34469 Commits (20)

 

All Branches   

Search

Author	SHA1	Message	Date
Gian Diego Javes	cc9b77d737	res_websocket_client: Release lock_obj when retries are exhausted ... ast_websocket_client_connect() takes lock_obj at the top of every pass through the retry loop. When the attempts run out it leaves through the break and returns NULL without releasing it, so the calling thread ends up holding it. The other two exits, the ast_calloc failure and the success path, do release it. The retry path also takes the lock again instead of releasing it before sleeping. The header describes lock_obj as "an ao2 object to lock while the connection is being attempted", and the usleep is the wait between attempts rather than an attempt, so that is where it should be released. chan_websocket.c passes the per-call instance as lock_obj, allocated with ao2_alloc and no flags, so it gets the default AO2_ALLOC_OPT_LOCK_MUTEX. When the object is freed, __ao2_ref() reaches ast_mutex_destroy() with the mutex still held and frees the memory anyway. There is no deadlock, Asterisk mutexes are recursive, but every call that fails to connect logs two ERROR lines. With each pass balanced, the lock is held exactly once when the loop breaks, so a single unlock before the return is enough. Fixes: #2121	1 day ago
Jacky W J Li	51b6dc7d7b	frame: Correct ast_frame_adjust_volume* documentation of 'adjustment' ... The documentation for ast_frame_adjust_volume() and ast_frame_adjust_volume_float() described the 'adjustment' parameter as a dB value. In reality the parameter is a linear gain factor applied directly to the audio samples: a positive value multiplies each sample, a negative value divides each sample, and 0 leaves the audio unchanged. The documentation block has also been reformatted to use the standard ' *' Doxygen style (asterisk in the second column) for consistency with the rest of the file. To convert a dB change to this factor use adjustment = 10^(dB/20) (for example, 2 ~= +6 dB and 10 ~= +20 dB). Resolves: #1167	2 days ago
Jaco Kroon	f3d5b76978	res_config_odbc: Missing SQLFreeHandle in error paths. ... Signed-off-by: Jaco Kroon <jaco@uls.co.za>	2 days ago
George Joseph	e8ce0c955f	res_cdrel_custom: Use extendable ast_str for JSON records. ... Since JSON records include the field names, the actual record size can be quite lengthy and since the thread-local backed ast_str used by the DSV writer can't be extended, we need to use a regular, extendable ast_str. Resolves: #2106	2 days ago
Sean Bright	1a47880d7f	res_geolocation: Don't crash on invalid sorcery configuration ... A crash will occur when you've configured object mappings using the realtime wizard but neglect to set up your realtime configuration in `extconfig.conf`. For example, adding the following to `sorcery.conf`: [res_geolocation] location = realtime,geo_location profile = realtime,geo_profile But not defining `geo_location` or `geo_profile` in `extconfig.conf` will cause Asterisk to crash on startup.	3 days ago
Sean Bright	2d42af12da	res_geolocation: Load when explicitly configured in sorcery.conf ... If you've explicitly configured your configuration objects in sorcery.conf, it is normal for `ast_sorcery_apply_default` to return something other than `AST_SORCERY_APPLY_SUCCESS` (e.g. `AST_SORCERY_APPLY_DEFAULT_UNNECESSARY`). This is one of the few places (outside of unit tests) that we actual pay attention to the return value of `ast_sorcery_apply_default` at all.	3 days ago
fsliwenjie	5bbca40381	res_stasis_recording: Allow recording a channel that is dialing via ARI ... When an ARI application dials a channel with channels.dial, the channel is placed into an internal, invisible holding bridge (the dial bridge) while dialing. record_file() rejected recording for any channel that was in a bridge, so channels.record failed with "Cannot record channel while in bridge" even though the application never joined the channel to a bridge of its own. Narrow the check so it only rejects recording when the channel is in a bridge that is NOT flagged invisible. The internal dial bridge is flagged AST_BRIDGE_FLAG_INVISIBLE (it is created by ARI internally, not by the application), so recording is still permitted there. A "real" bridge created by the application - including a user-created ARI holding bridge - is not invisible, so recording stays rejected. This keeps the fix targeted at the internal dial bridge only, rather than allowing recording in every holding bridge. Resolves: #2044	6 days ago
