Commit graph

106 commits

Author SHA1 Message Date
Sebastian Dröge
2c405da921 rtpmanager: Refactor RTCP packet loops to fix control flow
Mixing C loops with switch statements is a bad idea as break has a
different meaning in both. Breaking inside the switch statements wrongly
caused further loop iterations.

Instead use goto to get out of the loop and continue to do another loop
iteration, and never ever use break except for the end of a case.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2336>
2022-04-29 23:13:15 +03:00
Seungha Yang
6619f1611f rtpjitterbuffer: Initialize variables
Avoid use of uninitialized variable
Fixing MSVC warning
gstrtpjitterbuffer.c(4733) : warning C4700: uninitialized local variable 'have_sdes' used

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2315>
2022-04-28 12:37:13 +00:00
Sebastian Dröge
9d5179ad3f rtpjitterbuffer: add the reference timestamp meta in more situations
Previously, we only added it when actually performing synchronization
based on the NTP time.

The information can be useful downstream in other situations too, and
we can compute a NTP time as soon as we get a sender report with the
relevant information.

Co-authored-by: Mathieu Duponchelle <mathieu@centricular.com>
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2252>
2022-04-27 12:35:21 +00:00
Havard Graff
390ec99f1b rtptwcc: don't map the buffer twice
...and use the pt extracted rather than the one from RTPPacketInfo
when logging.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2271>
2022-04-26 10:27:25 +00:00
Thibault Saunier
d673a90aea rtpsession: Emit "notify::stats" when we update stats from RR or SR
Sensibily optimizing caching the pspecs and using them directly

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2266>
2022-04-26 08:49:42 +00:00
Havard Graff
b7b71e6974 rtprtxsend: mark RTX buffers with GST_RTP_BUFFER_FLAG_RETRANSMISSION
It is useful for elements downstream from rtxsend to know if the RTP
buffer they are dealing with is an RTX buffer or not.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2272>
2022-04-22 19:27:45 +00:00
Sebastian Dröge
02115a5efc rtpmanager: Move some duplicated constant and helper function to a single place
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2132>
2022-04-20 14:40:25 +00:00
Sebastian Dröge
c7e12974ba rtpbin/rtpjitterbuffer: Don't parse RTCP SRs twice unless needed
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2132>
2022-04-20 14:40:25 +00:00
Sebastian Dröge
82169aa140 rtpjitterbuffer: Add property to throttle handling of RTCP SR / NTP-64 syncing
This proxies the "rtcp-sync-interval" property of rtpbin.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2132>
2022-04-20 14:40:25 +00:00
Sebastian Dröge
ce38614e1a rtpsession: Handle RTCP-SR-REQ (RFC6051) RTCP feedback message
This causes an RTCP SR to be sent at the earliest possible time.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2132>
2022-04-20 14:40:25 +00:00
Sebastian Dröge
0c819d2f31 rtpbin/rtpjitterbuffer: Allow syncing to an SR without CNAME if the CNAME is already known
The RTCP SR packet might be without SDES in case of a reduced-size RTCP
packet. For syncing purposes the CNAME is needed but it might be known
already from an earlier RTCP packet or out of band, via the SDP for
example.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2132>
2022-04-20 14:40:25 +00:00
Sebastian Dröge
cbaac3cdba rtpbin/jitterbuffer: Use inband 64-bit NTP timestamps according to RFC6051 for faster synchronization
When signalled via the caps that the header extension is used, it will
be read and used in the same way as the RTP/NTP time mapping from RTCP
SRs.

If the CNAME of the stream's SSRC is provided out of band via e.g. the
SDP then this allows streams to be synchronized immediately on the first
packet instead of having to wait for the first RTCP SR to arrive.

