If, say, a rtx-packet arrives really late, this can have a dramatic
effect on the jitterbuffer clock-skew logic, having it being reset
and losing track of the current dts-to-pts calculations, directly affecting
the packets that arrive later.
This is demonstrated in the test, where a RTX packet is pushed in really
late, and without this patch the last packet will have its PTS affected
by this, where as a late RTX packet should be redundant information, and
not affect anything.
This patch corrects the delay set on EXPECTED timers that are added when
processing gaps. Previously the delay could be too small so that
'timout + delay' was much less than 'now', causing the following retries
to be scheduled too early. (They were sent earlier than
rtx-retry-timeout after the previous timeout.)
Turns out that the "big-gap"-logic of the jitterbuffer has been horribly
broken.
For people using lost-events, an RTP-stream with a gap in sequencenumbers,
would produce exactly that many lost-events immediately.
So if your sequence-numbers jumped 20000, you would get 20000 lost-events
in your pipeline...
The test that looks after this logic "test_push_big_gap", basically
incremented the DTS of the buffer equal to the gap that was introduced,
so that in fact this would be more of a "large pause" test, than an
actual gap/discontinuity in the sequencenumbers.
Once the test was modified to not increment DTS (buffer arrival time) with
a similar gap, all sorts of crazy started happening, including adding
thousands of timers, and the logic that should have kicked in, the
"handle_big_gap_buffer"-logic, was not called at all, why?
Because the number max_dropout is calculated using the packet-rate, and
the packet-rate logic would, in this particular test, report that
the new packet rate was over 400000 packets per second!!!
I believe the right fix is to don't try and update the packet-rate if
there is any jumps in the sequence-numbers, and only do these calculations
for nice, sequential streams.
gst_splitmux_src_activate_part() configures the pad information
before starting the pad task, but occasionally the changes it makes
to the pad are not seen in the pad task because they're not
protected by the right locking. Use the pad's object lock to
protect those variables.
Fix a deadlock around the pads list by using an RW lock to
allow simultaneous readers. The pad list doesn't really changes
except at startup and shutdown.
Make the debug output less confusing by not mentioning a src
pad when doing calculations on the sink pad side.
Improve debug around why a GOP is considered overflowing a fragment
AAC and various other audio codecs need a couple frames of lead-in to
decode it properly. The parser elements like aacparse take care of it
via gst_base_parse_set_frame_rate, but when inside a container, the
demuxer is doing the seek segment handling and never gives lead-in
data downstream.
Handle this similar to going back to a keyframe with video, in the
same place. Without a lead-in, the start of the segment is silence,
when it shouldn't, which becomes especially evident in NLE use cases.
There used to be some profile/level support in encoders. This code was moved to
GstV4l2Codecs and is now also used for decoders. The caps templates for the
H.264, H.265, MPEG4, VP8 and VP9 encoders and decoders should now reflect the
profiles and levels advertised by the kernel.
sethostent() seems to be using a global state and we endup with leaks from
that API when called through shout_init(). We had the option to only
ignore the shout case, but the impression is that if we have shout and
another sethostend user, as it's a global state, we may endup with a
different stack trace for the same leak. So in the end, we just ignore
memory allocated by sethostent in general.
In this change we now protect the internal srcpads list using the
stream lock and limit usage of the internal stream lock to
preventing data flowing on the other src pad type while creating
and signalling the new pad.
This fixes a deadlock with RTPBin shutdown lock. These two locks would
end up being taken in two different order, which caused a deadlock. More
generally, we should not rely on a streamlock when handling out-of-band
data, so as a side effect, we should not take a stream lock when
iterating internal links.
Allocator resources cannot be freed when a buffer pool is orphaned
while its buffers are in use. They should, however, be freed once those
buffers are no longer needed. This patch disposes of any buffers
belonging to an orphaned pool as they are released, and makes sure
that the allocator is cleaned up when the last buffer is returned.
When trying to orphan a buffer pool, successfully return and unref
the pool when the pool is either successfully stopped or orphaned.
Indicate failure and leave the pool untouched otherwise.
This means we can use some newer features and get rid of some
boilerplate code using the G_DECLARE_* macros.
As discussed on IRC, 2.44 is old enough by now to start depending on it.
It must be accurate for all samples to work in Final Cut properly, so
the best we can do is to assume that all samples are the same as the
first. Bigger samples are truncated, smaller samples are padded.
This takes the timestamp of the earliest stream and offsets it so that
it starts at 0. Some software (VLC, ffmpeg-based) does not properly
handle Matroska files that start at timestamps much bigger than zero.
Closes https://gitlab.freedesktop.org/gstreamer/gst-plugins-good/issues/449
There is only a single sink element in async-finalize mode, and we would
keep the running time from previous fragments set in that case. As we
don't ever set the running time for the very last fragment on EOS, this
would mean that the closing time reported for the very last fragment is
the same as the closing time of the previous fragment.
This is a tiny clarification as the storage was loosely named "storage".
This change clarify that the storage is specificaly used for received RTP
packets. This is unlike the storage found in rtprtxsend that stores a
backlog of sent RTP packets.
We recently added code to remove outdate NACK to avoid using bandwidth
for packet that have no chance of arriving on time. Though, this had a
side effect, which is that it was to get an early RTCP packet with no
feedback into it. This was pretty useless but also had a side effect,
which is that the RTX RTT value would never be updated. So we we stared
having late RTX request due to high RTT, we'd never manage to recover.
This fixes the regression by making sure we keep at least one NACK in
this situation. This is really light on the bandwidth and allow for
quick recover after the RTT have spiked higher then the jitterbuffer
capacity.
The second udpsrc (rtcp) might not have seen the segment event if it was
not enabled or if rtcp is not available on the server. So if the
application tries to send an EOS event it will try to set an invalid
seqnum to the event.
