Boris P. Korzun [Tue, 22 Feb 2022 20:51:02 +0000 (23:51 +0300)]
res_pjsip_sdp_rtp: Improve detecting of lack of RTP activity
Change RTP timer behavior for detecting RTP only after two-way
SDP channel establishment. Ignore detecting after receiving 183
with SDP or while direct media is used.
Make rtp_timeout and rtp_timeout_hold options consistent to rtptimeout
and rtpholdtimeout options in chan_sip.
Kevin Harwell [Wed, 23 Mar 2022 22:45:45 +0000 (17:45 -0500)]
res_pjsip_header_funcs: wrong pool used tdata headers
When adding headers to an outgoing request the headers were cloned using
the dialog's pool when they should have been cloned using tdata's pool.
Under certain circumstances it was possible for the dialog object, and
its pool to be freed while tdata is still active and available. Thus the
cloned header "disappeared", and when tdata tried to later access it a
crash would occur.
This patch makes it so all added headers are cloned appropriately using
tdata's pool.
Joshua C. Colp [Thu, 24 Feb 2022 17:48:19 +0000 (13:48 -0400)]
pjproject: Update bundled to 2.12 release.
This change removes patches which have been merged into
upstream and updates some existing ones. It also adds
some additional config_site.h changes to restore previous
behavior, as well as a patch to allow multiple Authorization
headers. There seems to be some confusion or disagreement
on language in RFC 8760 in regards to whether multiple
Authorization headers are supported. The RFC implies it
is allowed, as does some past sipcore discussion. There is
also the catch all of "local policy" to allow it. In
the case of Asterisk we allow it.
Naveen Albert [Sat, 5 Mar 2022 16:26:42 +0000 (16:26 +0000)]
pbx.c: Warn if there are too many includes in a context.
The PBX core uses the stack when it comes to includes, which
means that a context can only contain strictly fewer than
AST_PBX_MAX_STACK includes. If this is exceeded, then warnings
will be emitted for each number of includes beyond this if
searching for an extension in the including context, and if
the extension's inclusion is beyond the stack size, it will
simply not be found.
To address this, we now check if there are too many includes
in a context when the dialplan is reloaded so that if there
is an issue, the user is aware of at "compile time" as opposed
to "run time" only. Secondly, more details are printed out
when this message is encountered so it's clear what has happened.
George Joseph [Fri, 25 Mar 2022 14:33:30 +0000 (08:33 -0600)]
make_xml_documentation: Remove usage of get_sourceable_makeopts
get_sourceable_makeopts wasn't handling variables with embedded
double quotes in them very well. One example was the DOWNLOAD
variable when curl was being used instead of wget. Rather than
trying to fix get_sourceable_makeopts, it's just been removed.
George Joseph [Fri, 25 Mar 2022 19:00:48 +0000 (13:00 -0600)]
Makefile: Disable XML doc validation
make_xml_documentation was being called with the --validate
flag set when it shouldn't have been. This was causing
build failures if neither xmllint nor xmlstarlet were installed.
The correct behavior is to simply print a message that either
one of those tools should be installed for validation and
continue with the build.
Naveen Albert [Sat, 5 Feb 2022 00:36:17 +0000 (00:36 +0000)]
chan_iax2: Fix spacing in netstats command
The iax2 show netstats command previously didn't contain
enough spacing in the header to properly align the table
header with the table body. This caused column headers
to not align with the values on longer channel names.
Some spacing is added to account for the longest channel
names that display (before truncation occurs) so that
columns are always properly aligned.
ASTERISK-29895 #close
patches:
61205_misaligned2.patch submitted by Birger Harzenetter (license 5870)
Naveen Albert [Fri, 25 Feb 2022 17:01:41 +0000 (17:01 +0000)]
res_agi: Fix xmldocs bug with set music.
The XML documentation for the SET MUSIC AGI
command is invalid, as the parameter does not
have a name and the on/off enum options for
the on/off argument are listed separately, which
is incorrect. The cumulative effect of these currently
is that the Asterisk Wiki documentation for SET MUSIC
is broken and external documentation generators crash
on SET MUSIC due to the malformed documentation.
These issues are corrected so that the documentation
can be successfully parsed as with other similar AGI
commands.
Naveen Albert [Sun, 13 Mar 2022 17:46:36 +0000 (17:46 +0000)]
chan_iax2: Fix perceived showing host address.