Fixes https://gitlab.freedesktop.org/gstreamer/gst-plugins-good/-/issues/383

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2132>
2022-04-20 14:40:25 +00:00
Sebastian Dröge
7c796b3c05 rtpsession: Only add send latency to the running time if it is actually known
Otherwise we can't know the running time yet if rtcp-sync-send-time is
set, and have to wait until the latency is known later.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2132>
2022-04-20 14:40:25 +00:00
Sebastian Dröge
7ffc830959 rtpsession: Update 64-bit NTP header extensions with the actual NTP time in senders
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2132>
2022-04-20 14:40:25 +00:00
Sebastian Dröge
8980c35efe rtpmanager: Add header extension implementation for the 64-bit RFC6051 NTP header extension
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2132>
2022-04-20 14:40:25 +00:00
Robert Rosengren
e4a6521ac7 rtpbin: Fix division by zero when using ts-offset-smoothing-factor
avg_ts_offset may cause division by zero when calculating potential
overflow protection. This fix will avoid the division.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2151>
2022-04-11 15:29:49 +02:00
Sebastian Dröge
0813efc821 rtpstats: Remove non-existing twcc field docs from RTPPacketInfo and add missing field docs
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2121>
2022-04-06 10:15:13 +03:00
Sebastian Dröge
46d7763879 rtpsession: Remove unused twcc fields from the struct
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2121>
2022-04-06 10:15:13 +03:00
Matthew Waters
c1a3f958e7 rtpptdemux: fix leak of caps when ignoring a pt
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/2025>
2022-03-25 05:44:36 +00:00
Matthew Waters
206021e4d4 rtpmanager/rtx: implement initial support for reading/writing rid extensions
Two RTP Header extensions are very relevant for rtprtxsend/receive.
1. "urn:ietf:params:rtp-hdrext:sdes:rtp-stream-id": will always be removed
2. "urn:ietf:params:rtp-hdrext:sdes:repaired-rtp-stream-id": will be written
    instead of the "rtp-stream-id" header extension.

Currently it's only a simple replacement of one header extension for
another however a future change would only add the relevant extension
based on some heuristics (like, video frames only on one of the rtp key
frame buffers, or only until the rtx ssrc has been validated by the peer)
in order to reduce the required bandwidth.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1759>
2022-03-21 03:18:18 +00:00
Matthew Waters
1e55e2d654 rtpmanager: add support for RFC8852 (rid) RTP header extensions
Both for regular RID and for adding on a repaired (RTX) etc stream.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1759>
2022-03-21 03:18:18 +00:00
Matthew Waters
ecd9cce3b1 rtpmanager: add support for writing RFC8843 (BUNDLE mid) RTP header extension
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1759>
2022-03-21 03:18:18 +00:00
Sangchul Lee
7691c6776a rtpjitterbuffer: Fix invalid memory access in rtp_jitter_buffer_pop()
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1973>
2022-03-17 12:46:14 +00:00
Tim-Philipp Müller
c29d741c0e rtpbin: proxy new "add-reference-timestamp-meta" property from rtpjitterbuffer
When syncing to an RFC7273 clock this will add the original
reconstructed reference clock timestamp to buffers in form
of a GstReferenceTimestampMeta.

This is useful when we want to process or analyse data based
on the original timestamps untainted by any local adjustments,
for example reconstruct AES67 audio streams with sample accuracy.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1964>
2022-03-16 09:52:58 +00:00
Tim-Philipp Müller
c88bfc0b3e rtpjitterbuffer: add "add-reference-timestamp-meta" property
When syncing to an RFC7273 clock this will add the original
reconstructed reference clock timestamp to buffers in form
of a GstReferenceTimestampMeta.

This is useful when we want to process or analyse data based
on the original timestamps untainted by any local adjustments,
for example reconstruct AES67 audio streams with sample accuracy.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1964>
2022-03-16 09:52:58 +00:00
Sebastian Dröge
5ca39060f4 rtpjitterbuffer: Improve accuracy of RFC7273 clock time calculations
Previously the result of the calculations included inaccuracies caused
by the NTP clock estimation, which caused the timestamps to jitter
+/- 1/clockrate.

By reorganizing the calculations it is possible to get rid of this
inaccuracy and calculate deterministic and exact packet timestamps based
on the actual NTP clock as long as the estimation is not off by more
than 2**31 clockrate units.

The only remaining inaccuracy that is introduced now is caused by the
conversion from the NTP clock to the pipeline clock.

