Reverses the side effect of I225b43dff
Restores consistency and fixes tests and very slow systems
Change-Id: Ib4285e7aae95cd58eccd1d3c4f84e5786b4b0810
(cherry picked from commit ffacde5748)
Protect selected_sfd with in_lock.
Protect RTCP sending with in_lock and out_lock as appropriate.
Has the odd side effect of RTCP reports expected in tests to be sent one
packet later than before.
Closes#1966
Probably fixes#1927
Change-Id: I225b43dff8e8fbb938d3be6aad50249997615d77
(cherry picked from commit ffc539c0d8)
Using `l->next` as a test for whether only one element is left in the
list is not reliable. Use the actual list length instead.
Update one affected test.
Closes#1961
Change-Id: I5773715700220cd762e61090bac941ddd33afd9b
(cherry picked from commit 02897ccb46)
SDP attributes have been moved to the call's memory arena in 5115fe000.
Make sure the redis restore code uses the same.
Remove some redundant null checks as well.
Closes#1959
Change-Id: I2f37869f5f2c95999d028c811c6a0eb10215f576
(cherry picked from commit a7a43c9b75)
If DTX audio has already been produced for a particular timestamp and a
late packet matching that timestamp is encountered, we should just
discard the packet instead of immediately doing a timestamp reset. This
prevents unwanted TS resets. Only do a timestamp reset if this happens
multiple times in a row, as we don't want to have to discard too much
audio.
Change-Id: I46c8c20b08787f7e45145bd88463bb6878f36f15
(cherry picked from commit a2fc205505)
Keep outgoing RTP timestamps consistent between different instances of
the DTX buffer. Update affected tests.
Change-Id: I6cf03ab32f5c510bd781063a9e7241d1187c202b
(cherry picked from commit 940f84b23a)
Switch from thread-local allocated string object to a stack allocated
one. Use auto cleanup to manage storage duration and final logging.
Fixes intermittent random failures of the leak checker.
Change-Id: Ie6afb27e6fd1accbe641fc62175d553a0558de0d
(cherry picked from commit 56f171b64e)
These are normally held per thread, but in cases of threads not having
dedicated init/cleanup functions (GLib thread pool), we can use a global
one as fallback.
Fixes#1936
Change-Id: Ia2ff3523e6079baa73e0721862100ec2f8b66c88
(cherry picked from commit 5d613020e6)
Replace with hand-rolled requests made via libcurl.
Background: libxmlrpc-core-c3-dev packaging is currently broken in
Debian Sid and this is a good opportunity to move away from it.
Ref: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1102554
Change-Id: I8a09452220993afdac19654edf13d7f3f6ba64c9
(cherry picked from commit 5d985372d8)
If recording-method is pcap, then proc is zero-initialized, so
meta_filepath is empty. This shows many logs such as:
[core] Failed to open recording metadata file '(null)' for writing: Bad address
Prevent them by returning earlier.
Fixes#1889 for real this time.
Amends commit 759fd72dc6.
Closes#1924
Change-Id: I25c9acefba9ee129354fd799ef493af86b9eab2c
(cherry picked from commit 312997c21d)
If the last used PT was a different one, check if it was an unknown PT.
If it was, then DTX can remain active, as the packet would have been
blocked.
Change-Id: Icc8bf195c30a2252b392bbccfeef34b91078ddd6
The stub is used for unknown payload types. Specialise the handler
function for it with an extra check for DTX usage. This effectively
blocks RTP packets with unknown payload types if DTX has been
configured. Forwarding RTP packets with unknown payload types can
interfere with a clean output stream generated by the DTX buffer.
Change-Id: I6f0aa9654946d2877b963cd13cec7c3f5c8b1c54
Shutting down a DTX buffer means that it may restart at an unknown later
time. Reset TS tracking when shutting it down so that if it does
restart, it receives a fresh TS state, instead of continuing on the
previous timestamps, which may then be wrong.
Change-Id: I9c7978e2f335eab32982c4739917ca5ce0c27eb6
Extend PT tracker to not just track the most used PT, but also which PT
has been used last. Use a simple circular array.
Use this new tracking to determine whether a DTX buffer should remain
active. The "most used PT" approach isn't reliable for this, as it leads
to a DTX buffer restarting only after the PT has become the "most used,"
so only after about 10 packets. Meanwhile a newly appearing SSRC would
have a new DTX buffer active immediately. Being strict about which PT
was used last resolves this inconsistency.
Change-Id: If912e4d10e6737a82f6adc0b19d502075aa18a93
If a packet is received for a codec that should have an active DTX but
doesn't, drop the packet instead of forwarding it without being
processed. In this case we expect that some other codec has an active
DTX buffer, and forwarding the packet would interfere with a clean
output stream.
Change-Id: I162ba6f314b804123a6a3957feb7435a319ad2b9
The GLib slice allocator has been obsoleted. Having a mixture of two
allocators can lead to hidden issues on systems that use different
implementations. Take the leap and replace everything with g_new/g_free.
Change-Id: I025c8383ef47b2c2472573360407fd6a6ca339b3
Media IDs are supposed to be unique. Non-RFC OSRTP may however lead to
duplicated media sections with the same media ID. Use a hash table to
track which media IDs have already been seen, and ignore any duplicates.
Change-Id: I9de5fdf3165fc4326862af708aec3d4f6736cb12
This mode is only supposed to be used by cases when:
- in-dialog offerer doesn't have own MoH capabilities; and
- still want the recipient to hear the MoH music; hence
- an offerer checks whether a recipient is capable of MoH,
and launches a player based on given capabilities;
- the rest functions the same as with usual MoH hold;
Other information:
- `mode=reflect` is only to be used within dialog,
so, it's not meant for a session origination,
and hence stands alone from other MoH option flags
that actually give a metadata (e.g. blob data)
- `mode=reflect` hence contradicts with `mode=sendrecv`
which in its turn serves another purpose
Usage:
- `moh => { mode => 'reflect' }`
Change-Id: I8d0d55f2711c6b47bfca17691582c1ffe66eae3d
Decoding and encoding may have taken some time. Grab an updated "now"
before scheduling a packet to send.
Change-Id: I9380cb01d77519cf082d87f3a6c6cfc18b8213f8
With reuse-codecs, we still need to place the codec from the SDP into
our prefs list, even if it's already present, as the format options may
have changed. Update one affected test case.
Closes#1921
Change-Id: I688c57a8c45ec4c3bf159fe2193a0e00bbceeda2
Protects against double MoH played,
e.g. when inadvertently two rtpengine instances try to trigger MoH.
This gives a clue to the current rtpengine instance,
that another one already started MoH for this call.
By default is set to false.
Change-Id: I7cc36a177b0ce1bdb64ff8b42bf31f13fb1e4c91
Allocate all bufferpool shards of the same size, regardless of
underlying allocator. This way increase memory usage a bit, but we're
already quite heavy on that, so no big deal.
Change-Id: Icbe09cd2f9b33bc58ab1efe7de293dea00236fec
Use allocators that return memory blocks aligned to the requested size,
instead of generic malloc. This makes it easier to play tricks with the
memory blocks.
Change-Id: Iad4b1127c76e48c2e9b4ad8489118d4883a24f6a
Richard Fuchs
Orgad Shaneh
Donat Zenichev
Michal Popovič