George Kiagiadakis [Tue, 14 Jan 2014 11:07:58 +0000 (12:07 +0100)]
rtprtxreceive: change the rtx_pt_map directly in set_property() instead of delaying it for chain()
The same lock is held, so there is no point in complicating it...
George Kiagiadakis [Tue, 14 Jan 2014 10:55:00 +0000 (11:55 +0100)]
rtprtxreceive: simplify the code of finalize()
George Kiagiadakis [Tue, 14 Jan 2014 10:52:21 +0000 (11:52 +0100)]
rtprtxreceive: use the GstObject lock instead of a new one
George Kiagiadakis [Tue, 14 Jan 2014 10:45:52 +0000 (11:45 +0100)]
rtprtxsend: use the GstObject lock instead of a new one
Nicolas Dufresne [Tue, 10 Dec 2013 19:29:55 +0000 (14:29 -0500)]
v4l2: Add NV12_64Z32 support
https://bugzilla.gnome.org/show_bug.cgi?id=722127
Justin Joy [Tue, 14 Jan 2014 10:08:49 +0000 (19:08 +0900)]
osshelper: Don't leak fd when getting card name
https://bugzilla.gnome.org/show_bug.cgi?id=722163
Vincent Penquerc'h [Tue, 14 Jan 2014 09:43:33 +0000 (09:43 +0000)]
Revert "aacparse: relax the detection of ADTS"
This was pushed by mistake along with the V4L2 fix.
This reverts commit
8eb4b032bef444397c4d211f2095c173ba114187.
Justin Joy [Tue, 14 Jan 2014 06:42:01 +0000 (15:42 +0900)]
rtpg726pay: don't leak encoding_name string
https://bugzilla.gnome.org/show_bug.cgi?id=722159
Vincent Penquerc'h [Mon, 13 Jan 2014 09:14:00 +0000 (09:14 +0000)]
v4l2: fix build break using V4L2_CAP_VIDEO_M2M_MPLANE
This may not be defined. Since the previous version used
only the other define (V4L2_CAP_VIDEO_OUTPUT_MPLANE), fall
back on this only when not available.
Akihiro Tsukada [Tue, 26 Feb 2013 16:45:52 +0000 (01:45 +0900)]
aacparse: relax the detection of ADTS
According to ISO/IEC 13818-7, "channel_config" field in ADTS header
may have value of 0, as in the case of frame with PCE.
gst_aac_parse_detect_streams() returned FALSE for those frames
and discarded them.
Julien Isorce [Tue, 7 Jan 2014 11:58:23 +0000 (11:58 +0000)]
v4l2bufferpool: check set_config return value in gst_v4l2_buffer_pool_new
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Fri, 10 Jan 2014 17:40:31 +0000 (12:40 -0500)]
v4l2object: Add parsed=1 field for encoded output
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Fri, 10 Jan 2014 17:39:16 +0000 (12:39 -0500)]
v4l2object: Don't leak empty caps
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Julien Isorce [Wed, 8 Jan 2014 16:51:21 +0000 (16:51 +0000)]
v4l2bufferpool: do not stop a stream not previously started
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Thu, 12 Dec 2013 21:27:21 +0000 (16:27 -0500)]
v4l2object: Don't enforce dimension field on encoded formats
Don't enforce having width, height and framerate in template caps for encoded
formats. These don't always need to be exposed and may break negotiation for
decoder and decoding sink. If needed, these field will be automatically added
when probed caps are known.
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Julien Isorce [Thu, 12 Dec 2013 17:09:59 +0000 (17:09 +0000)]
v4l2object: unref downstream pool
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Julien Isorce [Wed, 18 Dec 2013 18:37:23 +0000 (13:37 -0500)]
v4l2bufferpool: add gst_v4l2_buffer_pool_flush
STREAMOFF set all v4l2buffers to DEQUEUE state.
Then for CAPTURE we call QBUF on each buffer.
For OUTPUT the buffers are just push back in the GstBufferPool
base class 's queue.
But the loop actually looks like the same.
