This crash could happen at any time a RTP and RTCP buffer arrived
simultaneously in ssrcdemux.
The problem was that sticky-event arriving while the rtp and rtcp pads
were being set up could arrive just too late to be included in the initial
forwarding.
The fix checks if the stickies have been sent on the srcpad about to be
pushed on, and if not sends them. It also blocks any stickes from
being forwarded *prior* to this happening, to avoid them arriving on
the srcpad multiple times.
Since the test loops 1000 times, this will make running under valgrind
take forever, so use the RUNNING_ON_VALGRIND variable to detect we
are running under valgrind, and reduce the loop-count to 2 in that case.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gst-plugins-good/-/merge_requests/992>
When used for processing bundled media streams within rtpbin the rtpssrcdemux element may
receive bad RTP and RTCP packets, these should not be treated as a fatal error.
The property is useful against atacks when the sender changes SSRC for
every RTP packet. The property with the same name introduced in rtpbin
was not enough, because we still can end up with thousands of pads
allocated in rtpssrcdemux.
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.
