Commit graph

468 commits

Author SHA1 Message Date
Olivier Crête
27a057962c address-pool: Verify that multicast addresses are used for multicast and vice-versa 2013-03-11 11:07:20 +01:00
Olivier Crête
d06e68abd1 address-pool: Add unicast addresses 2013-03-11 11:07:20 +01:00
Olivier Crête
4c61c6d308 rtsp-server: Limit the number of threads per server instance
If we exceed the maximum, just round robin the clients over the existing
threads.
2013-03-11 11:07:20 +01:00
Olivier Crête
4071e1b999 rtsp-server: No need to store the GMainContext in the client context 2013-03-11 11:07:20 +01:00
Olivier Crête
b9d111372e Document locking and its order 2013-03-11 11:07:19 +01:00
Olivier Crête
f0ab7ce1bf docs: Generate docs for GstRTSPAddressPool 2013-03-11 11:07:19 +01:00
Olivier Crête
773c48e22f client: Check client provided addresses against the address pool 2013-03-11 11:07:19 +01:00
Olivier Crête
cda75709bb address-pool: Add API to request a specific address from the pool
Also add relevant unit tests.
2013-03-11 11:07:19 +01:00
Olivier Crête
456f4367e3 address-pool: Fix off by one error
When splitting a port range, the port after a skip is not part of range.
2013-03-11 11:07:19 +01:00
Wim Taymans
6db0dbc76c client: make sure the watch exists while sending data
Protect the send_func with a lock. This allows us to wait for sending
to complete before changing the send_func and user_data. We add an
extra ref to the watch to make sure that it remains valid during
sending.
When closing the connection, set the send_func to NULL

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=692433
2013-01-28 11:11:46 +01:00
Wim Taymans
4100b20b0a rtsp-client: set the client backlog
Set the client backlog to a reasonable default
2012-12-14 11:58:29 +01:00
Ognyan Tonchev
0844e8afbc rtsp-media: Make the element a constructor parameter
https://bugzilla.gnome.org/show_bug.cgi?id=689594
2012-12-10 10:25:57 +01:00
Wim Taymans
6beabf1ed4 media: match prepare with unprepare
Really unprepare when there were an equal amount of prepare calls.
2012-11-30 15:03:15 +01:00
Wim Taymans
ca26588c7e media: media has to be unprepared in finalize
Because unprepare takes away the last ref on the media.
2012-11-30 14:58:46 +01:00
Wim Taymans
38addd7822 Revert "client: never call gst_rtsp_media_unprepare, let gst_rtsp_media_finalize do it"
This reverts commit ba5b78ff2f.

We can't use the refcount to trigger unprepare because it is the unprepare call
that removes the last refcount after all messages are consumed. What we should
probably do is make a prepared refcount and only unprepare when the refcount
reaches 0.
2012-11-30 14:36:30 +01:00
Wim Taymans
119674a828 media: let the source unref the last media ref
the last ref to the media is held by the source so we don't need to add more ref
and unrefs, we simply destroy the media when the source is gone.
2012-11-30 13:35:05 +01:00
Wim Taymans
339ea9b085 media: improve debug 2012-11-30 12:54:10 +01:00
Wim Taymans
241baba20a media: check state
Make sure we are in the right state when collecting the position and duration.
Only make ourselves PREPARED when we were previously PREPARING.
2012-11-30 12:53:02 +01:00
Wim Taymans
edf2ef4f0b media: use g_object_ref/unref for GObjects 2012-11-30 10:05:48 +01:00
Alessandro Decina
ba5b78ff2f client: never call gst_rtsp_media_unprepare, let gst_rtsp_media_finalize do it
Calling gst_rtsp_media_unprepare breaks shared medias. Just unref
GstRTSPMedia instances and let gst_rtsp_media_finalize unprepare when a media
isn't being used anymore.
2012-11-30 07:06:17 +01:00
Alessandro Decina
00d9a94e1a Fix compiler warning 2012-11-30 06:17:46 +01:00
Alessandro Decina
e2a7690cb3 Add missing g_type_class_add_private in GstRTSPMediaFactoryURI 2012-11-30 06:14:49 +01:00
Wim Taymans
1abc9be682 small cleanup 2012-11-29 17:21:12 +01:00
Wim Taymans
28fd887547 media: avoid element leak 2012-11-29 17:20:56 +01:00
Wim Taymans
4eb010824e media: require an element in media constructor 2012-11-29 17:20:26 +01:00
Wim Taymans
865c9a6b30 Revert "client: TEARDOWN brings that state to Init again"
This reverts commit 4b61fdad85.