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Benjamin Gaignard [Mon, 16 Dec 2013 22:29:30 +0000 (17:29 -0500)]
v4l2object: Add vp8 support
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Thu, 12 Dec 2013 21:46:09 +0000 (16:46 -0500)]
v4l2object: Don't force framerate field for OUTPUT
If there is nothing that seems to force a certain framerate on output device, it is
preferable to simply not set that feild. This allow negotiation with tsdemux in a
decoder for example.
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Thu, 12 Dec 2013 19:07:03 +0000 (14:07 -0500)]
v4l2object: _v4l2fourcc_to_structure() can be static
This function is not used anymore outside v4l2object.
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Thu, 12 Dec 2013 19:22:26 +0000 (14:22 -0500)]
v4l2object: Add MPEG1/2 support
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Thu, 12 Dec 2013 17:18:45 +0000 (12:18 -0500)]
v4l2object: Ask for a decent buffer size when dealing with encoded formats
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Sat, 7 Dec 2013 19:03:53 +0000 (14:03 -0500)]
v4l2bufferpool: On warn on size change if n_planes > 1
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Julien Isorce [Tue, 31 Dec 2013 16:38:09 +0000 (16:38 +0000)]
v4l2object: check if translated format is valid
Also add a FIXME in gst_v4l2_object_setup_format
to note that the whole function has to be improved
in order to support ENCODED formats.
It requires to have an encoder device which we do not
have right now.
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Sat, 7 Dec 2013 15:31:15 +0000 (10:31 -0500)]
v4l2object: Validate returned dimensions
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Fri, 6 Dec 2013 00:36:25 +0000 (19:36 -0500)]
v4l2object: Ensure max is not smaller then min in decide_allocation
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Fri, 6 Dec 2013 00:36:06 +0000 (19:36 -0500)]
v4l2object: Don't keep the max paramter when using our own pool
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Fri, 6 Dec 2013 00:34:44 +0000 (19:34 -0500)]
v4l2bufferpool: Respect the suggested min buffer
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Thu, 5 Dec 2013 23:48:44 +0000 (18:48 -0500)]
v4l2object: Allocate pool if needed in decide_allocation
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Thu, 5 Dec 2013 23:49:19 +0000 (18:49 -0500)]
v4l2object: Add V4L2_CID_MIN_BUFFERS_FOR_CAPTURE support
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Thu, 5 Dec 2013 23:48:15 +0000 (18:48 -0500)]
v4l2: Move decide allocation into v4l2object
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Thu, 5 Dec 2013 18:51:13 +0000 (13:51 -0500)]
v4l2object: Implement _setup_format()
This method allow setting up the object from the currently configured format on the
device. This is useful for M2M element where input data decides the format that will
be set on capture side.
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Tue, 10 Dec 2013 19:34:17 +0000 (14:34 -0500)]
v4l2object: Split out saving format from set_format()
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Julien Isorce [Tue, 31 Dec 2013 15:37:26 +0000 (15:37 +0000)]
v4l2object: set only one plane for encoded format
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Wed, 4 Dec 2013 21:49:13 +0000 (16:49 -0500)]
v4l2object: Move code block where it belongs
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Wed, 4 Dec 2013 21:26:12 +0000 (16:26 -0500)]
v4l2object: Don't check format specific information
The number of plane, and the stride does not represent a capability change. Same caps
can have different stride from the default GstVideoInfo and the number of planes will
never change for 1 format.
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Wed, 4 Dec 2013 21:23:18 +0000 (16:23 -0500)]
v4l2object: Move the extrapolation of stride at the right place
Now that we have a stride array, we should extrapolate only when
eeded (non multi-planar buffer).
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Wed, 4 Dec 2013 20:09:44 +0000 (15:09 -0500)]
v4l2object: Move back assertions where they should be
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Wed, 4 Dec 2013 20:09:10 +0000 (15:09 -0500)]
v4l2object: Move mplane logic into gst_v4l2_object_get_caps_info()
It makes the gst_v4l2_object_set_format() slightly simplier and will make that
logic reusable. Note that gst_v4l2_object_has_mplane() will always return the
same value for one device. There is no need to check against the caps as this
has already been done by _open.