Sean Bright	fd3d5759ba	chan_sip: Address Session Timer RFC compliance issues ... This commit addresses 2 issues with chan_sip's behavior as it relates to RFC 4028 (Session Timers): 1. When acting as a UAC, if Asterisk receives a response from the UAS that increases the Session-Expires refresh interval, Asterisk will honor that increase. This is not compliant with RFC 4028 Section 9. 2. When acting as a UAS, Asterisk includes the Session-Expires header in provisional responses (1xx) which is not specified by RFC 4028. This commit introduces 2 new global parameters to sip.conf to change this behavior. Asterisk's default behavior (as specified above) remains unchanged. Resolves: #2103 Resolves: #2104 UserNote: Two new global parameters have been added to chan_sip. `session-ignore-se-increase` and `session-include-in-provisional` can be used to influence how Asterisk handles Session Timers. See sip.conf.sample for more information.	6 days ago
OMAR.A	d5c837245b	alembic: Preserve sub-second precision in the queue_log time column. ... ast_queue_log() formats realtime queue_log timestamps with microseconds ("%F %T.%6q" in main/logger.c), but the queue_log table declares 'time' as a generic SQLAlchemy DateTime. That renders on MySQL and MariaDB as bare DATETIME, which has a fractional-seconds precision of 0, so MariaDB truncates the fraction and MySQL rounds it. Every event of a single queue entry can collapse onto one second, or move to the next one, and neither server warns, not even in strict mode. Once that happens, queue_log rows can no longer be ordered against each other, or against the CEL records that cel_odbc and cel_pgsql do stamp with microseconds. This adds a revision that widens the queue_log column to DATETIME(6) on MySQL and MariaDB. PostgreSQL is unaffected and left untouched: DateTime already renders as TIMESTAMP, which keeps microseconds by default. Fixes: #2072 UpgradeNote: (schema) On MySQL and MariaDB the queue_log 'time' column is widened from DATETIME to DATETIME(6) so that the microsecond timestamps Asterisk already writes are stored rather than discarded. This ALTER rewrites the table, a datetime precision change is neither INSTANT nor INPLACE, so queue_log writes will block for its duration on a large table. PostgreSQL installations are unaffected. Rows already written at whole-second precision will remain as-is and won't be updated.	1 week ago
Jeremy Lainé	43f6418ddd	main/file.c: Log the reason a file could not be opened. ... When ast_streamfile() failed to open a file it reported strerror(errno), but errno was not set by anything on that path. The value came from whatever syscall ran last, so the message routinely blamed an unrelated error. Drop errno from that message and log at the point of failure instead: filehelper() now reports the fopen() error, which does carry a meaningful errno, and an allocation failure from get_filestream(). Candidate files whose format does not match the channel are logged at debug level only, since filehelper() walks every registered format and skipping the ones that do not match is normal: a sound installed in several formats would otherwise warn on every successful playback. Failures inside the format module's open callback stay unlogged here, as those callbacks already report their own reason. Resolves: #1629	1 week ago
whitetreebug	249fac6f0f	chore: update .gitignore for clangd ... Ignore clangd generated artifacts: compile_commands.json, .cache Ignores for .clangd, .clang-format, .clang-tidy tooling files	1 week ago
Tinet-mucw	4b6f9db43e	func_speex: Hold channel lock while updating speex state. ... speex_write() previously unlocked the channel immediately after looking up the speex datastore, then continued to create, modify, or destroy SpeexPreprocessState and related direction data without the lock. speex_callback() runs from the media path with the channel already locked, so concurrent Set(DENOISE)/Set(AGC) could free or mutate that state while preprocess was running and crash inside speex_preprocess_run(). Keep the channel locked for the full configuration update, unlock before ast_audiohook_attach()/detach(), and hold the lock across speex_read() while copying values out of the datastore. Fixes: #2091	1 week ago
