Multiple writers may operate on a single WS connection simultaneously.
Make sure they don't get in each other's way while constructing their
messages/responses by holding the lock from the beginning of the
response until the point when it's fully ready. This fixes a problem of
parts of multiple messages getting mixed up with each other.
Change-Id: If84224fc06b423cd65c12981a5b09ee99b121df2
(cherry picked from commit 1480574001)
(cherry picked from commit a40dea5007)
Looks like partial writes are handled automatically internally, but add
a request for a "writeable" callback anyway so we know when we can write
more.
Closes#1943
Change-Id: I86a8e1924febc0524b90dc6559753e12e0de9dfb
(cherry picked from commit 9db8925846)
(cherry picked from commit c5eed15b10)
This makes it possible for the BIO callback to access these values,
instead of obtaining possibly mismatched values from packet_stream.
Restores behaviour prior to 5331506Closes#1996
Change-Id: I74ccb82f1739ca98eb033e68df4e27619ca7e46a
(cherry picked from commit 271b3b1520)
(cherry picked from commit 4411059353)
To release any references early
Change-Id: Ib4417301e4f02fbc7d9f4d77d78cb827bb481726
(cherry picked from commit b85e007ff1)
(cherry picked from commit 5a05467be1)
We already use the `dtls_connection` object as "app data" so let's use
it for the BIO callback arg as well.
Change-Id: I715b6756ea36c7a3cb2bed68fce17108eb480e23
(cherry picked from commit 8297449aa1)
(cherry picked from commit c811bbd96b)
Avoids redundant loop iteration.
Also change log level to INFO to match all other log output.
Change-Id: I5587c59398656dc03fcb346edcbc48f350a4a392
(cherry picked from commit d7d7a98071)
(cherry picked from commit 9b8ed29451)
Fixes regression from Ib4285e7aae
RTCP multiplexing requires the RTCP sender to maybe lock the same output
stream, maybe lock some other one. Allow for both.
Change-Id: I6fcef32e656f8f0de46ad777f11a19c259ce35c7
(cherry picked from commit cec0afa25e)
(cherry picked from commit d1d4110f0f)
(cherry picked from commit ea26245c4d)
With selected_sfd being protected by in_lock, we pretty much have to
hold at least in_lock everywhere, and end up requiring both locks in
many places. The distinction has become pointless.
Change-Id: Ic0ad976c2d68d9639b9434da7f0e6e9c0d84c185
(cherry picked from commit e03f814855)
(cherry picked from commit 222fcaa4b0)
(cherry picked from commit abf1ad73d2)
PTs that were remembered from a previous handshake to save codec options
must be flagged as such so that they're not considered as having been
present in the current offer, so that they can be flagged as transcoding
PTs.
closes#1989
Change-Id: I19c2aff7e83ed338a81be99544645821165304cd
(cherry picked from commit f2a9111fb6)
(cherry picked from commit 77c70d6665)
Reverses the side effect of I225b43dff
Restores consistency and fixes tests and very slow systems
Change-Id: Ib4285e7aae95cd58eccd1d3c4f84e5786b4b0810
(cherry picked from commit ffacde5748)
(cherry picked from commit e5a1aec247)
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)
(cherry picked from commit 2e714ebedd)
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)
(cherry picked from commit b38f4d81ce)
Report stats before monologues are removed from the hash table
Closes#1940
Change-Id: Ie6efad41359450ce788cdcda307cda9a24bb6d6c
NGCP-Flow: mr12.5
(cherry picked from commit 6bbf32e1d8)
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)
(cherry picked from commit 871162b13b)
(cherry picked from commit 97dfc13c9f)
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)
(cherry picked from commit 7fd87d07f7)
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)
(cherry picked from commit 83fa97fd72)
We need separate tracking for A and B sides
Change-Id: I9aeb91beb58e9efc5e8f79f336d120e46c25b8c9
(cherry picked from commit fa3de35448)
(cherry picked from commit ff25eecd32)
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
(cherry picked from commit b742dbc460)
(cherry picked from commit 87f5434415)
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
(cherry picked from commit 65bfe22d40)
(cherry picked from commit 789971d929)
This triggers an initial "no carrier" indication if the sender is slow
to start PCM.
Change-Id: I3513efafae06f992b3c9abd58b8ec797320c3b56
(cherry picked from commit dafb68ef80)
(cherry picked from commit d0c557a7dc)
A call that gets created but then doesn't get initialised would have its
Redis DB left at zero. At destruction it would then try to switch to DB
zero. Fix this by using an appropriate initial value.