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Tue, 3 Dec 2013 23:27:47 +0000 (18:27 -0500)]
v4l2object: Split _v4l2fourcc_to_video_format
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Mon, 2 Dec 2013 23:05:11 +0000 (18:05 -0500)]
v4l2bufferpool: Request buffers only once
VIDIOC_REQBUFS allocates buffer, it has no place inside set_config. Also, some driver do
no allow multiple calls to this ioctl.
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Mon, 2 Dec 2013 20:26:50 +0000 (15:26 -0500)]
v4l2object: Don't validate dimension for encoded format
We set the dimensions just in case but don't validate them
afterwards. For some codecs the dimensions are *not* in the
bitstream, IIRC VC1 in ASF mode for example.
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Thu, 28 Nov 2013 22:10:29 +0000 (17:10 -0500)]
v4l2object: Quirks for dev without initial format
Most M2M have undefined behaviour initially when VIDIOC_G_FMT is called.
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Thu, 28 Nov 2013 22:09:26 +0000 (17:09 -0500)]
v4l2object: Add gst_v4l2_object_open_shared()
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Thu, 28 Nov 2013 22:07:05 +0000 (17:07 -0500)]
v4l2object: Implement gst_v4l2_dup()
This will duplicated the FD from another object and copy over the probed result.
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Thu, 28 Nov 2013 21:59:59 +0000 (16:59 -0500)]
v4l2object: make IO_MODE enum public
This is to allow adding a second io-mode property on M2M device like decoder so
input and output can be controlled separatly.
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Wed, 5 Jun 2013 03:42:24 +0000 (23:42 -0400)]
v4l2: better handle quirks activation
This way we can activate deactivate those quirks all at once at one
place.
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Wed, 5 Jun 2013 03:34:04 +0000 (23:34 -0400)]
v4l2: Fix h264 caps
V4L2_PIX_FMT_H264 is documentated as byte-stream (with start code). The ensure proper
negotiation with element like h264parse.
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Nicolas Dufresne [Fri, 6 Dec 2013 19:44:51 +0000 (14:44 -0500)]
v4l2object: Split caps in different categories
This is need to correctly expose capabilities on specialized devices
like decoders and encoders.
https://bugzilla.gnome.org/show_bug.cgi?id=720568
Tim-Philipp Müller [Fri, 10 Jan 2014 14:16:00 +0000 (14:16 +0000)]
matroskademux: don't leak TOC chapter list
Vincent Penquerc'h [Fri, 10 Jan 2014 08:52:16 +0000 (08:52 +0000)]
matroskamux: remove obsolete write-dummy-and-overwrite-on-eos code
The need for rewriting apparently is obsolete 0.10 leftover.
We now have caps for subtitles when we create the headers,
so we always write the correct data in the first place.
Tim-Philipp Müller [Thu, 9 Jan 2014 23:55:16 +0000 (23:55 +0000)]
rtprtxsend: remove duplicate assignment
Coverity CID 1151680
Vincent Penquerc'h [Thu, 9 Jan 2014 18:25:04 +0000 (18:25 +0000)]
matroskamux: write subtitle codec ID and data at start when known
This avoids issues with writing dummy data first, then having
to come back and write correct data later. Doing so prevents
the muxed stream from being actually streamable.
https://bugzilla.gnome.org/show_bug.cgi?id=712134
Sebastian Dröge [Thu, 9 Jan 2014 16:32:15 +0000 (17:32 +0100)]
configure: Include AvailabilityMacros.h for osxvideo check
Otherwise MAC_OS_X_VERSION_MIN_REQUIRED might not be defined
Thiago Santos [Thu, 9 Jan 2014 14:56:31 +0000 (11:56 -0300)]
qtmux: respect the HDLR box string format for mov and isomedia
Mov spec says it uses a pascal style string, while isomedia uses
a null terminated one. Store the current atoms flavor into the HDLR
to be able to generate the correct output.
https://bugzilla.gnome.org/show_bug.cgi?id=705982
Wim Taymans [Wed, 8 Jan 2014 10:28:04 +0000 (11:28 +0100)]
Revert "matroskamux: Use the running time for container timestamps, not buffer timestamps"
This reverts commit
b3aa8755fe07639f22e4104f4932d769d6c9075a.
