Alexander Traud [Mon, 4 Dec 2017 11:27:02 +0000 (12:27 +0100)]
chan_sip: Peers with distinct source ports don't match, regardless of transport.
Previously, peers connected via TCP (or TLS) were matched by ignoring their
source port. One cannot say anything when protocol:IP:port match, yes (see
<http://stackoverflow.com/q/3329641>). However, when the ports do not match, the
peers do not match as well.
This change allows two peers connected to an Asterisk server via TCP (or TLS)
behind a NAT (= same source IP address) to be differentiated via their port as
well.
George Joseph [Thu, 30 Nov 2017 20:38:50 +0000 (13:38 -0700)]
AST-2017-013: chan_skinny: Call pthread_detach when sess threads end
chan_skinny creates a new thread for each new session. In trying
to be a good cleanup citizen, the threads are joinable and the
unload_module function does a pthread_cancel() and a pthread_join()
on any sessions that are active at that time. This has an
unintended side effect though. Since you can call pthread_join on a
thread that's already terminated, pthreads keeps the thread's
storage around until you explicitly call pthread_join (or
pthread_detach()). Since only the module_unload function was
calling pthread_join, and even then only on the ones active at the
tme, the storage for every thread/session ever created sticks
around until asterisk exits.
* A thread can detach itself so the session_destroy() function
now calls pthread_detach() just before it frees the session
memory allocation. The module_unload function still takes care
of the ones that are still active should the module be unloaded.
Alexander Traud [Sun, 26 Nov 2017 17:47:17 +0000 (18:47 +0100)]
translate: Transcode siren14, speex32, silk24, and silk12 via slin16.
When a format has no pre-recorded sound files, Asterisk has to transcode between
formats. For this, Asterisk has a fixed translation table. If the pre-recorded
sound files are not available in the same sample rate, Asterisk has not only to
transcode but also to resample.
Asterisk has pre-recorded files for SLN (8000 kHz) and SLN16 (16000 kHz).
However before this change, Asterisk did not take the sample rate into account,
because the translation paths to SLN and SLN16 got the same score/weight in the
table. Consequently, you might have got narrow-band audio with siren14, speex32,
silk24, and silk12 although those are (ultra) wide-band audio codecs.
With this change, the distance in sample-rates is taken into account. Now on the
Command-Line interface (CLI) 'core show channels', you should see:
(slin@16000)->(slin@32000)->(speex@32000).
Corey Farrell [Thu, 23 Nov 2017 00:37:44 +0000 (19:37 -0500)]
Add defaultbranch to .gitreview.
This will cause `git review` to post changes to the branch it is based
on instead of always using master. The defaultbranch setting should be
updated when new major branches are created.
Kevin Harwell [Wed, 22 Nov 2017 16:42:48 +0000 (10:42 -0600)]
pjsip: 183 without To tag does not negotiate media
If a 183 with sdp response is receive without a To tag the sdp is not
negotiated. According to RFC 3261 section 12.1.2 while a To tag is required,
the client needs to still be able to handle the missing tag case for
backwards compatibility.
This patch, accepted by and applied to pjproject, makes it so if an incoming
180/183 with SDP comes in without a To tag it gets appropriately handled.
Alexander Traud [Tue, 21 Nov 2017 12:39:34 +0000 (13:39 +0100)]
res_rtp_asterisk: ICE server-reflexive candidates (srflx) with Dual-Stack.
Previously, Asterisk sent srflx only when configured exclusively for IPv4. Now,
srflx is gathered and sent via SDP, even when Asterisk is enabled for
Dual Stack (IPv4+IPv6) and an IPv4 interface is available/used.
Corey Farrell [Mon, 20 Nov 2017 19:05:46 +0000 (14:05 -0500)]
res_parking: Make load_pri explicit.
res_parking has an implicit load_pri of 0 meaining it's one of the very
first modules loaded after modules with global symbols. Set it
explicitly in the AST_MODULE_INFO block.
Corey Farrell [Tue, 21 Nov 2017 15:16:24 +0000 (10:16 -0500)]
README: Convert to README.md.
Convert the README file to markdown format, remove the old README. This
causes websites like github to display the README in a much nicer
format with live links. The raw file is still very readable from
plain text editors and terminals.
Corey Farrell [Tue, 7 Nov 2017 21:34:40 +0000 (16:34 -0500)]
CLI: Refactor cli_complete.
* Stop using "_COMMAND NUMMATCHES" on remote consoles. Using this
command had doubled the amount of work needed from the Asterisk
daemon for each completion request.
* Fix code formatting.
* Remove static buffer used to send the command, use the same buffer
that will receive the results.
* Move sort from ast_cli_display_match_list.
Corey Farrell [Tue, 7 Nov 2017 20:47:51 +0000 (15:47 -0500)]
CLI: Refactor ast_cli_display_match_list.
* Stop estimating line count, just print until we run out of matches.
* Stop freeing entries, the caller does that anyways.
* Stop calculating / returning numoutput, it was ignored.
Corey Farrell [Thu, 9 Nov 2017 05:42:10 +0000 (00:42 -0500)]
CLI: Create ast_cli_completion_add function.
Some completion generators are very inefficent due to the way CLI
requests matches one at a time. ast_cli_completion_add can be called
multiple times during one invokation of a CLI generator to add all
results without having to reinitialize the search state for each match.
Corey Farrell [Thu, 9 Nov 2017 06:39:13 +0000 (01:39 -0500)]
CLI: Remove calls to ast_cli_generator.