The object is already disposed, there is no point in setting the state.
2012-11-29 17:07:30 +01:00
Wim Taymans
4b61fdad85 client: TEARDOWN brings that state to Init again 2012-11-29 12:30:20 +01:00
Wim Taymans
ad00c5e792 rtsp: make object details private
Make all object details private
Add methods to access private bits
2012-11-29 11:11:05 +01:00
Wim Taymans
e11287eb7c media: check if prepared for some methods
Check that the media object is prepared before doing seek and getting the
current position etc.
Add some g_return checks.
2012-11-28 14:45:30 +01:00
Wim Taymans
d3d74ab77b stream: improve debug 2012-11-28 12:40:18 +01:00
Wim Taymans
fe71114a7d media: unref pipeline in finalize to avoid leaking it 2012-11-28 12:39:37 +01:00
Wim Taymans
d43a31055e rtsp: use gst_object_unref on GstObjects 2012-11-28 12:10:47 +01:00
Wim Taymans
6b36241816 media-factory: require an url 2012-11-28 12:10:14 +01:00
Wim Taymans
20f09bf3e7 server: remove unused include 2012-11-28 11:17:27 +01:00
Wim Taymans
e5ba372808 client: fix factory leak
Keep the factory in the state object only for authorization checks and make
sure we unref it on failure. Also don't keep invalid objects in the state
object.
2012-11-28 11:05:08 +01:00
Wim Taymans
b4c168c71d mounts: add g_return_if guards 2012-11-28 10:40:14 +01:00
Wim Taymans
b3fe3357ab client: improve debug 2012-11-27 12:33:02 +01:00
Wim Taymans
d5389c940d client: improve debug and fix leaks
Cleanup the uri and session when there is a bad request.
2012-11-27 12:24:21 +01:00
Wim Taymans
a26e9b621e client: use 454 when session can't be found
We should use 454 when a session can't be found because there was no session
pool configured in the server. This is not a server configuration problem
because the server on which the request is done might not be the same one that
will keep the sessions for us and so it does not need to support sessions.
2012-11-27 12:11:41 +01:00
Wim Taymans
4782d08bdc client: only free connection when there is one
It's possible that the client doesn't have a connection when we try to free it.
2012-11-27 11:17:45 +01:00
Wim Taymans
18bb9ffa6b client: small cleanup 2012-11-26 17:35:51 +01:00
Wim Taymans
fc0f176a17 client: remove unused include 2012-11-26 17:34:35 +01:00
Wim Taymans
9f8e8bc02d client: fix compilation 2012-11-26 17:34:24 +01:00
Wim Taymans
eb88fa9e76 client: call destroy without the lock 2012-11-26 17:28:29 +01:00
Wim Taymans
33da3af265 client: make the client usable without a socket
Make a method to let the client handle a message and a callback when the client
wants us to send a response message back. This makes it possible to also use the
client object without the sockets, which should make it easier to test.
2012-11-26 17:20:39 +01:00
Wim Taymans
26a4b98ab0 client: small cleanup 2012-11-26 16:45:04 +01:00
Wim Taymans
8da4171055 client: remove reference to server
We don't need to keep a ref to the server
2012-11-26 16:39:26 +01:00
Wim Taymans
4fa7502fd9 client: add locking
Also add some g_return_if()
2012-11-26 16:31:43 +01:00
Wim Taymans
b21b46ec4d client: log more errors 2012-11-26 13:37:20 +01:00
Wim Taymans
f460e7360e client: fix compilation 2012-11-26 13:36:19 +01:00
Wim Taymans
84e72262d0 client: add generic close-after-send support
Add a property to send_response() to close the connection after the response has
been sent to the client.
2012-11-26 13:19:06 +01:00
Wim Taymans
1d53c46d23 MediaMapping -> MountPoints
Describes better what the object manages.
2012-11-26 12:37:55 +01:00
Wim Taymans
0f93879b2c media: fix seeking 2012-11-21 17:21:28 +01:00
Wim Taymans
5eb5fd45f3 media: support more Range formats
Use the new -base methods to convert the Range string into a seek start and stop
value.
2012-11-21 16:41:56 +01:00
Wim Taymans
37a7ec8033 factory: keep ref to factory while media active
While the media from a factory is alive, keep a ref to the factory.