Also split up debug output, demote many messages to the trace debug
level and output more intermediate results.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1955>
2022-03-15 23:33:37 +00:00
Nirbheek Chauhan
8c2ef0f025 twcc: Add some logging to debug TWCC feedback
This should allow people to debug when TWCC feedback is not enabled
because they haven't set the extmap in the caps.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1952>
2022-03-15 22:32:07 +00:00
Nirbheek Chauhan
a6bb63dcd7 twcc: Note that packet-loss-pct can count reordering as loss
This is difficult to encounter in ordinary networks, but is
encountered when using tc-netem to add random delays to packets, and
also when your UDP stream is bonded over multiple links with varying
characteristics.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1952>
2022-03-15 22:32:07 +00:00
Havard Graff
e5bd9839c4 rtprtxsend: don't require clock-rate in caps
For multiplexing, the rtpstreams you are multiplexing might not use
the same clock-rate.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1881>
2022-03-15 19:05:00 +00:00
Havard Graff
4d31641302 rtprtxsend: don't start the task unless we are doing rtx
The rtxsend element can do pass-through when not enabled (no pt-map set)
and in those cases there is no point in starting an additional task
that does absolutely nothing.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1880>
2022-03-15 12:03:27 +00:00
Havard Graff
6f57199958 rtprtxreceive: add ssrc-map property
Mirroring the rtxsend, this allows the application to "pre-map" the
retransmission-ssrcs to the "real" ssrc, if this information is known.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1878>
2022-03-14 09:14:10 +00:00
Tomasz Andrzejak
e74435008f rtpbin: allow FEC elements with Always pads
This patch enable picking up FEC decoder or enocder that have
static repair packets pad.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1860>
2022-03-10 08:33:27 +00:00
Havard Graff
a2c25ccd09 rtprtxsend: if no rtx is present, don't expose a rtx-ssrc in caps
The point here is that rtpsession will create a new rtpsource when
the field "rtx-ssrc" is present, and when not doing rtx, that means
a random ssrc will create a new rtpsource that will be included in RTCP
messages for the current session.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1882>
2022-03-09 15:30:37 +00:00
Havard Graff
2a8fa45ba8 rtprtxsend: don't process or warn if no map is set
This makes it more gentle when doing "pass-through"

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1879>
2022-03-09 12:01:22 +05:30
Mikhail Fludkov
815d279f2e rtprtxreceive: fix crash when RTX payload has zero length
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1875>
2022-03-08 09:07:41 +00:00
Havard Graff
86c7231dae rtprtxreceive: allow passthrough and non-rtp buffers
To avoid mapping rtp buffers when RTX is not in use, and to not
do a full error on receiving a non-rtp buffer, since you have no control
of what a rouge sender might send you.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1874>
2022-03-07 23:43:49 +00:00
Havard Graff
a475c93346 rtprtx: don't access type-system per buffer
When doing only a single stream of audio/video this hardly matters,
but when doing many at the same time, the fact that you have to get
a hold of the glib global type-system lock every time you process a buffer,
means that there is a limit to how many streams you can process in
parallel.

Luckily the fix is very simple, by doing a cast rather than a full
type-check.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1873>
2022-03-07 22:01:03 +00:00
Guillaume Desmottes
8bbdd9addb rtpsource: fix rtp_source_get_nack_deadlines doc
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1775>
2022-02-22 09:40:35 +00:00
Robert Rosengren
265878c4ba rtpbin: Safer ts-offset-smoothing-factor calculation
Protect the ts-offset-smoothing-factor calculation from overflow. Output
warning and fallback to ts-offset if it is detected.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1409>
2022-02-08 11:11:35 +00:00
Robert Rosengren
31dd9226ce rtpbin: add ts-offset-smoothing-factor property
Add property to set the TS offset smoothing factor and set default value
to not use it.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1409>
2022-02-08 11:11:35 +00:00
Danny Smith
bc964141c8 rtpbin: applied smoothing to jittery sender time-stamps
Applying a moving average filter to the timestamp offsets
for smoothing jittery and preventing aggressive skew handling.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1409>
2022-02-08 11:11:34 +00:00
Danny Smith
d5e257afd1 rtpbin: added option for setting min_ts_offset in ntp-sync mode
Constantly updating the ts_offset results in audiable glitches
when streaming audio using ntp-sync=true. By requiring a minimum
offset before updating ts_offset this can be mitigated. Added a
parameter which can be used to set min_ts_offset in ntp-sync mode.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1409>
2022-02-08 11:11:34 +00:00
Mathieu Duponchelle
a09b8ded30 rtpbin: add new request-fec-decoder-full signal for BUNDLE
When multiple streams are bundled together, the application needs
to know about the payload type in order to instantiate the appropriate
FEC decoder.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1429>
2021-12-14 17:34:53 +00:00
Mathieu Duponchelle
4412198c05 rtpfunnel: fix extmap handling on accept-caps
Follow-up on 97d83056b3, only check
for intersection with the current srccaps when checking if a sinkpad
can accept caps.

