* Groups of AGI commands that have similar functionality now reference
each other, and all reference the AGI application for ease of wiki
reference.
* The documentation for the AGI application has been improved, in
particular noting the various AGI types and how they are invoked.
* A warning message has been added to DeadAGI, noting that it is
deprecated.
Change-Id: I479ccdee8a7393f01b18692c3d4ab7e6bdd1875d
When compact_headers was set, we were sending a zero-length header name
for PAI and RPID because we always forced the short header name length
to 0. We did this because we cloned the header from "From" and wanted
to clear "f" from the sname. By cloning however, we bypass pjproject's
automatic logic that sets sname to name if there's no compact form of
the header, which there isn't for PAI and RPID. So now we force sname
to be the same as name right after we set name.
res_pjsip_diversion needed the same treatment for the Diversion header.
ASTERISK-26241 #close
Change-Id: I633ec139630cd83809aae00336cee4a10077e467
If debug was specified in the global configuration but left blank,
the logger would treat it as a wildcard and log all hosts. If
default_from_user was empty, a crash would result.
The global apply handler now checks for empty strings.
ASTERISK-26239 #close
ASTERISK-26238 #close
Change-Id: Ie75727f5cd5808845d92cc81f5713842fb203336
* Eliminated RAII_VAR() usage in
ast_sip_persistent_endpoint_update_state().
* Added a missing allocation failure check to
persistent_endpoint_find_or_create().
* Made persistent_endpoint_find_or_create() create the new object without
a lock as it isn't needed.
* Cleaned up some ao2 container allocation idioms.
* Reordered res_pjsip_mwi.c load_module() and unload_module()
Change-Id: If8ce88fbd82a0c72a37a2388f74f77237a6a36a8
* Eliminated most RAII_VAR() usage.
* Added several missing allocation failure checks.
* Made ast_sip_for_each_contact() allocate the wrapper ao2 object without
a lock as it is not needed.
Change-Id: Ie20913365156c95dd79e5d471cfd25e99ae880bc
The named aor lock was always being locked for writes so a rwlock adds no
benefit and may be slower because rwlocks are biased toward read locking.
Change-Id: I8c5c2c780eb30ce5441832257beeb3506fd12b28
A patch made to the master branch (Now the 14 branch) inadvertently made
libsrtp a required dependency in order to compile Asterisk. Rather than
create dummy defines to substitute for the defines supplied by libsrtp
when libsrtp is not available, most of the code in sdp_srtp.c is moved
into res_srtp.c. This gets more code out of Asterisk's core that isn't
used when SRTP is not available. This also makes another inadvertent
required dependency on libsrtp by Asterisk's core unlikely.
ASTERISK-26253 #close
Reported by: Ben Merrills
Change-Id: I0a46cde81501c0405399c2588633ae32706d1ee7
If both channels which should be masqueraded
are in the same serializer:
1st channel will be locked waiting condition 'complete'
2nd channel will be locked waiting condition 'suspended'
On heavy load system a chance that both channels will be in
the same serializer 'pjsip/distibutor' is very high.
To reproduce compile res_pjsip/pjsip_distributor.c with
DISTRIBUTOR_POOL_SIZE=1
Steps to reproduce:
1. Party A calls Party B (bridged call 'AB')
2. Party B places Party A on hold
3. Party B calls Voicemail app (non-bridged call 'BV')
4. Party B attended transfers Party A to voicemail using REFER.
5. When asterisk masquerades calls 'AB' and 'BV',
a deadlock is happened.
This patch adds a suspension indicator to the taskprocessor.
When a session suspends/unsuspends the serializer
it sets the indicator to the appropriate state.
The session checks the suspension indicator before
suspend the serializer.
ASTERISK-26145 #close
Change-Id: Iaaebee60013a58c942ba47b1b4930a63e686663b
libunbound at version 1.4.20 (which CentOS still uses) declared all
of their string function parameters as as 'char *'. 1.4.21 changed
them all to 'const char *'. Thankfully 1.4.21 also introduced the
UNBOUND_VERSION_MAJOR define so configure now checks for that and
sets HAVE_UNBOUND_CONST_PARAMS. res_resolver_unbound then checks
that and casts away the 'const' if it's not set.
Tested compile and testsuite on CentOS6 (1.4.20), Ubuntu14 (1.4.22) and
Fedora24 (1.5.4). There are a few failing tests to be addressed though.
ASTERISK-26283 #close
Change-Id: Ib708b19b706c5d0ba7b7d5473e6df339d9ae4148
When is executed CLI command "odbc show all" every time is show
information about variable last_negative_connect. If not there a fail
attempt of connection will show date like "1969-12-31 21:00:00".
This patch fix there situation for to show only this information when
exists a fail attempt before.
Change-Id: I7c058b0be6f7642e922de75ee6b82c7276c9f113
When an RTCP packet is sent or received, res_rtp_asterisk generates a
Stasis event that contains the RTCP report as well as the local and
remote addresses that the report pertains to.
The addresses are determined using ast_find_ourip(). For the local
address, this will typically result in a lookup of the hostname of the
server, and then a DNS lookup of that hostname. If you do not have the
host in /etc/hosts, then this results in a full DNS lookup, which can
potentially block for some time.
This is especially problematic when performing RTCP reads, since those
are done on the same thread responsible for reading and writing media.