We are already using the running-time because they were placed on the
buffers with gst_collect_pads_clip_running_time(). Arguably it would be
better to not modify the incomming buffers but collectpads seems to want
to use absolute timestamps from the buffers for finding the best buffer
(this can be changed with a custom compare function..).
Sebastian Dröge [Wed, 8 Jan 2014 09:41:24 +0000 (10:41 +0100)]
configure: Fix AC_COMPILE_IFELSE usage
Sebastian Dröge [Wed, 8 Jan 2014 09:31:18 +0000 (10:31 +0100)]
osxvideosink: Improve configure check for OSX >= 10.6
https://bugzilla.gnome.org/show_bug.cgi?id=721245
Aleix Conchillo Flaqué [Tue, 7 Jan 2014 20:13:51 +0000 (12:13 -0800)]
rtpbin: remove unused list of decoders
remove list of decoders, which are already handled by the list of elements.
https://bugzilla.gnome.org/show_bug.cgi?id=719938
Sebastian Dröge [Wed, 8 Jan 2014 08:46:55 +0000 (09:46 +0100)]
matroskamux: Error out if ADPCM caps don't contain the layout field
Nicola Murino [Fri, 3 Jan 2014 14:25:23 +0000 (15:25 +0100)]
matroskamux: Add support for g726 ADPCM
https://bugzilla.gnome.org/show_bug.cgi?id=720995
Wim Taymans [Tue, 7 Jan 2014 14:04:02 +0000 (15:04 +0100)]
rtspsrc: use new method to get media-type
Use the new method to get the media type of a transport.
Stefan Sauer [Mon, 6 Jan 2014 20:12:17 +0000 (21:12 +0100)]
wavparse: split the test
This way one failure won't shadow the other test and also if one fails we get
better disgnostics through the test-name.
Sebastian Dröge [Mon, 6 Jan 2014 13:54:46 +0000 (14:54 +0100)]
matroskamux: Add HEVC / h265 support
Sebastian Dröge [Mon, 6 Jan 2014 13:54:38 +0000 (14:54 +0100)]
matroskademux: Add HEVC / h265 support
Stefan Sauer [Mon, 6 Jan 2014 12:36:38 +0000 (13:36 +0100)]
wavparse: remove ifdef'ed code
We do have adtl and cue parse as part of toc handling alreday. The fmt code is a left over from <0.10 times.
Stefan Sauer [Mon, 6 Jan 2014 12:32:58 +0000 (13:32 +0100)]
avidemux, waveparse: more logging for unhandled chunks
Always print a warning with the tag and if possible do a memdump.
Stefan Sauer [Sun, 5 Jan 2014 21:47:42 +0000 (22:47 +0100)]
avidemux: expose 'strn' - stream name - as title tag
Stefan Sauer [Sun, 5 Jan 2014 21:41:24 +0000 (22:41 +0100)]
avidemux: parse fuji strd
We can get maker, model and capture date from this chunk.
Fixes #636143
Stefan Sauer [Sun, 5 Jan 2014 20:46:33 +0000 (21:46 +0100)]
avidemux: ... and use the local api both times
Stefan Sauer [Sun, 5 Jan 2014 20:38:14 +0000 (21:38 +0100)]
avidemux: copy the riff api for ncdt into the element
This chunk is avi specific, no need to expose this as public api.
Sebastian Dröge [Sun, 5 Jan 2014 09:28:21 +0000 (10:28 +0100)]
matroskamux: Add missing semicolon from last commit
Sebastian Dröge [Sun, 5 Jan 2014 09:22:37 +0000 (10:22 +0100)]
matroskamux: Use the running time for container timestamps, not buffer timestamps
Buffer timestamps have no real meaning here, and for selecting the next
buffer we already use the running time anyway.