I must have been lucky in my firefox testing then, and always entered
the code path with audio getting negotiated first, thus not failing
the is_subset check when srccaps had been negotiated as
application/x-rtp, and an accept-caps query was made for the video
caps with a defined extmap.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1384>
2021-11-23 20:26:30 +00:00
Mathieu Duponchelle
97d83056b3 rtpfunnel: don't enforce twcc during upstream negotiation
A previous patch has caused rtpfunnel to output twcc-related
information downstream, however this leaked into upstream
negotiation (through funnel->srccaps), causing payloader to
negotiate twcc caps even when not prompted to do so by the user.

Fix this by only enforcing that upstream sends us application/x-rtp
caps as was the case originally.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1278>
2021-11-12 18:40:32 +00:00
Mathieu Duponchelle
72118b9db4 rtptwcc: complete bufferlist fix
When dealing with bufferlists, we need to store one "SentPacket"
structure per buffer, not one per buffer list!

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1278>
2021-11-12 18:40:32 +00:00
Mathieu Duponchelle
792fb05cec st2022-1-fecdec: fix packet trimming
g_sequence_remove_range's end iter is exclusive, so if one
wants to remove that item as well, it should be called with
the next iter.

This could in theory fix an issue where:

* The sequence isn't entirely trimmed, with an old item lingering

* Following FEC packets are immediately discarded because they
  arrived later than corresponding media packets, long enough for
  seqnums to wrap around

* We now try to reconstruct a media packet with a completely obsolete
  FEC packet, chaos ensues.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1341>
2021-11-12 08:15:28 +00:00
Matthew Waters
71dd47516c rtpbin: separate out the two fec decoder locations
The pipeline flow for receiving looks like this:

rtpsession ! rtpssrcdemux ! session_fec_decoder ! rtpjitterbuffer ! \
  rtpptdemux ! stream_fec_decoder ! ...

There are two places where a fec decoder could be placed.
1. As requested from the 'request-fec-decoder' signal: after rtpptdemux
   for each ssrc/pt produced
2. after rtpssrcdemux but before rtpjitterbuffer: added for the
   rtpst2022-1-fecenc/dec elements,

However, there was some cross-contamination of the elements involved and
the request-fec-decoder signal was also being used to request the fec
decoder for the session_fec_decoder which would then be cached and
re-used for subsequent fec decoder requests.  This would cause the same
element to be attempted to be linked to multiple elements in different
places in the pipeline.  This would fail and cause all kinds of havoc
usually resulting in a not-linked error being returned upstream and an
error message being posted by the source.

Fix by not using the request-fec-decoder signal for requesting the
session_fec_decoder and instead solely rely on the added properties for
that case.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1300>
2021-11-10 10:38:26 +00:00
Mathieu Duponchelle
c414f9560a rtptwcc: don't assume all PacketInfo->data are buffers
They can also be buffer lists

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1268>
2021-10-29 03:46:25 +02:00
Matthew Waters
8c35850f23 rtpbin: fix leak of pad when a fec encoder and aux sender a created
The ghost sink pad retrieved by rtpbin from the aux sender was not freed
when there was a previous element (fec encoder) in the chain.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1222>
2021-10-21 13:46:03 +00:00
Olivier Crête
0dbe0e21fe rtphdrext-clientaudiolevel: Rename RFC 6464 element
Multiplying elements named after RFC numbers is confusing,
so let's give them meaningful names.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1125>
2021-10-20 00:03:09 +00:00
Olivier Crête
5ec82c1c4e rtphdrext: Pass just the attributes to the subclass
Since the base class now does the parsing, there is no need
to reproduce that code in all the subclasses, just pass the attributes
which are the only relevant bit anyway.

Also, only store the direction if the subclass accepted the caps

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/906>
2021-09-28 20:04:55 +00:00
Olivier Crête
2db9f42b36 rtphdrext-rfc6464: Use helper function to set caps field
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/906>
2021-09-28 20:04:55 +00:00
Olivier Crête
ba328fb98d rtphdrext: Set caps without attributes as the default
Most subclasses just use the simple function, so just let the base class
do it. It makes less code in subclasses.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/906>
2021-09-28 20:04:55 +00:00
Olivier Crête
498740fe57 rtphdrext: Put simple caps generation as the base class default
Instead of having a helper function that gets called by almost every
subclass, just let the base class set the caps fields automatically.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/906>
2021-09-28 20:04:55 +00:00
Thibault Saunier
5ff769d731 Move files from gst-plugins-good into the "subprojects/gst-plugins-good/" subdir 2021-09-24 16:13:50 -03:00