ASTERISK_22025 introduced a regression that shows
the host IP and port as the perceived IP and port
again, as opposed to showing the actual perceived
address. This fixes this by showing the correct
information.
Sean Bright [Fri, 4 Mar 2022 20:26:22 +0000 (15:26 -0500)]
conversions.c: Specify that we only want to parse decimal numbers.
Passing 0 as the last argument to strtoimax() or strtoumax() causes
octal and hexadecimal to be accepted which was not originally
intended. So we now force to only accept decimal.
MUSL defines BUFSIZ as 1024 which is not reasonable for log messages.
More broadly, BUFSIZ is the amount of buffering stdio.h does, which
is arbitrary and largely orthogonal to what logging should accept
as the maximum message size.
ASTERISK-29928
Signed-off-by: Philip Prindeville <philipp@redfish-solutions.com>
Change-Id: Iaa49fbbab029c64ae3d95e4b18270e0442cce170
Sean Bright [Thu, 10 Mar 2022 17:07:40 +0000 (12:07 -0500)]
stasis_recording: Perform a complete match on requested filename.
Using the length of a file found on the filesystem rather than the
file being requested could result in filenames whose names are
substrings of another to be erroneously matched.
We now ensure a complete comparison before returning a positive
result.
Alexei Gradinari [Tue, 15 Mar 2022 17:24:18 +0000 (13:24 -0400)]
res_pjsip_pubsub: RLS 'uri' list attribute mismatch with SUBSCRIBE request
When asterisk generates the RLMI part of NOTIFY request,
the asterisk uses the local contact uri instead of the URI to which
the SUBSCRIBE request is sent.
Because of this mismatch some IP phones (for example Cisco 5XX) ignore
this list.
According
https://datatracker.ietf.org/doc/html/rfc4662#section-5.2
The first mandatory <list> attribute is "uri", which contains the uri
that corresponds to the list. Typically, this is the URI to which
the SUBSCRIBE request was sent.
https://datatracker.ietf.org/doc/html/rfc4662#section-5.3
The "uri" attribute identifies the resource to which the <resource>
element corresponds. Typically, this will be a SIP URI that, if
subscribed to, would return the state of the resource.
This patch makes asterisk to generate URI using SUBSCRIBE request URI.
Naveen Albert [Mon, 14 Mar 2022 16:57:29 +0000 (16:57 +0000)]
pbx_builtins: Add missing options documentation
BackGround and WaitExten both accept options that are not
currently documented. This adds documentation for these
options to the xml documentation for each application.
res_pjsip_pubsub: update RLS to reflect the changes to the lists
This patch makes the Resource List Subscriptions (RLS) dynamic.
The asterisk updates the current subscriptions to reflect the changes
to the list on the subscriptions refresh. If list items are added,
removed, updated or do not exist anymore, the asterisk regenerates
the resource list.
Kfir Itzhak [Wed, 9 Feb 2022 10:28:29 +0000 (12:28 +0200)]
app_queue: Add QueueWithdrawCaller AMI action
This adds a new AMI action called QueueWithdrawCaller.
This AMI action makes it possible to withdraw a caller from a queue,
in a safe and a generic manner.
This can be useful for retrieving a specific call and
dispatching it to a specific extension.
It works by signaling the caller to exit the queue application
whenever it can. Therefore, it is not guaranteed
that the call will leave the queue.
George Joseph [Wed, 2 Mar 2022 14:57:26 +0000 (07:57 -0700)]
xml.c, config,c: Add stylesheets and variable list string parsing
Added functions to open, close, and apply XML Stylesheets
to XML documents. Although the presence of libxslt was already
being checked by configure, it was only happening if xmldoc was
enabled. Now it's checked regardless.
Added ability to parse a string consisting of comma separated
name/value pairs into an ast_variable list. The reverse of
ast_variable_list_join().
George Joseph [Tue, 1 Mar 2022 16:58:44 +0000 (09:58 -0700)]
xmldoc: Fix issue with xmlstarlet validation
Added the missing xml-stylesheet and Xinclude namespace
declarations in pjsip_config.xml and pjsip_manager.xml.
Updated make_xml_documentation to show detailed errors when
xmlstarlet is the validator. It's now run once with the '-q'
option to suppress harmless/expected messages and if it actually
fails, it's run again without '-q' but with '-e' to show
the actual errors.