The ability to add to localized storage cannot be supported by
ast_cli_generator. The only calls to ast_cli_generator should be by
functions that need to proxy the CLI generator, for example 'cli check
permissions' or 'core show help'.
* ast_cli_generatornummatches now retrieves the vector of matches and
reports the number of elements (not including 'best' match).
* test_substitution retrieves and iterates the vector.
res_mwi_external_ami specified AST_MODFLAG_LOAD_ORDER but didn't set
load_pri, resulting in an actual load priority of 0. This module only
provides AMI actions so it has no reason to load early.
Corey Farrell [Sun, 19 Nov 2017 19:52:59 +0000 (14:52 -0500)]
Build: Fix issues building without SSL.
* Fix conditional in libasteriskssl.
* Use variables produced by configure to link the SSL and uuid libraries
into libasteriskpj.so instead of hard-coding them.
Corey Farrell [Sun, 19 Nov 2017 19:28:41 +0000 (14:28 -0500)]
res_pjsip: Fix warning by deferring implicit type cast.
Mac doesn't like the comparison of -1 to an enum, so store the result of
ast_sip_str_to_dtmf to an int so we can check for the negative return
value. ast_sip_str_to_dtmf returns an int so this is only delaying the
implicit type cast.
Corey Farrell [Sun, 19 Nov 2017 15:57:28 +0000 (10:57 -0500)]
Build: Fix OSX build issues.
OSX does not support 'readlink -f' or 'sed -r'. Replace readlink with
the GNU make macro 'realpath'. Replace sed with grep in one place, cut
in the other.
Pirmin Walthert [Thu, 16 Nov 2017 08:47:39 +0000 (09:47 +0100)]
res_rtp_asterisk.c: Fix rtp source address learning for broken clients
Some clients do not send rtp packets every ptime ms. This can lead to
situations in which the rtp source learning algorithm will never learn
the address of the client. This has been discovered on a Mac mini with
a pjsip based softphone after updating to Sierra: as soon as USB
headsets are involved, the softphone will send the second packet 30ms
after the first, the third 30ms after the second and the fourth 1ms
after the third. So in the old implmentation the rtp source learning
algorithm was repeatedly reset on the fourth packet.
The patch changes the algorithm in a way that doesn't take the arrival
time between two consecutive packets into account but the time between
the first and the last packet of a learning sequence.
The patch also fixes a second problem: when a user was using a wrong
value for the probation setting there was a LOG_WARNING output stating
that the value had been set to the default value instead. However
the code for setting the value back to defaults was missing.
menuselect detects compiler support for multiple styles of weak
functions. This is a remnant from 2013 when OPTIONAL_API required weak
functions. It is no longer correct for menuselect to switch
dependencies from optional to required based on lack of weak function
support.
Note an issue remains - dependencies should switch from optional to
required based on OPTIONAL_API being enabled or disabled. I don't think
this is possible. menuselect needs to know at startup if OPTIONAL_API
is enabled or disabled, so the only way to fix this is to remove
OPTIONAL_API from menuselect and create a configure option. I've left
the code that switches in place but it's preprocessed out.
Additionally removed:
- WEAKREF variable from Asterisk makeopts.in.
- Related disabled code from test_utils.
- Pointless AC_REVISION call from menuselect/configure.ac.
Corey Farrell [Thu, 16 Nov 2017 15:48:36 +0000 (10:48 -0500)]
acl: Fix allocation related issues.
Add checks for allocation errors, cleanup and report failure when they
occur.
* ast_duplicate_acl_list: Replace log warnings with errors, add missing
line-feed.
* ast_append_acl: Add missing line-feed to logger message.
* ast_append_ha: Avoid ast_strdupa in loop by moving debug message to
separate function.
* ast_ha_join: Use two separate calls to ast_str_append to avoid using
ast_strdupa in a loop.
Corey Farrell [Tue, 7 Nov 2017 23:07:35 +0000 (18:07 -0500)]
CLI: Create ast_cli_completion_vector.
This is a rewrite of ast_cli_completion_matches using a vector to build
the list. The original function calls the vector version, NULL
terminates the vector and extracts the elements array.
One change in behavior the results are now sorted and deduplicated. This
will solve bugs where some duplicate checking was done before the list
was sorted.
Joshua Colp [Thu, 16 Nov 2017 15:04:55 +0000 (15:04 +0000)]
bridge_basic: Ignore answer from transfer target when they've timed out.
This is a fun one.
Given the following attended transfer scenario:
1. Transfer target is called
2. Transferer hangs up
3. Transfer target call attempt reaches timeout
4. Transfer target is told to hang up
5. Transfer target answers before channel is hung up
6. Transferer recall target is called
A crash would occur. This is because the transfer target call
attempt, despite being told to hang up, would raise a recall
target answer before the recall target had been answered. As it
had not answered there would be no recall target channel and it
would implode.
This change makes it so that if the transfer target has been
hung up we don't tell the attended transfer code that it has
answered. We also clear out the stimulus that the recall target
has been answered after telling the transfer target to hang up,
in case it was able to raise the information before we told it
to hangup.
Sean Bright [Thu, 16 Nov 2017 22:18:16 +0000 (17:18 -0500)]
res_pjsip_transport_websocket: Give transport a meaningful description
We were not \0 terminating this string, so any attempt to print it would
in the best case show an empty string and in the worst case potentially
crash.
projects / thirdparty / asterisk.git / log