Stefan Sauer [Sat, 4 Jan 2014 20:34:38 +0000 (21:34 +0100)]
avi: use new riff api to extract nikon metadata
Fixes #636143
Julien Isorce [Fri, 1 Nov 2013 16:41:43 +0000 (16:41 +0000)]
rtprtxsend/rtprtxreceive: generate gtk doc
George Kiagiadakis [Mon, 2 Dec 2013 10:26:09 +0000 (11:26 +0100)]
test/check: Verify rtprtxsend::ssrc-map property works as expected
George Kiagiadakis [Fri, 29 Nov 2013 18:35:44 +0000 (19:35 +0100)]
rtprtxreceive: modify to use a payload-type map like rtprtxsend
George Kiagiadakis [Fri, 29 Nov 2013 18:58:26 +0000 (19:58 +0100)]
rtprtxsend: do not keep history of packets with an unknown payload type
This allows to disable retransmission per payload type by not putting
a certain payload type in the map.
Wim Taymans [Thu, 2 Jan 2014 14:18:52 +0000 (15:18 +0100)]
rtprtxsend: Allow SSRC-multiplexing and multiple payload types in the original stream
Conflicts:
tests/examples/rtp/server-rtpaux.c
George Kiagiadakis [Mon, 25 Nov 2013 14:00:45 +0000 (15:00 +0100)]
rtprtxsend: Add an rtx-ssrc property to allow external control of the ssrc
This is useful when one needs to know the SSRC beforehands, so that it can
be used for SRTP for example.
Torrie Fischer [Wed, 13 Nov 2013 20:11:35 +0000 (15:11 -0500)]
examples: rtp: Add end-to-end rtpbin example with RTX elements
This example demonstrates how to use rtpbin with retransmission (rtx)
elements set in the place of rtpbin's "aux" elements in order to
enable RTP retransmission according to the rules of RFC4588.
Julien Isorce [Tue, 5 Nov 2013 17:35:01 +0000 (17:35 +0000)]
doc: add design-rtpauxiliary.txt to describe how rtpbin deals with auxiliary elements
Wim Taymans [Thu, 2 Jan 2014 13:48:49 +0000 (14:48 +0100)]
session: also push EOS event to RTCP srcpad
Wim Taymans [Thu, 2 Jan 2014 13:46:11 +0000 (14:46 +0100)]
session: place SSRC in Retransmission event
Julien Isorce [Fri, 1 Nov 2013 16:57:15 +0000 (16:57 +0000)]
tests/check: add rtpaux::test_simple_rtpbin_aux
It shows how to use "set-aux-receive" and "set-aux-send"
properties of rtpbin to set rtprtxsend and rtprtxreceive
Build 2 pipelines, one for rtpbin as a sender and one for
rtobin as a receive. Then transmit an audio stream.
It also drops some packets to activate restransmission and
check they are actually retransmited.
Julien Isorce [Fri, 1 Nov 2013 17:09:42 +0000 (17:09 +0000)]
tests/check: add rtpcollision::test_rtx_ssrc_collision unit test
check that rtxrtpsend changes its retransmission ssrc when
collision happens
George Kiagiadakis [Wed, 6 Nov 2013 10:34:13 +0000 (12:34 +0200)]
tests/check: add rtprtx::test_rtxreceive_data_reconstruction
This unit test verifies that retransmitted rtp packets coming out
of rtprtxreceive are the same as the original ones.
George Kiagiadakis [Tue, 5 Nov 2013 07:33:51 +0000 (09:33 +0200)]
rtprtxsend: use a realistic limit for the value of max-size-packets
G_MAXINT16 is chosen because if the queue contains more than
G_MAXINT16 packets, seqnum comparison will not work properly.
George Kiagiadakis [Mon, 4 Nov 2013 18:05:03 +0000 (20:05 +0200)]
rtprtxsend: use a GSequence to implement the buffer queue
This has the advantage that searching the queue to find the
buffer with the requested seqnum is done with binary search.