Maksym Tushkov	f5cf361ad6	pjsip_wizard.conf.sample: Add IP authentication and TLS examples. ... The only ITSP example in the file shows a single trunk that registers and authenticates. Many providers, including DIDWW which motivated this change, instead authenticate by IP address, deliver inbound calls from several regional POPs, and terminate outbound calls through a separate gateway that does require digest authentication. Several things about configuring that with the wizard are not obvious from the option reference: * remote_hosts alone generates the per-host identify matches, and is mandatory when registrations are not accepted. * Inbound and outbound need one wizard each, because a wizard has a single remote_hosts list. * A template carrying "type = wizard" is not turned into objects itself, so the two can share one. TLS has a pitfall of its own: the wizard cannot create transport objects, so the transport has to be defined in pjsip.conf and referenced here by name, while media_encryption belongs to the endpoint and needs the endpoint/ prefix. Add commented-out examples for an IP authenticated trunk pair sharing a template, and for a TLS trunk with SRTP. No functional change. UserNote: pjsip_wizard.conf.sample now includes examples for an ITSP that authenticates by IP address and delivers calls from several hosts, and for a TLS trunk with SRTP.	1 week ago
George Joseph	1381f3603a	res_http_websocket: Check for client before handling PONG frames. ... websocket_handled_pong_or_close() now checks that session->client is valid before trying to check missed_pong_count. Resolves: #2085	2 weeks ago
Michal Hajek	2cd6dbe0ba	res_pjsip_config_wizard: Fix concurrent Named ACL reloads ... Named ACL change notifications reloaded PJSIP sorcery directly from the Stasis callback. Under concurrent ACL, core, and config wizard reloads, that could overlap with config wizard object type observers that compare, prune, destroy, and replace the per-object-type last_config cache. Route the ACL-triggered reload through the PJSIP servant path and track when a Named ACL reload is pending or running so the observer still bypasses the unchanged-file optimization. Also serialize the observer cache update to prevent overlapping reload paths from mutating last_config at the same time. UserNote: Fixes a PJSIP config wizard reload hang during overlapping reloads.	2 weeks ago
Mehrdad Seifzadeh	cdc4e70bd1	res_pjsip_endpoint_identifier_ip: Clarify identify DNS warning ... When two DNS-based identify matches resolve to the same address, the first match makes the resolved address available for identify matching and the second match adds no new match address. The existing warning treated this as a DNS resolution failure and reported that the hostname did not resolve. Clarify the warning path so a hostname that resolves but only returns addresses already present for matching is not reported as unresolved. Fixes: #1723	2 weeks ago
Maksym Tushkov	5e7981dca9	res_pjsip_config_wizard: Fix documentation errors. ... The sample configuration and the module's XML documentation contained several errors, two of which actively misled users: * The sample said that sends_auth and accepts_auth require "outbound/username" and "inbound/username". The wizard looks for "outbound_auth/username" and "inbound_auth/username", so a configuration written by following the sample fails to load with "Wizard 'xxx' must have 'outbound_auth/username' if it sends authentication." * The sample claimed that hint_application defaults to "Dial(${HINT})". There is no such default. When hint_application is not specified the wizard removes the priority 1 extension it manages rather than creating one. The remaining corrections are a reversed object/field reference (match/identify), a leftover line in the hint_exten description, an option name (send_registrations), a stray space (remote _hosts), an unbalanced quote in a synopsis, a misspelling (nneds), an unbalanced bracket in the CLI usage text, and a copy/paste error in a log message that referred to the "sangoma wizard" instead of pjsip_wizard. Two "(default = ...)" annotations became "(default: ...)" to match the other fourteen in the same file. No functional change. UserNote: The pjsip_wizard.conf sample documented the wrong option names for sends_auth and accepts_auth, and claimed a default for hint_application that does not exist. Configurations written by following the sample would have failed to load.	2 weeks ago
