This ensures we don't end up calling any of transports' callbacks
with a potentially unreffed user_data (in practice, a client that
may have been removed)
Fixes#97
We cannot hold stream->lock while pushing data, but need
to consistently check the state of the backlog both from
the send_tcp_message function and the on_message_sent function,
which may or may not be called from the same thread.
This commit introduces internal API to allow for potentially
recursive locking of transport streams, addressing a race
condition where the RTSP stream could push items out of order
when popping them from the backlog.
The internal index of our appsinks, while it can be used to
determine whether a message is RTP or RTCP, is not necessarily
the same as the interleaved channel. Let the stream-transport
determine the channel to check backpressure for, the same way
it determines the channel according to whether it is sending
RTP or RTCP.
When removing transports an assertion was that the transports passed in
for removal are present in the list, however that can't be assumed.
As an example if a transport was removed from a thread running
send_tcp_message, the main thread can try to remove the same transport
again if it gets a handle_pause_request. This will not effect the
transport list but it will effect n_tcp_transports as it will be
decrement and then have the wrong value.
The previous implementation stopped sending TCP messages to
all clients when a single one stopped consuming them, which
obviously created problems for shared media.
Instead, we now manage a backlog in stream-transport, and slow
clients are removed once this backlog exceeds a maximum duration,
currently hardcoded.
Fixes#80
This adds new functions for passing buffer lists through the different
layers without breaking API/ABI, and enables the appsink to actually
provide buffer lists.
This should already reduce CPU usage and potentially context switches a
bit by passing a whole buffer list from the appsink instead of
individual buffers. As a next step it would be necessary to
a) Add support for a vector of data for the GstRTSPMessage body
b) Add support for sending multiple messages at once to the
GstRTSPWatch and let it be handled internally
c) Adding API to GOutputStream that works like writev()
Fixes https://gitlab.freedesktop.org/gstreamer/gst-rtsp-server/issues/29
Export rtsp-server library API in headers when we're building the
library itself, otherwise import the API from the headers.
This fixes linker warnings on Windows when building with MSVC.
Fix up some missing config.h includes when building the lib which
is needed to get the export api define from config.h
https://bugzilla.gnome.org/show_bug.cgi?id=797185
Before, the watch backlog size in GstRTSPClient was changed
dynamically between unlimited and a fixed size, trying to avoid both
unlimited memory usage and deadlocks while waiting for place in the
queue. (Some of the deadlocks were described in a long comment in
handle_request().)
In the previous commit, we changed to a fixed backlog size of 100.
This is possible, because we now handle RTP/RTCP data messages differently
from RTSP request/response messages.
The data messages are messages tunneled over TCP. We allow at most one
queued data message per stream in GstRTSPClient at a time, and
successfully sent data messages are acked by sending a "message-sent"
callback from the GstStreamTransport. Until that ack comes, the
GstRTSPStream does not call pull_sample() on its appsink, and
therefore the streaming thread in the pipeline will not be blocked
inside GstRTSPClient, waiting for a place in the queue.
pull_sample() is called when we have both an ack and a "new-sample"
signal from the appsink. Then, we know there is a buffer to write.
RTSP request/response messages are not acked in the same way as data
messages. The rest of the 100 places in the queue are used for
them. If the queue becomes full of request/response messages, we
return an error and close the connection to the client.
Change-Id: I275310bc90a219ceb2473c098261acc78be84c97
As long as gst-rtsp-server can successfully send RTP/RTCP data to
clients then the client must be reading. This change makes the server
timeout the connection if the client stops reading.
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=736647
Make a method to handle the data received on a channel. It sends the
data to the stream of the transport on the RTP or RTCP pads based on
the channel number.
Ignore streams that can't generate RTP-Info instead of failing.
Don't return the empty string when all streams are unconfigured but
return NULL so that we don't generate and empty RTP-Info header.
Improve docs a little.
Add method to create the RTPInfo for a stream-transport.
Add method to create the RTPInfo for all stream-transports in a
session-media.
Use the session-media RTPInfo code in client. This allows us to refactor
another method to link the TCP callbacks.
Use the address managed by the stream for multicast. This allows us to have 1
multicast address for each stream.
Because the address is now managed by the stream we don't have to pass it around
anymore.
Set the address pool on the streams.
Return a boxed GstRTSPAddress from the GstRTSPAddressPool. This allows us to
store more info in the structure and allows us to more easily return the address
to the right pool when no longer needed.
Pass the address to the StreamTransport so that we can return it to the pool
when the stream transport is freed or changed.
Make GObjects from the remaining simple structures.
Remove GstRTSPSessionStream, it's not needed.
Rename GstRTSPMediaStream -> GstRTSPStream: It is shorter
Rename GstRTSPMediaTrans -> GstRTSPStreamTransport: It describes how
a GstRTSPStream should be transported to a client.
Rename GstRTSPMediaFactory::get_element -> create_element because that
more accurately describes what it does.
Make nice methods instead of poking in the structures.
Move some methods inside the relevant object source code.
Use GPtrArray to store objects instead of plain arrays, it is more
natural and allows us to more easily clean up.
Move the allocation of udp ports to the Stream object. The Stream object
contains the elements needed to stream the media to a client.
Improve the prepare and unprepare methods. Unprepare should now undo
everything prepare did. Improve also async unprepare when doing EOS on
shutdown. Make sure we always unprepare correctly.