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=663555
2012-11-20 12:29:55 +01:00
Wim Taymans
8fcdca987d factory-uri: add some debug 2012-11-20 12:29:26 +01:00
Wim Taymans
1826844ee4 stream: set udp sources to PLAYING
Set the UDP sources to PLAYING and locked state before we add it to the pipeline
so that it doesn't cause our pipeline to produce ASYNC-DONE.
2012-11-20 12:24:13 +01:00
Wim Taymans
8211cdfdc2 factory-uri: take ref to factory
Take a ref to the factory that we place in our list.
2012-11-20 12:10:16 +01:00
David Svensson Fors
0eeb4a5c73 server: start and stop multiple times
Stop listening on the RTSP port when the GSource is removed, so clients
can't connect and the server can be started again.

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=688395
2012-11-20 11:30:37 +01:00
Wim Taymans
8a7197f078 server: fix small leak 2012-11-20 11:24:35 +01:00
Wim Taymans
989f004e24 media: unref source in finish_unprepare
The source is created in prepare, unref it in finish_unprepare.

See https://bugzilla.gnome.org/show_bug.cgi?id=688707
2012-11-20 09:46:40 +01:00
David Svensson Fors
01973c924d rtsp-media: remove bus watch before finalizing
* A GDestroyNotify function is set for the bus watch in gst_rtsp_media_prepare.
* An extra media ref is added for the bus watch. This extra ref is unreffed by
the GDestroyNotify function.
* gst_rtsp_media_unprepare destroys the source so the bus watch is removed.
* GstRTSPClient, which calls gst_rtsp_media_prepare, also calls
gst_rtsp_media_unprepare before unreffing the media.

This way, the bus watch will be removed before the media is finalized.

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=688707
2012-11-20 09:46:00 +01:00
Alessandro Decina
65042a9551 client: wait until the TEARDOWN response is sent to close the connection
Responses can be sent async so we need to wait until the TEARDOWN response has
been written before we close the connection to the client. This avoids the risk
of writing/polling closed sockets.