Gian Diego Javes	a35939885b	taskpool: count the in-flight task when selecting an executor ... The taskpool selectors measure load with ast_taskprocessor_size(), which only reports queued tasks. A taskprocessor that is executing a long running task has an empty queue, so it looks identical to an idle one and the least full selector hands it more work while other executors sit free. Add ast_taskprocessor_is_executing() and include the in-flight task in the load the selectors compare. This is not a new semantic. taskprocessor_push() already counts the in-flight task as outstanding work when it notifies the listener: /* The currently executing task counts as still in queue */ was_empty = tps->executing ? 0 : previous_size == 0; The taskpool selectors are the one place that still measures only the queue. This change applies the same rule there. The accessor reads tps->executing without the object lock, the way the CLI already does in main/taskprocessor.c, because taking the lock on every push cost roughly six times the throughput: taskpool push efficiency went from 324438 to 52966 tasks per second and the serializer variant from 217467 to 38313. Without the lock both are back to baseline. A stale read only means one push is routed as if the taskprocessor had just changed state, which the selector already tolerates. Also add a unit test that holds one executor of a four executor pool inside a long task and checks that a task pushed to a different serializer runs on one of the three free executors. It fails on every run without this change. Fixes: #2074 UserNote: Taskpool executors that are running a long task are no longer considered idle when work is distributed, so tasks are routed to free executors instead of queueing behind a busy one.	2 weeks ago
aabolfazl	19481fb36b	app_voicemail: Fix crash when VM_INFO reads an unset email address. ... The email member of struct ast_vm_user is a pointer rather than a fixed array, and populate_defaults() leaves it NULL when a mailbox has no email address configured. Every other attribute VM_INFO reads is a fixed array, so only the email attribute is affected. VM_INFO passed vmu->email straight to ast_copy_string(), which dereferences its source unconditionally. Reading the email attribute of a mailbox that has no email address therefore crashed Asterisk from the dialplan. Guard the copy with S_OR() so an unset email address yields an empty string, matching how the language attribute already handles its fallback. The make_email_file() call sites were already guarded and are left alone. Add a regression test for the unset case, and restore the voicemail configuration when the VM_INFO test finishes. That test was the only one in app_voicemail that did not do so, which left its test mailbox in the users list and made the test fail if it ran a second time. Fixes: #2063	2 weeks ago
Sean Bright	a7417e30c8	res_agi.c: Prevent out-of-bounds array access when parsing arguments ... The `parse_args(...)` function writes to the `argv` array in 2 locations but was only bounds checking in one of them. Moved the bounds check so that it is encountered on each iteration through the parsing loop. Resolves: #2069	2 weeks ago
Mike Bradeen	516d2c60ac	res_pjsip_sdp_rtp: Fix intermittent call drop on reINVITE sdp change ... A reINVITE that changed the offered codec set (e.g. ulaw+alaw -> alaw only) could lead to call teardown when a bridge write still using the old format occurred during the re-negotiation. This change avoids the termination by making the following changes. First, Asterisk now holds the lock across both the tx payload update and the channel format update so they commit atomically. Second, when processing the sdp on the new incoming offer, merge the new offer with the old until the negotiation is complete. Also adds unit tests for the sdp merge Fixes: #2059	3 weeks ago
Sven Kube	f14ea0897a	chan_websocket: Fix NULL requestor dereference in webchan_request. ... Two log statements in webchan_request() passed `requestor` straight to ast_channel_name(), which dereferences the channel with no NULL check. `requestor` is NULL whenever no originator channel is supplied: ARI POST /channels/externalMedia always passes NULL, and POST /channels/create passes NULL when `originator` is omitted.	3 weeks ago