This patch addresses the issue by performing a lookup of the local
address when RTCP is allocated. We then use this cached local address
for the Stasis events when necessary.
ASTERISK-26280 #close
Reported by Mark Michelson
Change-Id: I3dd61882c2e57036f09f0c390cf38f7c87e9b556
The PJSIP taskprocessors could be overflowed on startup
if there are many (thousands) realtime endpoints
configured with unsolicited mwi.
The PJSIP stack could be totally unresponsive for a few minutes
after boot completed.
This patch creates a separate PJSIP serializers pool for mwi
and makes unsolicited mwi use serializers from this pool.
This patch also adds 2 new global options to tune taskprocessor
alert levels: 'mwi_tps_queue_high' and 'mwi_tps_queue_low'.
This patch also adds new global option 'mwi_disable_initial_unsolicited'
to disable sending unsolicited mwi to all endpoints on startup.
If disabled then unsolicited mwi will start processing
on next endpoint's contact update.
ASTERISK-26230 #close
Change-Id: I4c8ecb82c249eb887930980a800c9f87f28f861a
This change replaces the custom unload process for the outbound
publish module with the common serializer shutdown group.
ASTERISK-25217 #close
Change-Id: I280a0384d860c486202d87d2d674394cca77ffb6
This changes the use of an empty regex for both res_sorcery_config
and res_sorcery_memory to "." instead. This is a more compatible
regular expression which also works on FreeBSD.
ASTERISK-26206 #close
Change-Id: Ia9166dd176f1597555ba22b6931180d0626c1388
Asterisk already supported iLBC 30. This change adds iLBC 20. Now, Asterisk
defaults to iLBC 20 but falls back to iLBC 30, when the remote party requests
this.
ASTERISK-26218 #close
ASTERISK-26221 #close
Reported by: Aaron Meriwether
Change-Id: I07f523a3aa1338bb5217a1bf69c1eeb92adedffa
The fax_detect_framehook() has the potential to deadlock if an incoming
fax happens during the Playback or similar application.
* Fixed the potential deadlock by not calling ast_async_goto() with the
channel lock held.
* Made always eat the fax detection frame whether there is a fax extension
or not.
* Made only detach the framehook if we detected a fax and not on other
possible frames.
ASTERISK-26216
Reported by: Richard Mudgett
Change-Id: I99da35c26d1cd802626ffb4c1b4eb5b015581b6d
The fax detection timeout option did not work because basically the wrong
variable was checked in fax_detect_framehook(). As a result, the timer
would timeout immediately and disable fax detection.
* Fixed ignoring negative timeout values. We'd complain and then go right
on using the negative value.
* Fixed destroy_faxdetect() in the off-nominal case of an incomplete
object creation.
* Added more range checking to FAXOPT(gateway) timeout parameter.
ASTERISK-26214 #close
Reported by: Richard Mudgett
Change-Id: Idc5e698dfe33572de9840bc68cd9fc043cbad976
The new endpoint option allows the PJSIP channel driver's fax_detect
endpoint option to timeout on a call after the specified number of
seconds into a call. The new feature is disabled if the timeout is set
to zero. The option is disabled by default.
ASTERISK-26214
Reported by: Richard Mudgett
Change-Id: Id5a87375fb2c4f9dc1d4b44c78ec8735ba65453d
This patch removed call of pjsip_tx_data_dec_ref in send_notify
if send_request failed.
The pjsip_dlg_send_request deletes the message on error by itself.
It seems this patch fixes next issues:
ASTERISK-26199
ASTERISK-26166
ASTERISK-26174
Change-Id: I8b05917c93d993f95d604c042ace5f1a5500f59a
This changes context switches from a linked list to a vector, makes
'struct ast_sw' opaque to pbx.c.
Although ast_walk_context_switches is maintained the procedure is no
longer efficient except for the first call (inc==NULL). This
functionality is replaced by two new functions implemented by vector
macros.
* ast_context_switches_count (AST_VECTOR_SIZE)
* ast_context_switches_get (AST_VECTOR_GET)
As with ast_walk_context_switches callers of these functions are
expected to have locked contexts. Only a few places in Asterisk walked
the switches, they have been converted to use the new functions.
Change-Id: I08deb016df22eee8288eb03de62593e45a1f0998
If AST_TEST_DEFINE is not conditional to TEST_FRAMEWORK it produces dead
code. This places all existing unit tests into a conditional block if
they weren't already.
ASTERISK-26211 #close
Change-Id: I8ef83ee11cbc991b07b7a37ecb41433e8c734686
The function create_mwi_subscriptions_for_endpoint checks
if there is active contacts by retrieving aors and contacts.
This function is used to create all unsolicited mwi subscriptions
on startup and is used when contact added.
In both cases it's not necessary to check if there are contacts.
The contacts are needed when asterisk sends mwi.
ASTERISK-26200 #close
Change-Id: I98e43bdc97f3c0829951cd9bf5f3c6348c6ac1fa
With this change, the initial RTP sequence number is randomly chosen not between
0 and 65535 (0xffff) but 0 and 32767 (0x7fff). This assures, the roll-over
counter (ROC) synchronization is not lost for sRTP, when the very first RTP
packets get lost; see http://srtp.sourceforge.net/faq.html#Q6
ASTERISK-26207 #close
Change-Id: I9a527e3aa3ce8f3becc5131d7ba32b57b5845464
This changes context ignore patterns from a linked list to a vector,
makes 'struct ast_ignorepat' opaque to pbx.c.