George Joseph [Sun, 20 Feb 2022 20:16:22 +0000 (13:16 -0700)]
core: Config and XML tweaks needed for geolocation
Added:
Replace a variable in a list:
int ast_variable_list_replace_variable(struct ast_variable **head,
struct ast_variable *old, struct ast_variable *new);
Added test as well.
Create a "name=value" string from a variable list:
'name1="val1",name2="val2"', etc.
struct ast_str *ast_variable_list_join(
const struct ast_variable *head, const char *item_separator,
const char *name_value_separator, const char *quote_char,
struct ast_str **str);
Added test as well.
Allow the name of an XML element to be changed.
void ast_xml_set_name(struct ast_xml_node *node, const char *name);
George Joseph [Mon, 14 Feb 2022 13:31:25 +0000 (06:31 -0700)]
Makefile: Allow XML documentation to exist outside source files
Moved the xmldoc build logic from the top-level Makefile into
its own script "make_xml_documentation" in the build_tools
directory.
Created a new utility script "get_sourceable_makeopts", also in
the build_tools directory, that dumps the top-level "makeopts"
file in a format that can be "sourced" from shell sscripts.
This allows scripts to easily get the values of common make
build variables such as the location of the GREP, SED, AWK, etc.
utilities as well as the AST* and library *_LIB and *_INCLUDE
variables.
Besides moving logic out of the Makefile, some optimizations
were done like removing "third-party" from the list of
subdirectories to be searched for documentation and changing some
assignments from "=" to ":=" so they're only evaluated once.
The speed increase is noticeable.
The makeopts.in file was updated to include the paths to
REALPATH and DIRNAME. The ./conifgure script was setting them
but makeopts.in wasn't including them.
So...
With this change, you can now place documentation in any"c"
source file AND you can now place it in a separate XML file
altogether. The following are examples of valid locations:
res/res_pjsip.c
Using the existing /*** DOCUMENTATION ***/ fragment.
res/res_pjsip/pjsip_configuration.c
Using the existing /*** DOCUMENTATION ***/ fragment.
res/res_pjsip/pjsip_doc.xml
A fully-formed XML file. The "configInfo", "manager",
"managerEvent", etc. elements that would be in the "c"
file DOCUMENTATION fragment should be wrapped in proper
XML. Example for "somemodule.xml":
It's the "appdocsxml.dtd" that tells make_xml_documentation
that this is a documentation XML file and not some other XML file.
It also allows many XML-capable editors to do formatting and
validation.
Other than the ".xml" suffix, the name of the file is not
significant.
As a start... This change also moves the documentation that was
in res_pjsip.c to 2 new XML files in res/res_pjsip:
pjsip_config.xml and pjsip_manager.xml. This cut the number of
lines in res_pjsip.c in half. :)
George Joseph [Thu, 17 Feb 2022 16:26:46 +0000 (09:26 -0700)]
build: Refactor the earlier "basebranch" commit
Recap from earlier commit: If you have a development branch for a
major project that will receive gerrit reviews it'll probably be
named something like "development/16/newproject" or a work branch
based on that "development" branch. That will necessitate
setting "defaultbranch=development/16/newproject" in .gitreview.
The make_version script uses that variable to construct the
asterisk version however, which results in versions
like "GIT-development/16/newproject-ee582a8c7b" which is probably
not what you want. It also constructs the URLs for downloading
external modules with that version, which will fail.
Fast-forward:
The earlier attempt at adding a "basebranch" variable to
.gitreview didn't work out too well in practice because changes
were made to .gitreview, which is a checked-in file. So, if
you wanted to rebase your work branch on the base branch, rebase
would attempt to overwrite your .gitreview with the one from
the base branch and complain about a conflict.
This is a slighltly different approach that adds three methods to
determine the mainline branch:
1. --- MAINLINE_BRANCH from the environment
If MAINLINE_BRANCH is already set in the environment, that will
be used. This is primarily for the Jenkins jobs.