Naveen Albert	f6c77533ab	backtrace.c: Include stdbool.h ... stdbool.h is needed now that we use the bool type directly instead of bfd_boolean. See also 1d95b744c06974f0a00c143c6e0fc979af930908. Resolves: #2061	3 weeks ago
aabolfazl	ef8b52852d	res_musiconhold: Fix mohclass reference leak on answeredonly early return. ... local_ast_moh_start() returns -1 from the answeredonly check without releasing the mohclass reference it holds, unlike every other exit path in the function. Nothing else ever releases that reference, so moh_class_destructor() never runs for the object. With realtime music on hold and cachertclasses disabled, each suppressed request leaks the class object, its monitor thread, the external application process and two file descriptors for the lifetime of Asterisk. For static classes the stale references prevent the class from ever being destroyed after it is replaced by a reload. Release the reference before returning, matching the other exit paths. Fixes: #2051	3 weeks ago
Naveen Albert	82d907ee8e	backtrace.c: Avoid removed bfd_boolean type. ... bfd_boolean has been deprecated for some time and has now been removed in gdb, so use bool directly instead. See: 1d95b744c0 Resolves: #2054	3 weeks ago
Sean Bright	4f1b9d068e	app_voicemail.c: Fix may-be-uninitialized error ... A pointer to `new` is passed to `inboxcount(...)` which increments the pointed-to value.	3 weeks ago
Mike Bradeen	94f399a0e7	pjproject: disable building pjsua when a sanitizer is used ... Disable building the pjsua test application when any of the SANITIZER flags are set. Fixes: #2056	3 weeks ago
George Joseph	d7ca743235	SECURITY.md: Add warning about reporting multiple issues in one advisory.	3 weeks ago
Mehrdad Seifzadeh	73332e54c2	chan_pjsip: Store transport info for outgoing channels ... Outgoing PJSIP channels did not have transport information stored in their session datastore when CHANNEL(pjsip,local_addr) or CHANNEL(pjsip,remote_addr) was read. As a result, the address fields were empty on B-leg channels even though other dialog fields such as call-id and URIs were available. Store the selected local transport address and destination address for outgoing UAC session requests once the request is transmitted and the transport information is known. Fixes: #871	3 weeks ago
aabolfazl	2b94217990	func_strings: Fix misspelling of "occurred" in FIELDNUM documentation. ... The FIELDNUM description misspelled "occurred" as "occured". This text is user-visible: it is rendered on docs.asterisk.org and printed by "core show function FIELDNUM".	4 weeks ago
Sean Bright	6d7dc4ef7b	build: Fix GCC discarded-qualifiers const errors. ... GCC 15.2.1 pays attention to the discarding of the const qualifier when strchr, strrchr, memchr, or memrchr are now used. This change fixes numerous errors with this throughout chan_sip. The fixes can be broken down into the following: 1. The value passed to strchr or strrchr can be cast as it is expected and allowed to be modified. 2. The pointer passed to strchr or strrchr is not meant to be modified and so the contents must be duplicated. 3. It was declared const and never should have been.	4 weeks ago
Jeremy Lainé	cf0cd22301	tcptls: Don't inhibit escalations for outbound client connections. ... handle_tcptls_connection() marks the current thread as inhibiting privilege escalations and as an external user interface so that dialplan functions considered 'dangerous' (STAT, SHELL, ...) cannot be executed on behalf of an external protocol. This is correct for inbound (server) connections, each of which runs on its own dedicated worker thread. However, the outbound (client) path calls handle_tcptls_connection() synchronously on the caller's own thread. When that caller is a channel/PBX thread, the thread-local flags are set and never cleared, permanently tainting the dialplan thread. Examples: - Calling Dial() for an outbound WebSocket and resuming dialplan execution with the "g" flag. - Running ExternalIVR() then continuing dialplan executing. Outbound connections are initiated by Asterisk itself and are not external user interfaces, so the flags should not be set for them. Gate the flag-setting on tcptls_session->client so it applies only to inbound connections. Resolves: #2038	1 month ago