Although ast_walk_context_ignorepats is maintained the procedure is no
longer efficient except for the first call (inc==NULL). This
functionality is replaced by two new functions implemented by vector
macros.
* ast_context_ignorepats_count (AST_VECTOR_SIZE)
* ast_context_ignorepats_get (AST_VECTOR_GET)
As with ast_walk_context_ignorepats callers of these functions are
expected to have locked contexts. Only a few places in Asterisk walked
the ignorepats, they have been converted to use the new functions.
Change-Id: I78f2157d275ef1b7d624b4ff7d770d38e5d7f20a
This changes context includes from a linked list to a vector, makes
'struct ast_include' opaque to pbx.c.
Although ast_walk_context_includes is maintained the procedure is no
longer efficient except for the first call (inc==NULL). This
functionality is replaced by two new functions implemented by vector
macros.
* ast_context_includes_count (AST_VECTOR_SIZE)
* ast_context_includes_get (AST_VECTOR_GET)
As with ast_walk_context_includes callers of these functions are
expected to have locked contexts. Only a few places in Asterisk walked
the includes, they have been converted to use the new functions.
const have been applied where possible to parameters for ast_include
functions.
Change-Id: Ib5c882e27cf96fb2aec67a39c18b4c71c9c83b60
This commit adds scaffolding in order to support the SILK audio format
on calls. Roughly, this is what is added:
* Cached silk formats. One for each possible sample rate.
* ast_codec structures for each possible sample rate.
* RTP payload mappings for "SILK".
In addition, this change overhauls the res_format_attr_silk file in the
following ways:
* The "samplerate" attribute is scrapped. That's native to the format.
* There are far more checks to ensure that attributes have been
allocated before attempting to reference them.
* We do not SDP fmtp lines for attributes set to 0.
These changes make way to be able to install a codec_silk module and
have it actually work. It also should allow for passthrough silk calls
in Asterisk.
Change-Id: Ieeb39c95a9fecc9246bcfd3c45a6c9b51c59380e
aor_observer_deleted() needs to operate on all contacts found for the
deleted AOR instead of only the first one found. This is really only a
problem if there is more than one contact for the AOR.
Change-Id: Id24ac0d5e8c931330231fb45dd2a331a84339dc1
* Fix some whitespace in various routines.
* Rename i to iter in persistent_endpoint_update_state().
* Fix off-nominal copy/paste message wording in
persistent_endpoint_contact_deleted_observer()
Change-Id: Id8e34f5d09e7eebac3af22501c44c1110a3e29d8
Since July 2014, TLS based protocols (SIP over TLS, Secure WebSockets, HTTPS)
support PFS thanks to ASTERISK-23905. In July 2015, the same feature was added
for DTLS. The source code from main/tcptls.c should have been re-used to ease
security audits. Therefore, this change rolls back the change from July 2015 and
re-uses the code from July 2014. This has the additional benefits to work under
CentOS 7 and enabling not just ECDHE but DHE based cipher suites as well.
ASTERISK-25659 #close
Reported by: StefanEng86, urbaniak, pay123
Tested by: sarumjanuch, traud
patches:
res_rtp_asterisk.patch submitted by sarumjanuch
dtls_centos_step_1.patch submitted by traud
dtls_centos_step_2.patch submitted by traud
Change-Id: I537cadf4421f092a613146b230f2c0ee1be28d5c
It is possible in a hypothetical situation for a session refresh to be
invoked on a PJSIP when the negotiatior on the INVITE session has not
yet been established. While this shouldn't occur with existing uses of
ast_sip_session_refresh, the crashes that occur due to improperly
calling PJSIP functions that expect a non-NULL negotiatior are
avoidable. PJSIP will create the negotiator in pjsip_inv_reinvite; this
means that simply checking for the presence of the negotiator before
passing it to other PJSIP functions that use it is allowable. As such,
this patch adds checks for the presence of the negotiator before calling
PJSIP functions that assume it is non-NULL.
Change-Id: I1028323e7e01b0a531865e5412a71b6f6ec4276d
When something very sad and wrong occurs, it's challenging sometimes to
figure out why. This patch adds some additional debug statements on
off-nominal paths to try and make debugging easier.
Change-Id: I7bffb73cc733b6f80193a23340881db4a102b640
When res_corosync detects that a node leaves or joins, it currently is
informed of this via Corosync callbacks. However, there are a few
limitations with the information presented:
(1) While we have information that Corosync is aware of - such as the
Corosync nodeid - that information is really only useful inside of
Corosync or res_corosync. There's no way to translate a Corosync
nodeid to some other internally useful unique identifier for the
Asterisk instance that just joined or left the cluster.
(2) While res_corosync is notified of the instance joining or leaving
the cluster, it has no mechanism to inform the Asterisk core or
other modules of this event. This limits the usefulness of res_corosync
as a heartbeat mechanism for other modules.
This patch addresses both issues.
First, it adds the notion of a cluster discovery message both within the
Stasis message bus, as well as the binary event messages that
res_corosync uses to transmit data back and forth within the cluster.