2. --- .develvars
Instead of storing the basebranch in .gitreview, it can now be
stored in a non-checked-in ".develvars" file and keyed by the
current branch. So, if you were working on a branch named
"new-feature-work" based on "development/16/new-feature" and wanted
to push to that branch in Gerrit but wanted to pull the external
modules for 16, you'd create the following .develvars file:
[branch "new-feature-work"]
mainline-branch = 16
The .gitreview file would still look like:
[gerrit]
defaultbranch=development/16/new-feature
...which would cause any reviews pushed from "new-feature-work" to
go to the "development/16/new-feature" branch in Gerrit.
The key is that the .develvars file is NEVER checked in (it's been
added to .gitignore).
3. --- Well Known Development Branch
If you're actually working in a branch named like
"development/<mainline_branch>/some-feature", the mainline branch
will be parsed from it.
4. --- .gitreview
If none of the earlier conditions exist, the .gitreview
"defaultbranch" variable will be used just as before.
Naveen Albert [Sun, 9 Jan 2022 13:32:48 +0000 (13:32 +0000)]
ami: Allow events to be globally disabled.
The disabledevents setting has been added to the general section
in manager.conf, which allows users to specify events that
should be globally disabled and not sent to any AMI listeners.
This allows for processing of these AMI events to end sooner and,
for frequent AMI events such as Newexten which users may not have
any need for, allows them to not be processed. Additionally, it also
cleans up core debug as previously when debug was 3 or higher,
the debug was constantly spammed by "Analyzing AMI event" messages
along with a complete dump of the event contents (often for Newexten).
Naveen Albert [Sat, 5 Feb 2022 01:27:27 +0000 (01:27 +0000)]
channel.c: Clean up debug level 1.
Although there are 10 debugs levels, over time,
many current debug calls have come to use
inappropriately low debug levels. In particular,
a select few debug calls (currently all debug 1)
can result in thousands of debug messages per minute
for a single call.
This can adds a lot of noise to core debug
which dilutes the value in having different
debug levels in the first place, as these
log messages are from the core internals are
are better suited for higher debug levels.
Some debugs levels are thus adjusted so that
debug level 1 is not inappropriately overloaded
with these extremely high-volume and general
debug messages.
Naveen Albert [Sat, 5 Feb 2022 01:11:43 +0000 (01:11 +0000)]
documentation: Add since tag to xmldocs DTD
Adds the since tag to the documentation DTD so
that individual applications, functions, etc.
can now specify when they were added to Asterisk.
This tag is added at the individual application,
function, etc. level as opposed to at the module
level because modules can expand over time as new
functionality is added, and granularity only
to the module level would generally not be useful.
This enables the ability to more easily determine
when new functionality was added to Asterisk, down
to minor version as opposed to just by major version.
This makes it easier for users to write more portable
dialplan if desired to not use functionality that may
not be widely available yet.
Naveen Albert [Wed, 16 Feb 2022 11:34:34 +0000 (11:34 +0000)]
app_voicemail: Emit warning if asking for nonexistent mailbox.
Currently, if VoiceMailMain is called with a mailbox, if that
mailbox doesn't exist, then the application silently falls back
to prompting the user for the mailbox, as if no arguments were
provided.
However, if a specific mailbox is requested and it doesn't exist,
then no warning at all is emitted.
This fixes this behavior to now warn if a specifically
requested mailbox could not be accessed, before falling back to
prompting the user for the correct mailbox.
res_pjsip_pubsub: fix Batched Notifications stop working
If Subscription refresh occurred between when the batched notification
was scheduled and the serialized notification was to be sent,
then new schedule notification task would never be added.
There are 2 threads:
thread #1. ast_sip_subscription_notify is called,
if notification_batch_interval then call schedule_notification.
1.1. The schedule_notification checks notify_sched_id > -1
not true, then
send_scheduled_notify = 1
notify_sched_id =
ast_sched_add(sched, sub_tree->notification_batch_interval, sched_cb....
1.2. The sched_cb pushes task serialized_send_notify to serializer
and returns 0 which means no reschedule.
1.3. The serialized_send_notify checks send_scheduled_notify if it's false
the just returns. BUT notify_sched_id is still set, so no more ast_sched_add.
thread #2. pubsub_on_rx_refresh is called
2.1 it pushes serialized_pubsub_on_refresh_timeout to serializer
2.2. The serialized_pubsub_on_refresh_timeout calls pubsub_on_refresh_timeout
which calls send_notify
2.3. The send_notify set send_scheduled_notify = 0;
The serialized_send_notify should always unset notify_sched_id.