Mike Bradeen	e68c5c77a9	res_rtp_asterisk: Avoid two lock inversion deadlocks with pj project ... Fixes two related instance, group lock inversion deadlocks. When writing the RTP stream via rtp_sendto, the RTP instance was only unlocked before being passed to pj project when the instance was un-bundled (ie video was not bundled onto the audio stream's ICE transport.) The first change unlocks the instance whether bundled or unbundled to avoid the deadlock, then re-checks the bundled status upon return to know wether or not the transport should be used. For the second change, when an incoming RTCP NACK record was recieved, the parent (transport) and child (instance) locks were both held before the call to rtp_sendto, which is only able to release the child lock. Now in the un-bundled case the parent and child will be unlocked and re-locked in the correct order via a new helper function before and after the call to send_to. Fixes: #1946	1 month ago
Stanislav Abramenkov	ee58f361c7	jansson: Upgrade version to jansson 2.15.1 ... UpgradeNote: jansson has been upgraded to 2.15.1. For more information visit jansson Github page: https://github.com/akheron/jansson/releases/tag/v2.15.1 Resolves: #2033	1 month ago
Jeremy Lainé	1d164595d4	file.c: Ensure opening a stream opens at most one file ... The functions used to stream files eventually end up calling into `filehelper(.., ACTION_OPEN)` which takes care of iterating over supported extensions and opening the first existing file. To do this, `filehelper` uses two nested loops: - An outer loop over the supported file formats. - An inner loop over the possible extensions for that format. We need to break out of both loops as soon as a file was successfully opened. Otherwise if a file exists in multiple formats (e.g `foo.wav` and `foo.alaw`) both files will be opened successively. Resolves: #1625	1 month ago
Joshua C. Colp	9f24612ba8	pjsip: Eliminate some unique taskprocessors. ... When placing outbound calls each call would get its own taskprocessor for things related to the call. In practice this is overkill as few things actually occur. Inbound calls on the other hand already use one of the fixed number of distributor taskprocessors. This also occurred for OPTIONS requests with each AOR having its own taskprocessor. This change moves both to using a distributor taskprocessor instead.	1 month ago
Roel van Meer	2b85c53cc5	bridging: Add support for TOUCH_MIXMONITOR_OPTIONS ... This channel variable can be used to override the default automixmon options (which are 'b' or 'bB(<n>)'). If both TOUCH_MIXMONITOR_OPTIONS and TOUCH_MIXMONITOR_BEEP are set, TOUCH_MIXMONITOR_BEEP is ignored with a warning. It is the responsibility of the user to configure valid options in TOUCH_MIXMONITOR_OPTIONS. Fixes: #1981 UserNote: The TOUCH_MIXMONITOR_OPTIONS channel variable can now be used to configure the options used by the mixmonitor application when it is started via automixmon. If both TOUCH_MIXMONITOR_OPTIONS and TOUCH_MIXMONITOR_BEEP are set, TOUCH_MIXMONITOR_BEEP is ignored with a warning.	1 month ago
George Joseph	da3de0d905	res_pjsip_refer: Fix issues with Refer-Sub:false and ARI. ... When an attended transfer is controlled by ARI using PJSIP_TRANSFER_HANDLING()=ari-only, and a REFER request is received with a Refer-Sub: false header to suppress the automatic progress subscription, a segfault occurs in res_pjsip_refer:refer_incoming_ari_request(). However, even if the segfault were prevented, there's another issue... Suppressing the subscription and NOTIFYs to the referer also prevents the ARI app from getting any progress events for the transfer which it's controlling. So... * The segfault has been fixed. * Suppressing the subscription no longer suppress the ARI events from being sent to the controlling ARI app. * A good amount of tracing has also been added. Resolves: #2021	1 month ago
Mike Bradeen	ab63dece36	build: re-add pjsua test application ... pjsua and it's python bindings were removed as part of PR 1854. Testing for RTT requires the pjsua application be re-added to the third pary build process. This applies to the pjsua application ONLY and not the associated python bindings. Resolves: #2024	1 month ago
Naveen Albert	80e3ac64e5	res_pjsip: Document 'none' option for 'dtmf_mode'. ... The 'none' value can be used to disable DSP processing for DTMF on PJSIP channels and is sometimes necessary for this reason, and already exists in the code, but is not documented. Add it to the documentation enum. Resolves: #2026	1 month ago