When Asterisk joins the cluster, it sends a discovery message to the other
nodes in the cluster, which correlates the Corosync nodeid along with
the Asterisk EID. res_corosync now maintains a hash of Corosync nodeids
to Asterisk EIDs, such that it can map changes in cluster state with the
Asterisk instance that has that nodeid. Likewise, when an Asterisk
instance receives a discovery message from a node in the cluster, it now
sends its own discovery message back to the originating node with the
local Asterisk EID. This lets Asterisk instances within the cluster
build a complete picture of the other Asterisk instances within the
cluster.
Second, it publishes the discovery messages onto the Stasis message bus.
Said messages are published whenever a node joins or leaves the cluster.
Interested modules can subscribe for the ast_cluster_discovery_type()
message under the ast_system_topic() and be notified when changes in
cluster state occur.
Change-Id: I9015f418d6ae7f47e4994e04e18948df4d49b465
* We weren't properly subscribing to the channel and it's originator
on create.
* We weren't doing a publish_dial after calling ast_call on dial.
* We weren't calling depart_bridge when a channel left the dial bridge.
The first 2 issues were causing events to not be generated and the third
was actually causing channels to not get properly destroyed when hung up.
Together these 3 issues were causing the new
rest_apichannels/create_dial_bridge tests to fail.
As a result of the fixes, the cdr state machine had to be slightly
tweaked to allow bridge leave events without asserting and the tests
themselves had to be updated to account for the channels now cleaning
themselves up.
Change-Id: Ibf23abf5a62de76e82afb4461af5099c961b97d8
The ASTERISK-25904 change-id I8fad8aae9305481469c38d2146e1ba3a56d3108f
patch introduced several regressions when the newly created "Updated"
state goes out for each endpoint registration refresh.
1) It restarted any OPTIONS RTT ping cycle.
2) It would interfere with a currently active ping and throw off that
ping's resulting RTT calculation.
3) It cleared the RTT time each time the endpoint was refreshed.
4) The cleared RTT time was sent out as a statsd update each time.
5) It created two AMI events for each update.
* Revert the original patch and reimplement it. Now the current contact
status state is re-sent instead of the state being momentarily toggled
every time the endpoint refreshes its registration. The statsd events are
not created for the re-sent refresh because they are sent after every
OPTIONS ping.
ASTERISK-26160 #close
Reported by: Matt Jordan
Change-Id: Ie072be790fbb2a8f5c1c874266e4143fa31f66d1
The func_odbc module was modified to ensure that the
previous behavior of using a single database connection
was maintained. This was done by getting a single database
connection and holding on to it. With the new multiple
connection support in res_odbc this will actually starve
every other thread from getting access to the database as
it also maintains the previous behavior of having only
a single database connection.
This change disables the func_odbc specific behavior if
the res_odbc module is running with only a single database
connection active. The connection is only kept for the
duration of the request.
ASTERISK-26177 #close
Change-Id: I9bdbd8a300fb3233877735ad3fd07bce38115b7f
If the SQL UPDATE statement changes nothing then SQLRowCount returns 0.
This value should be treated as success.
But the function sorcery_realtime_update treats it as failed.
This bug was found using stress tests on PJSIP.
If there are 2 consecutive SIP REGISTER requests with the same contact data
during 1 second then res_pjsip_registrar adds contact location on 1st request
and tries to update contact location on 2nd.
The update fails and res_pjsip_registrar even removes correct contact location.
The test "object_update_uncreated" was removed from test_sorcery_realtime.c
because it's now a valid situation.
This patch also adds missing debug of extra SQL parameter.
ASTERISK-26172 #close
Change-Id: I05a7f3051455336c9dda29efc229decf86071303
Some T.38 implementations may send another re-invite after the initial
one which adds additional negotiation details (such as the max bitrate).
Currently this will fail when passthrough is being done in chan_sip as we
do nothing if T.38 is already active.
Other handlers of T.38 inside of Asterisk (such as res_fax) handle this
scenario so this change adds support for it to chan_sip and res_pjsip_t38.
If a request to negotiate is received while T.38 is already enabled a
new re-INVITE is sent and negotiation is done again.
ASTERISK-26179 #close
Change-Id: I0298494d3da6df3219bbfa4be9aa04015043145c
When using TCP transport with chan_pjsip, the TCP_NODELAY
option value was allocated on the stack, then passed as a
pointer to the tcp transport configuration structure, and
later re-used on subsequently created sockets when it was
no longer valid. This patch changes the allocation to be
a static.
ASTERISK-26180 #close
Reported by: Scott Griepentrog
Change-Id: I3251164c7f710dbdab031282f00e30a9770626a0
If specified, incoming SUBSCRIBE requests will be searched for the matching
extension in the indicated context. If no "subscribe_context" is specified,
then the "context" setting is used.
ASTERISK-25471 #close
Change-Id: I3fb7a15f5bc154079bd348c08b7ad1cdd2d5e514
When an answer SDP is invalid we were disconnecting the outgoing call and
sending two BYE requests. The first BYE was sent by PJPROJECT because of
the invalid SDP answer. The second BYE was sent by Asterisk because it
thought the canceled call was the result of the RFC5407 section 3.1.2 race
condition.
* Made not send the BYE on a canceled session if the SDP negotiation is
incomplete because PJPROJECT has already sent a BYE for the failed
negotiation.