Naveen Albert [Fri, 18 Feb 2022 12:09:47 +0000 (12:09 +0000)]
func_db: Add validity check for key names when writing.
Adds a simple sanity check for key names when users are
writing data to AstDB. This captures four cases indicating
malformed keynames that generally result in bad data going
into the DB that the user didn't intend: an empty key name,
a key name beginning or ending with a slash, and a key name
containing two slashes in a row. Generally, this is the
result of a variable being used in the key name being empty.
If a malformed key name is detected, a warning is emitted
to indicate the bug in the dialplan.
res_pjsip_pubsub: provide a display name for RLS subscriptions
Whereas BLFs allow to show a display name for each RLS entry,
the asterisk provides only the extension now.
This is not end user friendly.
This commit adds a new resource_list option, resource_display_name,
to indicate whether display name of resource or the resource name being
provided for RLS entries.
If this option is enabled, the Display Name will be provided.
This option is disabled by default to remain the previous behavior.
If the 'event' set to 'presence' or 'dialog' the non-empty HINT name
will be set as the Display Name.
The 'message-summary' is not supported yet.
Naveen Albert [Sat, 5 Feb 2022 01:46:27 +0000 (01:46 +0000)]
documentation: Adds missing default attributes.
The configObject tag contains a default attribute which
allows the default value to be specified, if applicable.
This allows for the default value to show up specially on
the wiki in a way that is clear to users.
There are a couple places in the tree where default values
are included in the description as opposed to as attributes,
which means these can't be parsed specially for the wiki.
These are changed to use the attribute instead of being
included in the text description.
Naveen Albert [Sat, 5 Feb 2022 12:39:42 +0000 (12:39 +0000)]
app_mp3: Document and warn about HTTPS incompatibility.
mpg123 doesn't support HTTPS, but the MP3Player application
doesn't document this or warn the user about this. HTTPS
streams have become more common nowadays and users could
reasonably try to play them without being aware they should
use the HTTP stream instead.
This adds documentation to note this limitation. It also
throws a warning if users try to use the HTTPS stream to
tell them to use the HTTP stream instead.
Naveen Albert [Sat, 22 Jan 2022 22:52:45 +0000 (22:52 +0000)]
app_mf: Add max digits option to ReceiveMF.
Adds an option to the ReceiveMF application to allow specifying a
maximum number of digits.
Originally, this capability was not added to ReceiveMF as it was
with ReceiveSF because typically a ST digit is used to denote that
sending of digits is complete. However, there are certain signaling
protocols which simply transmit a digit (such as Expanded In-Band
Signaling) and for these, it's necessary to be able to read a
certain number of digits, as opposed to until receiving a ST digit.
This capability is added as an option, as opposed to as a parameter,
to remain compatible with existing usage (and not shift the
parameters).
Mike Bradeen [Thu, 3 Feb 2022 01:18:06 +0000 (18:18 -0700)]
taskprocessor.c: Prevent crash on graceful shutdown
When tps_shutdown is called as part of the cleanup process there is a
chance that one of the taskprocessors that references the
tps_singletons object is still running. The change is to allow for
tps_shutdown to check tps_singleton's container count and give the
running taskprocessors a chance to finish. If after
AST_TASKPROCESSOR_SHUTDOWN_MAX_WAIT (10) seconds there are still
container references we shutdown anyway as this is most likely a bug
due to a taskprocessor not being unreferenced.
Alexei Gradinari [Fri, 21 Jan 2022 19:00:53 +0000 (14:00 -0500)]
app_queue: load queues and members from Realtime when needed
There are a lot of Queue AMI actions and Queue applications
which do not load queue and queue members from Realtime.
AMI actions
QueuePause - if queue not in memory - response "Interface not found".
QueueStatus/QueueSummary - if queue not in memory - empty response.
Applications:
PauseQueueMember - if queue not in memory
Attempt to pause interface %s, not found
UnpauseQueueMember - if queue not in memory
Attempt to unpause interface xxxxx, not found
This patch adds a new function load_realtime_queues
which loads queue and queue members for desired queue
or all queues and all members if param 'queuename' is NULL or empty.
Calls the function load_realtime_queues when needed.
Also this patch fixes leak of ast_config in function set_member_value.