Mike Bradeen	6916dfe360	Audiohooks: Allow whisper audiohooks to work without an underlying stream ... Audiohook whispers work by mixing audio with the current stream of frames being written to the hooked channel. Prior to this change if there were no audio frames being written to the channel outside of the audiohook, then any whisper audio has nothing to mix with and so is not heard on the hooked channel. This change removes that requirement by creating a framehook driven timer on the channel that creates silent frames the whispered frames can be mixed into. The framehook is automatically added with the first attached whisper audiohook and is removed when the last one is detached. If there are no attached streams being written to the channel, the timer is used to generate a default one. If a stream is attached; the timer is ignored until the stream stops. This stop and start is automatically triggered by the presence of a written stream and does not require any external action. Resolves: #1966 UserNote: There is no longer a requirement to play silence or otherwise have a stream of audio being sent to a hooked channel for whispered frames to be heard.	1 month ago
Ben Ford	394b324871	Logger: Don't discard WARNING or ERROR messages. ... When the logger reaches it's threshold for messages, don't discard the WARNING or ERROR messages up to a certain amount. If THAT threshold reaches its maximum, start discarding the oldest non-WARNING/ERROR messages until we only have WARNING and ERROR messages left. Everything will be discarded after that until we have room again for more. Also added a test that limits the queue sizes and logs many messages to see the queues hitting their maximums and discarding the appropriate messages. Fixes: #1882	2 months ago
Aqeel Abbas	2770e2bb83	func_strings: Fix syntax error in STRBETWEEN documentation example ... The STRBETWEEN function documentation example was missing a closing brace in the SendDTMF call. Corrected the syntax to ensure the dialplan example is functional.	2 months ago
Mehrdad Seifzadeh	a6fdedebae	res_pjsip_registrar: Resolve AOR for identified endpoints ... When an inbound REGISTER is matched to an endpoint using an identification method such as IP, header, or request URI, the endpoint can be identified and authenticated successfully while registrar AOR resolution still fails. find_registrar_aor() only resolved AOR names for username and auth_username identify methods. A header-only endpoint therefore left aor_name unset and failed registration with "AOR '' not found" even when the REGISTER To URI matched one of the endpoint's configured AORs. Resolve the REGISTER AOR from the To URI for endpoint identification methods that do not directly provide an AOR name, while still constraining the match to the already-identified endpoint's configured AOR list. Fixes: #1398	2 months ago
Thomas Guebels	0330d37340	res_http_websocket: Unref session when it fails to establish. ... Two error paths in res_http_websocket could leak an ast_websocket session by failing to drop a reference after the session had been allocated. The first occurs when session ID generation fails. While this is largely theoretical, as it would require an out-of-memory condition, the session reference should still be released correctly. The second occurs when the ast_websocket_pre_callback session_attempted callback returns an error. This path is reachable through ari_websockets, which implements this callback and can legitimately fail under certain conditions. Fixes: #2020	2 months ago
George Joseph	b4cd1ec359	Bundled pjproject: Make it easier to override options in config_site.h. ... Bundled pjproject uses ./third-party/pjproject/patches/config_site.h to set many pjproject build options like PJ_IOQUEUE_MAX_HANDLES, PJSIP_MAX_URL_SIZE, etc. Editing that file however, causes the git tree to become dirty which is inconvenient. So... * Updated the Asterisk configure scripts to pass the PJPROJECT_CFLAGS variable down to the pjproject configure scripts. See the UserNote below for details. * Updated ./third-party/pjproject/patches/config_site.h to allow the following options to be overridden: PJ_MAX_HOSTNAME, PJSIP_MAX_URL_SIZE, PJ_IOQUEUE_MAX_HANDLES. Other options in config_site.h are not overridable because they can have a bad effect on the overall operation of pjproject. This