ASTERISK-25772 #close
Reported by: Dmitriy Serov
Change-Id: I44ad0bd0605e8eeb7035c890d6f97a1331f1a836
When an incoming call defers SDP negotiation and then sends us an invalid
SDP in the ACK, we need to send a BYE to disconnect the call. In this
case SDP negotiation has failed and we don't have valid media streams
negotiated.
ASTERISK-25772
Change-Id: Ia358516b0fc1e6c4c139b78246f10b9da7a2dfb8
Registering the PJMEDIA error codes allows errors found when parsing an
incoming SDP to be easier to figure out.
"Missing SDP rtpmap for dynamic payload type (PJMEDIA_SDP_EMISSINGRTPMAP)"
is much easier to understand than "Unknown error 220030".
ASTERISK-25772
Change-Id: I44b2dcea656fedd7593171be9e845880a2c70ca0
pjsip_inv_end_session() is documented as being able to return the
passed in tdata parameter set to NULL on success.
Change-Id: I09d53725c49b7183c41bfa1be3ff225f3a8d3047
This change removes hardcoded SDP parsing and generation for
Siren7 and Siren14 from chan_sip and moves it to format attribute
modules so it can also be used by chan_pjsip.
With this the fmtp lines for both are added with the bitrate
information.
ASTERISK-26021
Change-Id: Ibb004eda37a14c0a35ef0613f6237977fc800037
fax_v21_session_new created a session details object but only released
the allocation reference during error conditions. fax_session_new adds
it's own reference to details if needed so the caller is always
responsible for cleaning it's own reference.
ASTERISK-26141 #close
Change-Id: Ie7fc52a83b6596ce9ce2d5a2bd9f3e204f48fc88
The patch removes updating all Endpoints' status on startup.
Instead, only non-qualified aors with static contact
and non-qualified non-expired contacts are retrieved from the realtime to
update the endpoint status to ONLINE.
The endpoint name was added to the contact object to simply find the endpoint
that created this contact.
The status of endpoints with qualified aors will be updated by 'qualify'
functions.
ASTERISK-26061 #close
Change-Id: Id324c1776fa55d3741e0c5457ecac0304cb1a0df
gcc 6 caught a previously unidentified self-comparison in
ice_candidate_cmp. Fixed it and re-ordered the predicates for better
short-circuiting.
ASTERISK-26140 #close
Change-Id: I3da713c568e24064430257b3502fbdafd35af7a7
A non-existent constraint was being referenced in the upgrade script.
This patch corrects the problem by removing the reference.
In addition, the head of the alembic branch referred to a non-existent
revision. This has been fixed by referring to the proper revision.
This patch fixes another realtime problem as well. Our Alembic scripts
store booleans as yes or no values. However, Sorcery tries to insert
"true" or "false" instead. This patch introduces a new boolean type that
translates to "yes" or "no" instead.
ASTERISK-26128 #close
Change-Id: I51574736a881189de695a824883a18d66a52dcef
Occasionally under load we'll attempt to send a final NOTIFY on a
subscription that's already been terminated and a SEGV will occur
down in pjproject's evsub_destroy function. This is a result of a
race condition between all the paths that can generate a notify
and/or destroy the underlying pjproject evsub object:
* The client can send a SUBSCRIBE with Expires: 0.
* The client can send a SUBSCRIBE/refresh.
* The subscription timer can expire.
* An extension state can change.
* An MWI event can be generated.
* The pjproject transaction timer (timer_b) can expire.
Normally when our pubsub_on_evsub_state is called with a terminate,
we push a task to the serializer and return at which point the dialog
is unlocked. This is usually not a problem because the task runs
immediately and locks the dialog again. When the system is heavily
loaded though, there may be a delay between the unlock and relock
during which another event may occur such as the subscription timer
or timer_b expiring, an extension state change, etc. These may also
cause a terminate to be processed and if so, we could cause pjproject
to try to destroy the evsub structure twice. There's no way for us to
tell that the evsub was already destroyed and the evsub's group lock
can't tolerate this and SEGVs.
The remedy is twofold.
* A patch has been submitted to Teluu and added to the bundled
pjproject which adds add/decrement operations on evsub's group lock.
* In res_pjsip_pubsub:
* configure.ac and pjproject-bundled's configure.m4 were updated
to check for the new evsub group lock APIs.
* We now add a reference to the evsub group lock when we create
the subscription and remove the reference when we clean up the
subscription. This prevents evsub from being destroyed before
we're done with it.
* A state has been added to the subscription tree structure so
termination progress can be tracked through the asyncronous tasks.
* The pubsub_on_evsub_state callback has been split so it's not doing
double duty. It now only handles the final cleanup of the
subscription tree. pubsub_on_rx_refresh now handles both client
refreshes and client terminates. It was always being called for
both anyway.
* The serialized_on_server_timeout task was removed since
serialized_pubsub_on_rx_refresh was almost identical.
* Missing state checks and ao2_cleanups were added.
* Some debug levels were adjusted to make seeing only off-nominal
things at level 1 and nominal or progress things at level 2+.
ASTERISK-26099 #close
Reported-by: Ross Beer.
Change-Id: I779d11802cf672a51392e62a74a1216596075ba1
Do not use DTLSv1_method() but DTLS_method() when available in OpenSSL of the
underlying platform. This change enables DTLS 1.2 since OpenSSL 1.0.2, for
WebRTC (DTLS-SRTP via SIP-over-WebSockets). This change enables AEAD-based
cipher-suites.
