chan_pjsip_indicate was missing a case for the recently added
AST_CONTROL_STREAM_TOPOLOGY_CHANGED condition and was returning an
error and causing the call to be hung up instead of just ignoring
it.

ASTERISK-27260
Reported by: Daniel Heckl

Change-Id: I4fecbb00a0b8a853da85155065c1a6bddf235e80

pubsub_on_rx_notify_request wasn't checking for a null
Content-Type header before checking that it was
application/simple-message-summary.

ASTERISK-27279
Reported by: Ross Beer

Change-Id: Iec2a6c4d2e74af37ff779ecc9fd35644c5c4ea52

This change makes it so that the conference recorder channel
that is created only contains audio formats and an audio stream.
This is because the underlying application used by ConfBridge to
record, MixMonitor, only allows recording audio.

Having additional streams (and in particular a video stream) can
result in clients needlessly renegotiating to add a video stream
that will never receive video.

Change-Id: I89d38aedc9205eca7741d5435e73e73bb9de97a0

ast_variables_destroy is NULL safe, so there is no need to check its
argument before passing it.

ASTERISK-25524 #close
Reported by: Jesper

Change-Id: Ib0f8057642e9d471960f1a79fd42e5a3ce587d3b

Change-Id: I7de0a5adc89824a5f2b696fc22c80fc22dff36b0

Incoming requests with non sip(s) URIs in the Request, To, From
or Contact URIs are now rejected with
PJSIP_SC_UNSUPPORTED_URI_SCHEME (416).  This is performed in
pjsip_message_filter (formerly pjsip_message_ip_updater) and is
done at pjproject's "TRANSPORT" layer before a request can even
reach the distributor.

URIs read by res_pjsip_outbound_publish from pjsip.conf are now
also checked for both length and sip(s) scheme.  Those URIs read
by outbound registration and aor were already being checked for
scheme but their error messages needed to be updated to include
scheme failure as well as length failure.

Change-Id: Ibb2f9f1d2dc7549da562af4cbd9156c44ffdd460

The Websocket implementation will steal the underlying stream of
TCP/TLS sessions. This results in an error message being output
about a stream not being present when in reality this is actually
fine.

This change moves it to a debug message instead.

Change-Id: I66cc639080b4b4599beadb4faa7d313f2721d094

* The way that we were looking at XML elements for CalDAV was extremely
  fragile, so use SAX2 for increased robustness.

* Don't complain about a 'channel' not be specified if autoreminder is
  not set. Assume that if 'channel' is not set, we don't want to be
  notified.

* Fix some truncated CLI output in 'calendar show calendar' and make the
  'Autoreminder' description a bit more clear

ASTERISK-24588 #close
Reported by: Stefan Gofferje

ASTERISK-25523 #close
Reported by: Jesper

Change-Id: I200d11afca6a47e7d97888f286977e2e69874b2c

Multicast/Unicast RTP do not use SDP so we need to use a format that
cleanly maps to one of the static RTP payload types. Without this
change, an Originate to a Multicast or Unicast channel without a format
specified would produce no audio on the receiving device.

ASTERISK-21399 #close
Reported by: Tzafrir Cohen

Change-Id: I97e332b566e85da04b0004b9b0daae746cfca0e3

A new endpoint parameter "incoming_mwi_mailbox" allows Asterisk to
receive unsolicited MWI NOTIFY requests and make them available to
other modules via the stasis message bus.

res_pjsip_pubsub has a new handler "pubsub_on_rx_mwi_notify_request"
that parses a simple-message-summary body and, if
endpoint->incoming_mwi_account is set, calls ast_publish_mwi_state
with the voice-message counts from the message.

Change-Id: I08bae3d16e77af48fcccc2c936acce8fc0ef0f3c

Change-Id: I3f20ce428777cc4ce9c13b2f808d29ff8c873998

The downgrade function was missing "_v2" at the end of the
alter column type.

Change-Id: Iaa9bcef48d6f3590ce07a61342d8e66f00263d8e

In 45744fc5, I mistakenly broke SDP media address rewriting by
misinterpreting which address was checked in the localnet comparison.

Instead of checking the remote peer address to decide whether we need
media address rewriting, we check our local media address: if it's
local, then we rewrite. This feels awkward, but works and even made
directmedia work properly if you set local_net. (For the record: for
local peers, the SDP media rewrite code is not called, so the
comparison does no harm there.)

ASTERISK-27248 #close

Change-Id: I566be1c33f4d0a689567d451ed46bab9c3861d4f

Change-Id: I43f25976aa3069793ddbe0086833965a6fb0a518

MS-SQL has no native Enum-type support and therefore
needs to work with constraints.
Since these constraints need unique names the suggested approach
referenced in the following alembic documentation has been applied:
http://bit.ly/2x9r8pb

ASTERISK-27255 #close

Change-Id: I8b579750dae0c549f1103ee50172644afb9b2f95

Fixes ${CDR(...,u)} when used in cdr_custom.conf

ASTERISK-27165 #close

Change-Id: Ia4e0b6ba93e03d27886354c279737790e2cd6a83

* WaitForSilence completes successfully if it receives no media in the
  specified timeout, but when acting as WaitForNoise that logic needs
  to be reversed.

* Use standard argument parsing macros and add some error checking for
  invalid values.

* The documentation indicated that the first argument to both
  WaitForSilence and WaitForNoise was required when it was not. Update
  the documentation to reflect that.

* Wrap up some behavior in structs to avoid boolean checks all over the
  place.

ASTERISK-24066 #close
Reported by: M vd S

Change-Id: I01d40adc5b63342bb5018a1bea2081a0aa191ef9

In handle_request_invite, when processing a pickup, a call
is made to get_sip_pvt_from_replaces to locate the pvt for
the subscription. The pvt is assumed to be valid when zero
is returned indicating no error, and is dereferenced which
can cause a crash if it was not found.

This change checks the not found case and returns -1 which
allows the calling code to fail appropriately.

ASTERISK-27217 #close
Reported-by: Bryan Walters

Change-Id: I6bee92b8b8b85fcac3fd66f8c00ab18bc1765612

Change-Id: Ia0edb7dc0dbbb879c079ff7000f1b722d86ce7dc

If an error occurs during a bridge impart it's possible that
the "bridge_after" callback might try to run before
control_swap_channel_in_bridge has been signalled to continue.
Since control_swap_channel_in_bridge is holding the control lock
and the callback needs it, a deadlock will occur.

* control_swap_channel_in_bridge now only holds the control
  lock while it's actually modifying the control structure and
  releases it while the bridge impart is running.
* bridge_after_cb is now tolerant of impart failures.

Change-Id: Ifd239aa93955b3eb475521f61e284fcb0da2c3b3

If directmedia=yes is configured, when call is answered, Asterisk sends reINVITE
to both parties to set up media path directly between the endpoints.
In this reINVITE msg SDP origin line (o=) contains IP address of endpoint
instead of IP of asterisk. This behavior violates RFC3264, sec 8:
"When issuing an offer that modifies the session,
the "o=" line of the new SDP MUST be identical to that in the
previous SDP, except that the version in the origin field MUST
increment by one from the previous SDP."
This patch assures IP address of Asterisk is always sent in
SDP origin line.

ASTERISK-17540
Reported by:  saghul

Change-Id: I533a047490c43dcff32eeca8378b2ba02345b64e