DOCUMENTATION tag that it
finds within a file. If a file did not previously have a DOCUMENTATION tag
but received one due to it having an AMI event, then the XML fragment
associated with the AMI event was erroneously placed in the resulting XML
file. Without the python scripts, these XML fragments will not validate.
This patch adds DOCUMENTATION tags at the top of those files that did
not previously have them to prevent the awk script from pulling AMI event
documentation.
Matthew Jordan [Tue, 10 Jul 2012 22:26:27 +0000 (22:26 +0000)]
Add some additional documentation for core AMI events
This patch adds some basic documentation for a number of modules. This
includes core source files in Asterisk (those in main), as well as
chan_agent, chan_dahdi, chan_local, sig_analog, and sig_pri. The DTD
has also been updated to allow referencing of AMI commands.
Kinsey Moore [Tue, 10 Jul 2012 15:36:37 +0000 (15:36 +0000)]
Fix failing SDP_offer_answer test
Asterisk now generates image stream declinations with the same
transport case that it used to before the stream declination
improvements. (udptl vs UDPTL)
Respect codec preference order when adding codecs to a media description.
This change allows an endpoint in motif.conf to be configured with a preference of G.722 and fallback of ulaw. With Google this allows communication with Google Talk clients to use G.722 while when using Google Voice ulaw will be used.
Matthew Jordan [Tue, 10 Jul 2012 13:34:15 +0000 (13:34 +0000)]
Fix initial loading problem with res_curl
When the OpenSSL duplicate initialization issues were resolved in r351447,
res_curl could fail to load if it checked SSL_library_init after SSL
initialization completed. This is due to the SSL_library_init stub returning
a value of 0 for success, as opposed to a value of 1. OpenSSL uses a value of
1 to indicate success - in fact, SSL_library_init is documented to always return
1. Interestingly, the CURL libraries actually checked the return value - the fact
that nothing else that depends on OpenSSL was having problems loading probably means
they don't check the return value.
When receiving a STUN binding request send one out as the Google Talk client uses this as a method to determine if the remote party is still reachable or not.
Failure to do this results in the Google Talk client ignoring RTP packets after a specific period of time. This is also done as a result of receiving a STUN binding request so that the username information can be used from the inbound request, thus not requiring it to be stored on a per candidate basis.
Jonathan Rose [Mon, 9 Jul 2012 14:54:22 +0000 (14:54 +0000)]
chan_sip: Fix small behavioral change accidentally introduced in r369750
When removing the warning for AST_CONTROL_FLASH from sip_indicate, I also
inadvertently changed the return value, which would likely make the indication
not be sent in audio. This fixes that while still removing the warning message.
........
Merged revisions 369792 from http://svn.asterisk.org/svn/asterisk/branches/1.8
........
Merged revisions 369793 from http://svn.asterisk.org/svn/asterisk/branches/10
Add a new unified Jingle, Google Jingle, and Google Talk channel driver written from scratch called chan_motif.
This channel driver is a replacement for both chan_gtalk and chan_jingle but adds additional features not found in either.
These features include full configuration reload, video, full codec support, bidirectional cause code mapping, hold,
unhold, and ringing indication. It is also compliant with the current published Jingle and Google Jingle specifications.
The original Google Talk protocol is also supported for Google Voice interoperability.
You may ask yourself though where the name motif comes from... and I would say to you... music!
motif: a perceivable or salient recurring fragment or succession of notes
Kinsey Moore [Fri, 6 Jul 2012 22:03:44 +0000 (22:03 +0000)]
Remove unnecessary generation of informational cause frames
It is not necessary to generate information cause code frames on every
protocol event that occurs. This removes all the instances where the
frame was not conveying a cause code and was instead just conveying a
protocol-specific message. This also corrects the generation of the
message associated with disconnects for MFC/R2 to use the MFC/R2
specific text for the disconnect cause.
Jonathan Rose [Fri, 6 Jul 2012 21:28:26 +0000 (21:28 +0000)]
chan_sip: Add case for FLASH control frames so that we don't display a warning.
chan_sip channels can receive flash control frames when connected to analog
phones and possibly for other reasons. There really isn't a reason to warn when
these frames are received, we can safely ignore them.
Patches:
dahdi_sip_flash.diff uploaded by Jonathan Rose (license 6182)
........
Merged revisions 369750 from http://svn.asterisk.org/svn/asterisk/branches/1.8
........
Merged revisions 369751 from http://svn.asterisk.org/svn/asterisk/branches/10
Mark Michelson [Fri, 6 Jul 2012 18:49:17 +0000 (18:49 +0000)]
Remove a superfluous and dangerous freeing of an SSL_CTX.
The problem here is that multiple server sessions share
a SSL_CTX. When one session ended, the SSL_CTX would be
freed and set NULL, leaving the other sessions unable to
function.
The code being removed is superfluous because the SSL_CTX
structures for servers will be properly freed when ast_ssl_teardown
is called.
(closes issue ASTERISK-20074)
Reported by Trevor Helmsley
Patches:
ASTERISK-20074.diff uploaded by Mark Michelson (license #5049)
Testers:
Trevor Helmsley
........
Merged revisions 369731 from http://svn.asterisk.org/svn/asterisk/branches/1.8
........
Merged revisions 369732 from http://svn.asterisk.org/svn/asterisk/branches/10
Mark Michelson [Fri, 6 Jul 2012 15:31:52 +0000 (15:31 +0000)]
Fix bridging thread leak.
The bridge thread was exiting but was never being
reaped using pthread_join(). This has been fixed now
by calling pthread_join() in ast_bridge_destroy().