ASTERISK-26130 #close
Change-Id: I41f24448d6d2953e8bdb97c9f4a6bc8a8f055fd0
The receipt of a SIP MESSAGE may occur over any transport including TCP
and TLS. When the message is received, the original URI is added to the
message in the field PJSIP_RECVADDR, but this is insufficient to ensure
a reply message can reach the originating endpoint. This patch adds the
PJSIP_TRANSPORT field populated with the transport type.
ASTERISK-26132 #close
Change-Id: I28c4b1e40d573a056c81deb213ecf53e968f725e
When shutting down res_pjsip_session will get unloaded before res_pjsip.
The act of unloading unregisters all the PJSIP services and sets
their module IDs to -1. In some cases it is possible for a timer to
occur after this happens which calls into res_pjsip_session. The
res_pjsip_session module can then try to get the session from the
INVITE session using the module ID. Since the module ID is now -1
this fails.
This change stores a copy of the module ID and uses it for the timer
callback scenario. If the module ID is -1 the callback immediately
returns but if the module ID is valid then it continues as normal.
This works as the original ID of the module is guaranteed to still
be valid when used with the INVITE session.
ASTERISK-26127 #close
Change-Id: I88df72525c4e9ef9f19c13aedddd3ac4a335c573
Announcer channels were not being destroyed because the
stasis_app_control structure that referenced them was not being
destroyed. The control structure was not being destroyed because it was
not being unlinked from its container. It was not being unlinked from
its container because the after bridge callback for the announcer
channel was not being run. The after bridge callback was not being run
because the after bridge datastore was not being removed from the
channel on destruction. The channel was not being destroyed because the
hangup that used to destroy the channel was now only reducing the
reference count to one. The reference count of the channel was only
being reduced to one because the stasis_app_control structure was
holding the final reference...
The control structure used to not keep a reference to the channel, so
that loop described above did not happen.
The solution is to manually remove the control structure from its
container when the playback on a bridge is complete.
ASTERISK-26083 #close
Reported by Joshua Colp
Change-Id: I0ddc0f64484ea0016245800b409b567dfe85cfb4
* In unload_module(), reordered destroying things to minimize the window
that the global transports container could be used by other threads on
shutdown. When shutting down you need to stop things in the opposite
order of creation.
* Put the global transports container into an AO2_GLOBAL_OBJ_STATIC to
eliminate the crash potential by other threads using the container on
shutdown.
* Made struct monitored_transport.sip_received not use
ast_atomic_fetchadd_int() since it is used as a boolean value that is only
set TRUE. It was previously incremented for every received SIP message
and could theoretically overflow.
* In monitored_transport_state_callback(), allocated the monitored
transport object without a lock since the lock was unused.
* In keepalive_global_loaded(), removed releasing the transports container
if the keepalive_thread could not be started. I set it up to be tried
again if the user reloads the configuration.
Change-Id: I8d12d16ef564290fa6d25a32334bb5ce8fdf87ff
ARI was recently outfitted with operations to create and dial channels.
This leads to the ability to try funny stuff. You could create a channel
and then immediately try to play back media on it. You could create a
channel, dial it, and while it is ringing attempt to make it continue in
the dialplan.
This commit attempts to fix this by adding a channel state check to
operations that should not be able to operate on outbound channels that
have not yet answered. If a channel is in an invalid state, we will send
a 412 response.
ASTERISK-26047 #close
Reported by Mark Michelson
Change-Id: I2ca51bf9ef2b44a1dc5a73f2d2de35c62c37dfd8
This patch fixes a race condition processing received REGISTER requests
and their retransmissions caused by REGISTER requests being processed by
two threads. The "sip_transaction Unable to register REGISTER transaction
(key exists)" message is a notable symptom of this issue.
This issue was more likely to happen before the pjsip/distributor
serializers were created. Instead of steps one and two below placing the
REGISTER messages into the same pjsip/distributor they were placed in
random pjsip/default serializers.
1) REGISTER requests come in and get placed on the pjsip/distributor
serializer.
2) Before the first request is processed a retransmission comes in and is
placed on the same pjsip/distributor serializer.
3) The first request goes up the pjsip stack and is then shunted off to
the pjsip/aor/<aor> serializer.
4) Before the first request is completed processing in the pjsip/aor/<aor>
serializer, the second request goes up the pjsip stack and is also shunted
off to the pjsip/aor/<aor> serializer.
5) The first request completes processing and sends out its response.
6) The second request completes processing and tries to send out its
response but pjlib complains that the REGISTER transaction key already
exists.
7) Sadness ensues.
* The race is eliminated by removing the pjsip/aor/<aor> serializer and
continuing the processing in the pjsip/distributor serializer. Now any
retransmissions queued in the pjsip/distributor serializer will be
processed after the first message is completely processed.
ASTERISK-26088 #close
Reported by: Richard Mudgett
Change-Id: I842d714346088bf717ea27437f1dd85bff0bab5a
Sorcery creates taskprocessors for object types to process object observer
callbacks. An API call is needed to be able to set the congestion levels
of these taskprocessors for selected object types.
* Updated PJSIP's contact and contact_status sorcery object type observer
default congestion levels based upon stress testing. Increased the
congestion levels to reduce the potential for bursty register/unregister
and subscribe/unsubscribe activity from triggering the taskprocessor
overload alert.