Closes#1905
Change-Id: I852e48c5a06b732b37d2ccd5c478de4760aacd4e
(cherry picked from commit add46f265a)
(cherry picked from commit 127540cefa)
These are potentially computed from inside each subdir, and in addition
due to what appears to be a regression in GNU make 4.4, where it is
reevaluating variables that contain $(shell) functions, many times (in
the order of thousands, this was slowing down the build, were on the
Debian amd64 build daemons it went from 5m with GNU make 4.3 to 2h40m
with GNU make 4.4. Although the bulk of the slow down has been fixed
with previous commits, the remaining optimizations are only to avoid
this potentially happening again in the future, and to reduce useless
duplicate work.
Instead of trying to cache the values from within make itself, where
programming this there is extremely painful, and does not seem to be
able to greatly reduce the number of calls, because the build system
is going to be called multiple times for different targets. Simply
externalize the generation into several shell scripts, that we call
to generate a make fragment that then we include from the various
Makefiles.
For a Debian build with GNU make 4.3, this reduces the amount of total
pkg-config calls from around ~1600 to 128, for dpkg-buildflags from
~1100 down to 6, and for dpkg-parsechangelog from ~56 to 17, but the
slow down is not as significant there anyway.
For a Debian build with GNU make 4.4, this reduces the amount of total
pkg-config calls from around ~2600 to 128, for dpkg-buildflags from
~2800 down to 6, and for dpkg-parsechangelog from ~350 to 21.
For a Debian build with GNU make 4.4, this reduces the build time
on this system from 2m10s to ~ 1m30s.
Change-Id: I427d0ea5106dc6ed1ff9e664ccdba2fa0725b7d0
(cherry picked from commit 7dbe24e4e5)
(cherry picked from commit 1e1075bf38)
Starting with GNU make 4.4, build time have massively regressed
where before they would take 5m on amd64 now can take 2h40m. While this
seems clearly broken, the release notes are filled with notices for
breaking changes, and in particular the one for passing all make
variables down to the invoked programs executed via the «shell» GNU make
function, so it is not clear what is expected breakage and what is not.
This has been reported in Debian, but not yet upstream, and while it
seems like a clear regression, it's not clear what will be the upstream
take on it. For now apply workarounds that do not change semantics, and
which do not regress with older GNU make versions.
Use the GNU make «origin» function instead of «?=» which defaults to
defining a variable as a recursive one. Coerce already defined variables
into simple ones to avoid GNU make re-evaluating these variables for
each «shell» function invocation.
Ref: https://bugs.debian.org/1092051
Change-Id: I076fc05dd616918473a22e7e942fecfdc9851d47
(cherry picked from commit 887fb40f3f)
(cherry picked from commit 67a2b222c7)
Make sure we properly return a failure if no supported payload type is
present.
Change-Id: Ia483e0819b2d8ca0c2c5184c929dfe3d05c96ca1
(cherry picked from commit 5f6609de5d)
(cherry picked from commit 357c4f1e4d)
Make sure a codec is not only known to us, but that it can actually be
used, in places where it makes sense. This is partially redundant
because ensure_codec_def_type already takes care of this, but a codec
definition may come from a different source, so it doesn't help to
double check.
Change-Id: I91af84afc2477840f1400674b2538ad8fb7746ee
(cherry picked from commit 7b60e85970)
(cherry picked from commit ea3cfdbd38)
If the receiver of a previously passthrough T.30 stream gets switched to
a generated PCM stream from a T.38 gateway, continue the sequencing of
the previous SSRC. Technically this is not necessary as the generated
PCM stream gets a new SSRC, but at least Asterisk seems to ignore this
and expect sequencing to continue, and will ignore PCM if the sequence
jump is too large.
Change-Id: Ia4656770db11f5fa1a1e9bf5bd71a0398deb1e00
(cherry picked from commit 18579227a9)
(cherry picked from commit bca98ab764)
Add a simple function that acts as an encryption callback to just update
the ext_seq (index) of the egress SSRC context. The kernel module
already does this, but the daemon only did it when SRTP was involved.
This now tracks egress packet indexes in all cases.
Change-Id: I9460744de55ead4b05aceb322fd8482442ff2b41
(cherry picked from commit 0da49000ba)
(cherry picked from commit 3e8a35511b)
Richard Fuchs
Alex Brown
Orgad Shaneh
Michal Popovič
Dennis Yurasov
Guillem Jover