may be revisited in the future. Options not already set in config_site.h can still be set. * Fixed an issue where if the Linux `epoll` facility is used (which it is by default) the default PJ_IOQUEUE_MAX_HANDLES and PJSIP_MAX_TRANSPORTS were being left at 1024 instead of being icnreased to 5000. * The `pjproject show buildopts` CLI command previously only showed options from the top level `PJ` pjproject layer but now also shows many from the `PJSIP` layer. Many of these, such as PJSIP_MAX_TRANSPORTS, can be set using PJPROJECT_CFLAGS provided they're not already unconditionally set in config_site.h. * The `pjsip dump endpt` CLI command previously required that the pjproject log level be already set to at least 3 or no output would be produced. The command now does that automatically then sets it back to whatever it was. This isn't strictly related to this PR but was just nagging me. UserNote: Bundled pjproject: It's now possible to override some of the pjproject build options contained in ./third-party/pjproject/patches/config_site.h by adding PJPROJECT_CFLAGS to your Asterisk ./configure command line. For example: `./configure ... PJPROJECT_CFLAGS='-DPJ_OPT1=8192 -DPJ_OPT2=512'` Any option in config_site.h that's wrapped in a `#ifndef` block can be overridden and many of the `PJSIP` options displayed by `pjproject show buildopts` can be set. WARNING: Adjusting these options without understanding their effect can cripple your Asterisk instances. You shouldn't adjust them unless you need to solve a specific issue. Resolves: #2011	2 months ago
Joshua C. Colp	2363297479	test_performance: Add performance experimentation test module. ... This is a module which is only built when TEST_FRAMEWORK is enabled and provides CLI commands for testing performance of certain things on the system they are invoked on. The first CLI commands added cover the most common container usage: storage of objects with a lookup based on a string key. These commands take various arguments and allow you to see how they perform. There is also an "all" command named "performance test container_key_lookup_all" that will execute all of these container tests and pass through the given arguments, which makes it easy to run all of the tests for given usage. To facilitate a vector bsearch test a new AST_VECTOR_BSEARCH macro has been added that allows more efficient searching of sorted vectors.	2 months ago
Joshua C. Colp	0a774c6223	manager: Move away from shared linked list for events. ... This change moves manager away from a shared linked list for events to a per-session vector of pending events. This allows early filtering of events which eliminates the need to wake up threads for events that they have no interest in. TCP based connections have also been moved to an alert pipe based wakeup mechanism to reduce locking and simplify usage. UserNote: The "manager show eventq" CLI command has been removed as there is no longer a single event queue to display.	2 months ago
George Joseph	26551a4dc0	chan_websocket: Use leftover data if no frames are available when the timer fires. ... When the 20ms channel timer fires but there are no frames available in the queue, we now check for leftover data in the buffer and if there is any, we create a frame with it and send it to the core. This resolves an issue with the leftover data being delayed if a STOP_MEDIA_BUFFERING command is delayed. Some existing comments were also clarified to account for the new behavior. Resolves: #2006	2 months ago
Mehrdad Seifzadeh	05dc5b2d6f	res_pjsip_session: Bound delayed BYE behind UAC INVITE ... When a confirmed session is being terminated while an outgoing in-dialog INVITE transaction is still outstanding, the BYE is delayed until the outstanding transaction terminates. If that INVITE has already received a provisional response and the final response is malformed and rejected before transaction processing, the transaction can remain outstanding and the delayed BYE can keep the session, media state, RTP instance, and PJPROJECT pools referenced after the channels are gone. When a BYE is delayed behind an outstanding UAC INVITE, set a PJPROJECT transaction timeout on that INVITE so the delayed cleanup path has a bounded wait. If PJPROJECT terminates the dialog as a result of the timeout, discard the delayed BYE instead of sending a duplicate BYE. Fixes: #1965	2 months ago