ASTERISK-26088
Reported by: Richard Mudgett
Change-Id: I4542e83b556f0714009bfeff89505c801f1218c6
When taskprocessors get backed up, there is a good chance that we are
being overloaded and need to defer adding new work to the system.
* Implemented a high/low water alert mechanism for modules to check if the
system is being overloaded and take appropriate action. When a
taskprocessor is created it has default congestion levels set. A
taskprocessor can later have those congestion levels altered for specific
needs if stress testing shows that the taskprocessor is a symptom of
overloading or needs to handle bursty activity without triggering an
overload alert.
* Add CLI "core show taskprocessor" low/high water columns.
* Fixed __allocate_taskprocessor() to not use RAII_VAR(). RAII_VAR() was
never a good thing to use when creating a taskprocessor because of the
nature of how its references needed to be cleaned up on a partial
creation.
* Made res_pjsip's distributor check if the taskprocessor overload alert
is active before placing a message representing brand new work onto a
distributor serializer.
ASTERISK-26088
Reported by: Richard Mudgett
Change-Id: I182f1be603529cd665958661c4c05ff9901825fa
We must continue using the serializer that the original INVITE came in on
for the dialog. There may be retransmissions already enqueued in the
original serializer that can result in reentrancy and message sequencing
problems.
Outgoing call legs create the pjsip/outsess/<endpoint> serializers for
their dialogs.
ASTERISK-26088
Reported by: Richard Mudgett
Change-Id: I24d7948749c582b8045d5389ba3f6588508adbbc
* Resolves potential reentrancy problems if system restarted in the middle
of subscription message transactions.
* Fixes memory leak recreating persistent subscriptions when the
subscription resource tree could not be created.
ASTERISK-26088
Reported by: Richard Mudgett
Change-Id: I71e34d7ae8ed35a694f1030e820e2548c48697be
We must continue using the serializer that the original SUBSCRIBE came in
on for the dialog. There may be retransmissions already enqueued in the
original serializer that can result in reentrancy and message sequencing
problems. The "sip_transaction Unable to register SUBSCRIBE transaction
(key exists)" message is a notable symptom of this issue.
Outgoing subscriptions still create the pjsip/pubsub/<endpoint>
serializers for their dialogs.
ASTERISK-26088
Reported by: Richard Mudgett
Change-Id: I18b00bb74a56747b2c8c29543a82440b110bf0b0
Incoming messages that are not part of a dialog or a recognized response
to one of our requests need to be sent to a consistent serializer. Under
load we may be queueing retransmissions before we can process the original
message. We don't need to throw these messages onto random serializers
and cause reentrancy and message sequencing problems.
* Created a pool of pjsip/distributor serializers that get picked by
hashing the call-id and remote tag strings of the received messages.
* Made ast_sip_destroy_distributor() destroy items in the reverse order of
creation.
ASTERISK-26088
Reported by: Richard Mudgett
Change-Id: I2ce769389fc060d9f379977f559026fbcb632407
We should not be processing any incoming messages until we are fully
booted. We may not have dialplan or other needed configuration loaded
yet.
ASTERISK-26089 #close
Reported by: Scott Griepentrog
ASTERISK-26088
Reported by: Richard Mudgett
Change-Id: I584aefb4f34b885a8927e1f13a2c64babd606264
POSIX defines signal.h. sys/signal.h should not be used as it is
c-library internal header which may or may not exist. Notably with
musl it generates warning of being incorrect.
Change-Id: Ia56b0aa1d84b5c590114867b1b384a624f39a6fc
A crash can occur in res_hep_pjsip or res_hep_rtcp if res_hep has not
loaded and does not have a configuration file. Previously when this
occurred, checks were put in to see if the configuration was loaded
successfully. While this is a good idea - and has been added to the
offending function in res_hep - the reality is res_hep_pjsip and
res_hep_rtcp have no business running if res_hep isn't also running.
As such, this patch also adds a function to res_hep that returns whether
or not it successfully loaded. Oddly enough, ast_module_check returns
"everything is peachy" even if a module declined its load - so it cannot
be solely relied on. res_hep_pjsip and res_hep_rtcp now also check this
function to see if they should continue to load; if it fails, they
decline their load as well.
ASTERISK-26096 #close
Change-Id: I007e535fcc2e51c2ca48534f48c5fc2ac38935ea
Testing has shown that our usage of UnixODBC is problematic
due to bugs within UnixODBC itself as well as the heavy weight
cost of connecting and disconnecting database connections, even
when pooling is enabled.
For users of UnixODBC 2.3.1 and earlier crashes would occur due
to insufficient protection of the disconnect operation. This was
fixed in UnixODBC 2.3.2 and above.
For users of UnixODBC 2.3.3 and higher a slow-down would occur
under heavy database use due to repeated connection establishment.
A regression is present where on each connection the database
configuration is cached again, with the cache growing out of
control.
The connection pool implementation present in this change helps
to mitigate these issues by reducing how much we connect and
disconnect database connections. We also solve the issue of
crashes under UnixODBC 2.3.1 by defaulting the maximum number of
connections to 1, returning us to the previous working behavior.
For users who may have a fixed version the maximum concurrent
connection limit can be increased helping with performance.
The connection pool works by keeping a list of active connections.