Also this patch fixes incorrect LOG_WARNING when pausing/unpausing
already paused/unpaused member.
The function ast_update_realtime returns 0 when no record modified.
So 0 is not an error to warn about.
Sean Bright [Mon, 31 Jan 2022 18:52:26 +0000 (13:52 -0500)]
manager.c: Generate valid XML if attribute names have leading digits.
The XML Manager Event Interface (amxml) now generates attribute names
that are compliant with the XML 1.1 specification. Previously, an
attribute name that started with a digit would be rendered as-is, even
though attribute names must not begin with a digit. We now prefix
attribute names that start with a digit with an underscore ('_') to
prevent XML validation failures.
This is not backwards compatible but my assumption is that compliant
XML parsers would already have been complaining about this.
Added the following APIs:
pjsip_multipart_find_part_by_header()
pjsip_multipart_find_part_by_header_str()
pjsip_multipart_find_part_by_cid_str()
pjsip_multipart_find_part_by_cid_uri()
George Joseph [Wed, 26 Jan 2022 13:56:15 +0000 (06:56 -0700)]
build: Add "basebranch" to .gitreview
If you have a development branch for a major project that
will receive gerrit reviews it'll probably be named something
like "development/16/newproject". That will necessitate setting
"defaultbranch=development/16/newproject" in .gitreview. The
make_version script uses that variable to construct the asterisk
version however, which results in versions like
"GIT-development/16/newproject-ee582a8c7b" which is probably not
what you want. Worse, since the download_externals script uses
make_version to construct the URL to download the binary codecs
or DPMA. Since it's expecting a simple numeric version, the
downloads will fail.
To get this to work, a new variable "basebranch" has been added
to .gitreview and make_version has been updated to use that instead
of defaultversion:
Now git-review will send the reviews to the proper branch
(development/16/myproject) but the version will still be
constructed using the simple branch number (16).
If "basebranch" is missing from .gitreview, make_version will
fall back to using "defaultbranch".
George Joseph [Mon, 31 Jan 2022 13:09:09 +0000 (06:09 -0700)]
res_pjsip_outbound_authenticator_digest: Prevent ABRT on cleanup
In dev mode, if you call pjsip_auth_clt_deinit() with an auth_sess
that hasn't been initialized, it'll assert and abort. If
digest_create_request_with_auth() fails to find the proper
auth object however, it jumps to its cleanup which does exactly
that. So now we no longer attempt to call pjsip_auth_clt_deinit()
if we never actually initialized it.
Naveen Albert [Wed, 15 Dec 2021 18:36:42 +0000 (18:36 +0000)]
cdr: allow disabling CDR by default on new channels
Adds a new option, defaultenabled, to the CDR core to
control whether or not CDR is enabled on a newly created
channel. This allows CDR to be disabled by default on
new channels and require the user to explicitly enable
CDR if desired. Existing behavior remains unchanged.
In order to get around the issue of certain frames
having names that could overlap, func_frame_drop
surrounds names with commas for the purposes of
comparison.
The buffer is allocated and printed to properly,
but the original buffer is used for comparison.
In most cases, this wouldn't have had any effect,
but that was not the intention behind the buffer.
This updates the code to reference the modified
buffer instead.
Torrey Searle [Thu, 20 Jan 2022 12:56:27 +0000 (13:56 +0100)]
res/res_rtp_asterisk: fix skip in rtp sequence numbers after dtmf
When generating dtmfs, asterisk can incorrectly think packet loss
occured during the dtmf generation, resulting in a jump in sequence
numbers when forwarding voice frames resumes. This patch forces
asterisk to re-learn the expected sequence number after each DTMF
to avoid this
Sean Bright [Fri, 21 Jan 2022 16:34:38 +0000 (11:34 -0500)]
build: Rebuild configure and autoconfig.h.in
autoconfigh.h.in was missed in the original review for this
issue. Additionally it looks like I have newer pkg-config autoconf
macros on my development machine.
Kevin Harwell [Thu, 13 Jan 2022 22:31:27 +0000 (16:31 -0600)]
res_http_websocket: Add a client connection timeout
Previously there was no way to specify a connection timeout when
attempting to connect a websocket client to a server. This patch
makes it possible to now do such.