George Kiagiadakis [Mon, 4 Nov 2013 16:38:24 +0000 (18:38 +0200)]
rtprtxsend: retransmit packets in the same order as the rtx requests
George Kiagiadakis [Sat, 2 Nov 2013 17:56:44 +0000 (19:56 +0200)]
tests/check: Add unit test for rtxsend's max_size_time property
George Kiagiadakis [Tue, 29 Oct 2013 17:27:00 +0000 (18:27 +0100)]
rtprtxsend: Handle the max_size_time property
This property allows you to specify the amount of buffers
to keep in the retransmission queue expressed as time (ms)
instead of buffer count (which is the max_size_buffers property).
George Kiagiadakis [Sat, 2 Nov 2013 13:21:08 +0000 (15:21 +0200)]
rtprtxsend: keep important buffer information in a private structure
This is to avoid mapping a buffer every time we need to read a seqnum
or a timestamp.
George Kiagiadakis [Fri, 1 Nov 2013 10:58:47 +0000 (11:58 +0100)]
tests/check: Add rtprtx::test_rtxsender_packet_retention
This unit test verifies that the rtxsend element correctly maintains
a buffer of already transmitted rtp packets and that it can
re-transmit all of them correctly on demand. It also verifies
that the limit of this buffer (max-size-packets property) is respected.
Julien Isorce [Fri, 1 Nov 2013 16:22:13 +0000 (16:22 +0000)]
tests/check: add rtprtx::test_drop_multiple_sender unit test
Several senders / one receiver
Similar than test_drop_one_sender but with multiple senders
mixed through the funnel element.
It drops some packets and checks that they are retransmited
correctly.
Julien Isorce [Fri, 1 Nov 2013 16:21:00 +0000 (16:21 +0000)]
tests/check: add rtprtx::test_drop_one_sender unit test
Test for one sender / one receiver
Build the pipeline
videotestsrc ! rtpvrawpay ! rtprtxsend ! rtprtxreceive ! fakesink
and drop some buffers between rtprtxsend and rtprtxreceive
Then it checks that every dropped packet has been re-sent.
It also checks that not too much requests has been sent.
Julien Isorce [Fri, 1 Nov 2013 16:17:51 +0000 (16:17 +0000)]
tests/check: add rtprtx::test_push_forward_seq
add simple unit test that manually push buffers
in rtprtxsend connected to rtprtxreceive.
Drops some buffers and make sure they are retransmisted.
Julien Isorce [Fri, 1 Nov 2013 15:52:03 +0000 (15:52 +0000)]
rtpmanager: add new rtprtxsend / rtprtxreceive elements
The purpose of the sender RTX object is to keep a history
of RTP packets up to a configurable limit (in time). It will
listen for custom retransmission events from downstream. When
it receives a request for retransmission, it will look up the
requested seqnum in its list of stored packets. If the packet
is available, it will create a RTX packet according to RFC 4588
and send this as an auxiliary stream.
The receiver will listen to the custom retransmission events
from the downstream jitterbuffer and will remember the SSRC1
of the stream and seqnum that was requested. When it sees a
packet with one of the stored seqnum, it associates the SSRC2
of the stream with the SSRC1 of the master stream. From then
on it knows that SSRC2 is the retransmission stream of SSRC1.
This algorithm is stated in RFC 4588. For this algorithm to
work, RFC4588 also states that no two pending retransmission
requests can exist for the same seqnum and different SSRCs or
else it would be impossible to associate the retransmission with
the original requester SSRC.
When the RTX receiver has associated the retransmission packets,
it can depayload and forward them to the source pad of the element.
RTX is SSRC-multiplexed
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=711084
Julien Isorce [Tue, 5 Nov 2013 16:36:46 +0000 (16:36 +0000)]
doc: add design for rtp retransmission
Describe how rtprtxsend and rtprtxreceive generally work
but also how the association algorithm is implemented.
Reynaldo H. Verdejo Pinochet [Thu, 2 Jan 2014 23:23:05 +0000 (20:23 -0300)]
souphttpsrc: use status code macro instead of 407
Rest of the code is using the _PROXY_AUTHENTICATION_REQUIRED
macro too. Easier to understand if you don't recall HTTP
error codes by heart.