If the connection limit has not been reached a new connection is
established. If the connection limit has been reached then the
request waits until a connection becomes available before
continuing.
ASTERISK-26074 #close
ASTERISK-26054 #close
Change-Id: I6774bf4bac49a0b30242c76a09c403d2e856ecff
Since libSRTP 1.5, its Random Number Generator (RNG) is not maintained anymore.
Therefore, the symbol RAND_bytes is used instead of crypto_get_random.
ASTERISK-24436 #close
Change-Id: Iea0bae4d4e3c9aa0926ea442b6484b5159789d96
If you create a local channel and don't specify an originator channel
to take capabilities from, we automatically add all audio formats to
the new channel's capabilities. When we try to make the channel
compatible with another, the "best format" functions pick the best
format available, which in this case will be slin192. While this is
great for preserving quality, it's the worst for performance and
overkill for the vast majority of applications.
In the absense of any other information, adding all formats is the
correct thing to do and it's not always possible to supply an
originator so a new parameter 'formats' has been added to the channel
create/originate functions. It's just a comma separated list of formats
to make availalble for the channel. Example: "ulaw,slin,slin16".
'formats' and 'originator' are mutually exclusive.
To facilitate determination of format names, the format name has been
added to "core show codecs".
ASTERISK-26070 #close
Change-Id: I091b23ecd41c1b4128d85028209772ee139f604b
These flags are non-portable GNU extensions. Make their use
optional. This fixes complication error on e.g. musl c-library
based systems.
Change-Id: I0aa06efc62aa8995f091445c8b762a75a91042f3
The pjproject doxygen for rdata->msg_info.info says to call
pjsip_rx_data_get_info() instead of accessing the struct member directly.
You need to call the function mostly because the function will generate
the struct member value if it is not already setup.
Change-Id: I4d519385a577f3e9d9193a88125e493cf17fa799
Re-ordered the body items so Message-Account is second.
Messages-Waiting: no
Message-Account: sip:1571@<IP Removed>:5060
Voice-Message: 0/0 (0/0)
ASTERISK-26065 #close
Reported-by: Ross Beer
Change-Id: If5d35a64656eac98c2dd5e490cc0b2807bed80c3
Added notes about when you can read or write headers. Specifically
about being able to read on the inbound channel and write on an
outbound channel.
ASTERISK-26063 #close
Reported by: Private Name
Tested by: Rusty Newton
Change-Id: Ibeb64af17d1f6451028b3c29855a3f151a01d8c5
This adds a new parameter to the end of a multicast RTP dialing string.
This parameter defines the following options:
* i: Set the interface from which multicast RTP is sent
* l: Set whether multicast packets are looped back to the sender
* t: Set the TTL for multicast packets
* c: Set the codec to use for RTP
ASTERISK-26068 #close
Reported by Mark Michelson
Change-Id: I033b706b533f0aa635c342eb738e0bcefa07e219
ARI dial had been implemented using the Dial API. This made great sense
when dialing was 100% separate from bridging. However, if a channel were
to be added to a bridge during the dial attempt, there would be a
conflict between the dialing thread and the bridging thread. Each would
be attempting to read frames from the dialed channel and act on them.
The initial attempt to make the two play nice was to have the Dial API
suspend the channel in the bridge and stay in charge of the channel
until the dial was complete. The problem with this was that it was
riddled with potential race conditions. It also was not well-suited for
the case where the channel changed which bridge it was in during the
dial.
This new approach removes the use of the Dial API altogether. Instead,
the channel we are dialing is placed into an invisible ARI dialing
bridge. The bridge channel thread handles incoming frames from the
channel. If the channel is added to a real bridge, it is departed from
the invisible bridge and then added to the real bridge. Similarly, if
the channel is removed from the real bridge, it is automatically added
back to the invisible bridge if the dial attempt is still active.
This approach keeps the threading simple by always having the channel
being handled by bridge channel threads.
ASTERISK-25925
Change-Id: I7750359ddf45fcd45eaec749c5b3822de4a8ddbb
As res_pjsip_nat rewrites contact's address, only the last Via header
can contain the source address of registered endpoint.
Also Call-Id header may contain the source address of registered
endpoint.
Added "via_addr", "via_port", "call_id" to contact.
Added new fields ViaAddress, CallID to AMI event ContactStatus.
ASTERISK-26011
Change-Id: I36bcc0bf422b3e0623680152d80486aeafe4c576
There are a lot of verbose messages about Endpoint and Contact status
changes if there are many dynamic endpoints.
The patch sets verbose level 2 for Endpoint status changes
and verbose level 3 for Contact status changes.
ASTERISK-26055 #close
Change-Id: Ie64e261ddbbc41bfff0f0190241152cc123fe6d7
The pjproject doxygen for rdata->msg_info.info says to call
pjsip_rx_data_get_info() instead of accessing the struct member directly.
You need to call the function mostly because the function will generate
the struct member value if it is not already setup.
Change-Id: Iafe8b01242b7deb0ebfdc36685e21374a43936d2
Richard Mudgett
Matt Jordan
zuul
Joshua Colp
George Joseph
Alexei Gradinari
Rodrigo Ramírez Norambuena
Mark Michelson
Kevin Harwell
Alexander Traud
Corey Farrell
Scott Griepentrog
Torrey Searle
Timo Teräs
Rusty Newton