(closes issue ASTERISK-19834)
Reported by Marcus Hunger
Import revision 4196 from pjproject trunk. Fix a crash issue when starting ICE connectivity checks and immediately destroying the ICE session. This was exposed by the SIP CCSS test.
Full fix for this issue will be worked on as a medium to long term roadmap item.
pjroject issue viewable at https://trac.pjsip.org/repos/ticket/1548
Matthew Jordan [Thu, 5 Jul 2012 21:36:41 +0000 (21:36 +0000)]
Add 'stun show status' command
This patch adds a new CLI command, 'stun show status'. This command will show
a table describing all known STUN servers and statuses.
(closes issue ASTERISK-18046)
Reported by: Jeremy Kister
Tested by: Jeremy Kister
patches:
(stun-show-status-v4-trunk.patch license #6232 uploaded by Jeremy Kister)
Kinsey Moore [Thu, 5 Jul 2012 19:36:21 +0000 (19:36 +0000)]
AST-2012-011: Resolve heap corruption issue with voicemail
The heard and deleted arrays in the voicemail state structure were not
handled properly following the memory leak fix in r354890 and a fix for
an invalid free in r356797. This could result in accessing and writing
into freed memory. The allocation for these arrays has been reworked
to avoid the possibility of invalid frees, access of freed memory, and
crashes that were occurring as a result of this.
Locking around accesses and modifications of the voicemail state
structure members dh_arraysize, heard, and deleted has been added to
prevent simultaneous modification and access when IMAP storage is in
use. If IMAP storage is not in use, this locking is not compiled in.
Review: https://reviewboard.asterisk.org/r/1994/
(closes issue ASTERISK-19923)
Reported by: Dan Delaney
Tested by: Dan Delaney, Julian Yap
Patches:
vm_alloc_fix.diff uploaded by kmoore (license 6273)
........
Merged revisions 369652 from http://svn.asterisk.org/svn/asterisk/branches/1.8
........
Merged revisions 369653 from http://svn.asterisk.org/svn/asterisk/branches/10
Matthew Jordan [Thu, 5 Jul 2012 17:03:43 +0000 (17:03 +0000)]
Do not send a BYE when a provisional response arrives during a re-INVITE
Commits r369557 and r369579 were done to improve handling of re-INVITEs
when the UA that was supposed to receive the re-INVITE fails to respond.
A limitation of those patches occurred when a UA sent a provisional
response to the re-INVITE. This triggered a sending of a BYE in
check_pending. This patch tweaks the handling of the re-INVITE such that
a BYE is not sent in response to those messages.
(issue ASTERISK-19992)
Reported by: Steve Davies
Tested by: Steve Davies
patches:
(reinvite_tweak.diff license #5012 by Steve Davies)
........
Merged revisions 369626 from http://svn.asterisk.org/svn/asterisk/branches/1.8
........
Merged revisions 369627 from http://svn.asterisk.org/svn/asterisk/branches/10
Alexandr Anikin [Wed, 4 Jul 2012 21:42:05 +0000 (21:42 +0000)]
Added direct media support to ooh323 channel driver
options are documented in config sample
sample config rename to proper name - ooh323.conf
To change media address ooh323 send empty TCS if there was
completed TCS exchange or send facility forwardedelements
with new fast start proposal if not.
Then close transmit logical channels and renew TCS exchange.
If new fast start proposal is received then ooh323 stack call back
channel driver routine to change rtp address in the rtp instance.
If empty TCS is received then close transmit logical channels and
renew TCS exchange
Terry Wilson [Tue, 3 Jul 2012 17:07:20 +0000 (17:07 +0000)]
More improvements to re-INVITEs timing out after a provisional response
There is no need to call check_pendings() on a final response to an INVITE
when destroying the scheduler entry as it will be done later during normal
processing.
(issue ASTERISK-19992)
........
Merged revisions 369579 from http://svn.asterisk.org/svn/asterisk/branches/1.8
........
Merged revisions 369580 from http://svn.asterisk.org/svn/asterisk/branches/10
Terry Wilson [Tue, 3 Jul 2012 14:49:19 +0000 (14:49 +0000)]
Better handle re-INVITEs with provisional but no final repsonses
A previous attempt at fixing this issue had negative side effects related
to attended transfers which this patch should resolve. Many thanks to
Steve Davies for all of the good suggestions and testing.
(closes issue ASTERISK-19992)
Reported by: Steve Davies
Tested by: Steve Davies, Terry Wilson
Review: https://reviewboard.asterisk.org/r/2009/
........
Merged revisions 369557 from http://svn.asterisk.org/svn/asterisk/branches/1.8
........
Merged revisions 369558 from http://svn.asterisk.org/svn/asterisk/branches/10
Add a cleaned up drop-in replacement for res_jabber called res_xmpp. This provides the same externally facing functionality but is implemented differently internally.
This is currently not built by default but this will be changed once chan_jingle2 (insert actual name in your head when reading this after it has been merged)
is in the tree.
Richard Mudgett [Fri, 29 Jun 2012 17:02:32 +0000 (17:02 +0000)]
Hangup handlers - Dialplan subroutines that run when the channel hangs up.
Hangup handlers are an alternative to the h extension. They can be used
in addition to the h extension. The idea is to attach a Gosub routine to
a channel that will execute when the call hangs up. Whereas which h
extension gets executed depends on the location of dialplan execution when
the call hangs up, hangup handlers are attached to the call channel. You
can attach multiple handlers that will execute in the order of most
recently added first.
(closes issue ASTERISK-19549)
Reported by: Mark Murawski
Tested by: rmudgett
projects / thirdparty / asterisk.git / log