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=688535
2012-11-20 09:32:19 +01:00
David Svensson Fors
0996266342 rtsp-stream: plug socket leak
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=688703
2012-11-20 09:26:28 +01:00
Tim-Philipp Müller
0006ca6d60 rtsp-server: don't use deprecated API 2012-11-17 00:11:27 +00:00
Tim-Philipp Müller
290968eb8c rtsp-client: fix unused-but-set-variable compiler warning
rtsp-client.c:1260:21: error: variable 'protocols' set but not used
2012-11-17 00:03:42 +00:00
Wim Taymans
26ff5fc073 rtsp: cleanups 2012-11-15 17:11:16 +01:00
Wim Taymans
e4ea72ccdf stream: use the address managed by the stream
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.
2012-11-15 16:18:29 +01:00
Wim Taymans
ba21661ce4 rtsp: improve debug 2012-11-15 16:15:20 +01:00
Wim Taymans
c34f5d1c1a media: add signal for new streams
This allows applications to listen for new streams and configure properties on
them, like the address pool.
2012-11-15 15:41:42 +01:00
Wim Taymans
4168a67992 media: configure address pool in new streams 2012-11-15 15:41:19 +01:00
Wim Taymans
44a2855eb3 stream: add methods to deal with address pool
Add methods to get and set the address pool for the stream
Add method to allocate and get the multicast addresses for this stream.
2012-11-15 15:36:21 +01:00
Wim Taymans
1b4ac6e5b0 media: remove MTU property
It is a stream property
2012-11-15 15:32:43 +01:00
Wim Taymans
2160d6dbd3 client: set blocksize only on stream
Set the blocksize only on the current stream.
2012-11-15 15:29:35 +01:00
Wim Taymans
6c2947e68b stream: share src and sink sockets
the allocated socket is in the used-socket property, not socket.
2012-11-15 13:52:07 +01:00
Wim Taymans
45b6693b39 rtsp: make address-pool return an address object
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.
2012-11-15 13:25:14 +01:00
Wim Taymans
f15ffb521c rtsp: use AddressPool
Remove the multicast_group property.
Use the configured addresspool to allocate multicast addresses.
2012-11-14 17:23:59 +01:00
Wim Taymans
d0ffc8e679 address-pool: add clear method 2012-11-14 16:20:36 +01:00
Wim Taymans
6085b1fcc1 address-pool: small cleanups 2012-11-14 16:10:45 +01:00
Wim Taymans
b30202b174 address-pool: add object to manage multicast addresses
Make an object that can manage a rage of multicast addresses and ports.
2012-11-14 15:49:06 +01:00
Wim Taymans
7d6e4606fa server: set default max-threads property 2012-11-13 12:05:42 +01:00
Wim Taymans
dfe3efef74 media: wait for concurrent _prepare
If a prepare is busy, wait for the result.
2012-11-13 11:54:17 +01:00
Wim Taymans
47127bd270 media: add lock around message handler
We don't want to dispatch messages while we are still processing the result of
the state change.
2012-11-13 11:49:08 +01:00
Wim Taymans
9a97de88ea media: add lock to protect state changes 2012-11-13 11:15:35 +01:00
Wim Taymans
4753588b09 stream: add locking 2012-11-13 11:14:49 +01:00
Wim Taymans
c7d20e5603 stream-transport: add keep-alive method 2012-11-12 17:11:18 +01:00
Wim Taymans
75473fc88d stream-transport: add method to handle RTP/RTCP
Call new methods instead of poking into the structures directly.
2012-11-12 17:06:42 +01:00
Wim Taymans
883cf794e4 session-media: add locking 2012-11-12 16:51:03 +01:00
Wim Taymans
11cf3f3ccb session: add locking 2012-11-12 16:42:37 +01:00
Wim Taymans
65fa516677 server: free old socket 2012-11-12 16:30:16 +01:00
Wim Taymans
04881bd632 mapping: add locking 2012-11-12 16:18:57 +01:00
Wim Taymans
b8cba7719c media-factory: add locking 2012-11-12 16:14:19 +01:00
Wim Taymans
e61c84c9bb auth: add locking 2012-11-12 16:03:21 +01:00
Wim Taymans
06cadebe71 server: add max-thread property 2012-11-12 15:53:28 +01:00
Wim Taymans
8523c9ca92 server: use a threadpool for the mainloops 2012-11-12 15:29:39 +01:00
Wim Taymans
c431592976 client: rename method
gst_rtsp_client_create_from_socket -> gst_rtsp_client_use_socket: we
don't really create the client from the socket, we use the socket for the
client.
2012-11-12 15:01:13 +01:00
Wim Taymans
a58d404e1f server: rework maincontext handling in clients
Make a separate method to attach a client to a MainContext.

Let the server decide in what GMainContext the client will operate and give this
context to the client in attach. Then the server can later decide to use a
separate thread for each client or just use the mainthread.
2012-11-12 15:01:09 +01:00
Wim Taymans
5b4340067a session: move session header code in session object 2012-11-12 12:40:34 +01:00
Tim-Philipp Müller
4dba434f16 Fix FSF address 2012-11-04 00:14:25 +00:00
Sebastian Pölsterl
75598337a9 rtsp-server: added annotations to indicate type of ownership transfer of return values
https://bugzilla.gnome.org/show_bug.cgi?id=680777
2012-10-28 15:39:04 +00:00