Mike Bradeen [Wed, 8 Dec 2021 21:14:48 +0000 (14:14 -0700)]
sched: fix and test a double deref on delete of an executing call back
sched: Avoid a double deref when AST_SCHED_DEL_UNREF is called on an
executing call-back. This is done by adding a new variable 'rescheduled'
to the struct sched which is set in ast_sched_runq and checked in
ast_sched_del_nonrunning. ast_sched_del_nonrunning is a replacement for
now deprecated ast_sched_del which returns a new possible value -2
if called on an executing call-back with rescheduled set. ast_sched_del
is modified to call ast_sched_del_nonrunning to maintain existing code.
AST_SCHED_DEL_UNREF is also updated to look for the -2 in which case it
will not throw a warning or invoke refcall.
test_sched: Add a new unit test sched_test_freebird that will check the
reference count in the resolved scenario.
Michał Górny [Thu, 11 Nov 2021 03:40:40 +0000 (04:40 +0100)]
build_tools/make_version: Fix sed(1) syntax compatibility with NetBSD
Fix the sed(1) invocation used to process git-svn-id not to use "\s"
that is a GNU-ism and is not supported by NetBSD sed. As a result,
this call did not work properly and make_version did output the full
git-svn-id line rather than the revision.
Michał Górny [Thu, 11 Nov 2021 04:06:27 +0000 (05:06 +0100)]
include: Remove unimplemented HMAC declarations
Remove the HMAC declarations from the includes. They are
not implemented nor used anywhere, and their presence breaks the build
on NetBSD that delivers an incompatible hmac() function in <stdlib.h>.
Naveen Albert [Tue, 11 Jan 2022 18:46:08 +0000 (18:46 +0000)]
res_rtp_asterisk: Fix typo in flag test/set
The code currently checks to see if an RFC3389
warning flag is set, except if it is, it merely
sets the flag again, the logic of which doesn't
make any sense.
This adjusts the if comparison to check if the
flag has NOT been set, and if so, emit a notice
log event and set the flag so that future frames
do not cause an event to be logged.
Michał Górny [Thu, 11 Nov 2021 02:05:02 +0000 (03:05 +0100)]
BuildSystem: Fix misdetection of gethostbyname_r() on NetBSD
Fix the configure script not to detect the presence of gethostbyname_r()
on NetBSD incorrectly. NetBSD includes it as an internal libc symbol
that is not exposed in system headers and that is incompatible with
other implementations. In order to avoid misdetecting it, perform
the symbol check only if the declaration is found in the public header
first.
George Joseph [Tue, 18 Jan 2022 14:04:24 +0000 (07:04 -0700)]
bundled_pjproject: Fix srtp detection
Reverted recent change that set '--with-external-srtp' instead
of '--without-external-srtp'. Since Asterisk handles all SRTP,
we don't need it enabled in pjproject at all.
George Joseph [Mon, 10 Jan 2022 13:44:12 +0000 (06:44 -0700)]
res_pjsip: Make message_filter and session multipart aware
Neither pjsip_message_filter's filter_on_tx_message() nor
res_pjsip_session's session_outgoing_nat_hook() were multipart
aware and just assumed that an SDP would be the only thing in
a message body. Both were changed to use the new
pjsip_get_sdp_info() function which searches for an sdp in
both single- and multi- part message bodies.
George Joseph [Wed, 12 Jan 2022 17:12:01 +0000 (10:12 -0700)]
build: Fix issues building pjproject
The change to allow easier hacking on bundled pjproject created
a few issues:
* The new Makefile was trying to run the bundled make even if
PJPROJECT_BUNDLED=no. third-party/Makefile now checks for
PJPROJECT_BUNDLED and JANSSON_BUNDLED and skips them if they
are "no".
* When building with bundled, config_site.h was being copied
only if a full make or a "make main" was done. A "make res"
would fail all the pjsip modules because they couldn't find
config_site.h. The Makefile now copies config_site.h and
asterisk_malloc_debug.h into the pjproject source tree
when it's "configure" is performed. This is how it used
to be before the big change.
pjsip_msg_find_hdr(), pjsip_msg_find_hdr_by_name(), and
pjsip_msg_find_hdr_by_names() require a pjsip_msg to be passed in
so if you need to search a header list that's not in a pjsip_msg,
you have to do it yourself. This commit adds generic versions of
those 3 functions that take in the actual header list head instead
of a pjsip_msg so if you need to search a list of headers in
something like a pjsip_multipart_part, you can do so easily.
Sean Bright [Wed, 12 Jan 2022 19:20:34 +0000 (14:20 -0500)]
say.c: Prevent erroneous failures with 'say' family of functions.
A regression was introduced in ASTERISK~29531 that caused 'say'
functions to fail with file lists that would previously have
succeeded. This caused affected channels to hang up where previously
they would have continued.
We now explicitly check for the empty string to restore the previous
behavior.
Naveen Albert [Sat, 8 Jan 2022 20:35:09 +0000 (20:35 +0000)]
documentation: Document built-in system and channel vars
Documentation for built-in special system and channel
vars is currently outdated, and updating is a manual
process since there is no XML documentation for these
anywhere.
This adds documentation for system vars to func_env
and for channel vars to func_channel so that they
appear along with the corresponding fields that would
be accessed using a function.
Naveen Albert [Sat, 8 Jan 2022 15:09:13 +0000 (15:09 +0000)]
pbx_variables: add missing ASTSBINDIR variable
Every config variable in the directories
section of asterisk.conf currently has a
counterpart built-in variable containing
the value of the config option, except
for the last one, astsbindir, which should
have an ASTSBINDIR variable.
However, the actual corresponding ASTSBINDIR
variable is missing in pbx_variables.c.
This adds the missing variable so that all
the config options have their corresponding
variable.
George Joseph [Tue, 30 Nov 2021 22:35:27 +0000 (15:35 -0700)]
bundled_pjproject: Make it easier to hack
There are times when you need to troubleshoot issues with bundled
pjproject or add new features that need to be pushed upstream
but...
* The source directory created by extracting the pjproject tarball
is not scanned for code changes so you have to keep forcing
rebuilds.
* The source directory isn't a git repo so you can't easily create
patches, do git bisects, etc.
* Accidentally doing a make distclean will ruin your day by wiping
out the source directory, and your changes.
* etc.
This commit makes that easier.
See third-party/pjproject/README-hacking.md for the details.
Naveen Albert [Wed, 15 Dec 2021 14:23:06 +0000 (14:23 +0000)]
cli: Add module refresh command
Adds a command to the CLI to unload and then
load a module. This makes it easier to perform
these operations which are often done
subsequently to load a new version of a module.
"module reload" already refers to reloading of
configuration, so the name "refresh" is chosen
instead.
Naveen Albert [Mon, 3 Jan 2022 01:13:33 +0000 (01:13 +0000)]
app_mp3: Throw warning on nonexistent stream
Currently, the MP3Player application doesn't
emit a warning if attempting to play a stream
which no longer exists. This can be a common
scenario as many mp3 streams are valid at some
point but can disappear at any time.
Now a warning is thrown if attempting to play
a nonexistent MP3 stream, instead of silently
exiting.
Kevin Harwell [Mon, 15 Nov 2021 22:13:19 +0000 (16:13 -0600)]
tcptls.c: refactor client connection to be more robust
The current TCP client connect code, blocks and does not handle EINTR
error case.
This patch makes the client socket non-blocking while connecting,
ensures a connect does not immediately fail due to EINTR "errors",
and adds a connect timeout option.
The original client start call sets the new timeout option to
"infinite", thus making sure old, orginal behavior is retained.
Steve Davies [Wed, 15 Dec 2021 12:23:45 +0000 (12:23 +0000)]
app_queue: Fix hint updates, allow dup. hints
A previous patch for ASTERISK_29578 caused a 'leak' of
extension state information across queues, causing the
state of the first member of unrelated queues to be
updated in addition to the correct member. Which queues
and members depended on the order of queues in the
iterator.
Additionally, it is possible to use the same 'hint:' on
multiple queue members, so the update cannot break out
of the update loop early when a match is found.
Sean Bright [Thu, 23 Dec 2021 21:57:59 +0000 (16:57 -0500)]
say.c: Honor requests for DTMF interruption.
SayAlpha, SayAlphaCase, SayDigits, SayMoney, SayNumber, SayOrdinal,
and SayPhonetic all claim to allow DTMF interruption if the
SAY_DTMF_INTERRUPT channel variable is set to a truthy value, but we
are failing to break out of a given 'say' application if DTMF actually
occurs.
projects / thirdparty / asterisk.git / log
