platform/upstream/gst-plugins-good.git
6 years agomatroskamux: add properties to control cluster duration
Nicola Murino [Mon, 17 Jul 2017 22:26:11 +0000 (00:26 +0200)]
matroskamux: add properties to control cluster duration

https://bugzilla.gnome.org/show_bug.cgi?id=784971

6 years agov4l2: UVC driver is named uvcvideo these days
Nicolas Dufresne [Tue, 18 Jul 2017 00:47:26 +0000 (20:47 -0400)]
v4l2: UVC driver is named uvcvideo these days

The quirk to avoid probing interlacing didn't work anymore as the driver
is now name uvcvideo. This should slightly speed up camera startup.

6 years agov4l2object: Remove unused defines
Nicolas Dufresne [Thu, 13 Jul 2017 01:02:39 +0000 (21:02 -0400)]
v4l2object: Remove unused defines

6 years agov4l2: Make gst_v4l2_get_capabilities static
Nicolas Dufresne [Thu, 13 Jul 2017 00:53:51 +0000 (20:53 -0400)]
v4l2: Make gst_v4l2_get_capabilities static

It's not used outside of v4l2_calls.c

6 years agov4l2: Merge v4l2_calls.h into gstv4l2object.h
Nicolas Dufresne [Thu, 13 Jul 2017 00:49:47 +0000 (20:49 -0400)]
v4l2: Merge v4l2_calls.h into gstv4l2object.h

First step of a larger cleanup, all function from v4l2_calls are in fact
methods on GstV4l2Object. This split makes the code really confusing.
This also remove no longer unused macros.

6 years agomeson: po: use glib preset and read language list from LINGUAS
Tim-Philipp Müller [Sat, 15 Jul 2017 11:48:19 +0000 (12:48 +0100)]
meson: po: use glib preset and read language list from LINGUAS

Supported since meson 0.37, so we can use it now.

6 years agov4l2object: Trace unknown fourcc as text
Nicolas Dufresne [Fri, 14 Jul 2017 16:12:56 +0000 (12:12 -0400)]
v4l2object: Trace unknown fourcc as text

This makes it easier to find out what is not supported.

6 years agov4l2: Don't probe for unneeded format
Nicolas Dufresne [Fri, 14 Jul 2017 15:54:57 +0000 (11:54 -0400)]
v4l2: Don't probe for unneeded format

For v4l2videodec/enc, we generate elements per formats, and in
this case we can speed up the start up by only probing the format
we care about.

6 years agov4l2videodec: Implement stable element names
Nicolas Dufresne [Thu, 13 Jul 2017 16:32:00 +0000 (12:32 -0400)]
v4l2videodec: Implement stable element names

Before that, each m2m node would be wrapped as a single, multi-format
decoder element. As a unique name was needed, we where using the device
name, which changes between re-boots. This led to unpredictable element
names. In this patch, we generate an element per codec, using
v4l2<codec>dec name. If there is multiple decoder for the same format,
the following elements will be named v4l2<node><codec>dec.

https://bugzilla.gnome.org/show_bug.cgi?id=784908

6 years agosouphttpsrc: Post an element message with the HTTP headers on the bus too
Sebastian Dröge [Thu, 13 Jul 2017 11:50:44 +0000 (14:50 +0300)]
souphttpsrc: Post an element message with the HTTP headers on the bus too

Instead of just sending a sticky event with them downstream. This allows
getting the HTTP headers easily in the application, and especially also
on errors.

6 years agoqtdemux: Fix parsing of RLE depth
Sebastian Dröge [Thu, 13 Jul 2017 09:47:02 +0000 (12:47 +0300)]
qtdemux: Fix parsing of RLE depth

Regression introduced by 86b427dc70562f891a551ffc9f96cefe1cafcddd

https://bugzilla.gnome.org/show_bug.cgi?id=784812

7 years agoshout2: use gint and guint in place of int and uint
Philippe Renon [Mon, 10 Jul 2017 16:57:11 +0000 (18:57 +0200)]
shout2: use gint and guint in place of int and uint

this fixes a compilation error with gcc 7.1.0 on mys2 where uint is not defined

https://bugzilla.gnome.org/show_bug.cgi?id=784758

7 years agortpgsmpay: fix accidental garbage data before actual payload
Yasushi SHOJI [Fri, 7 Jul 2017 12:15:57 +0000 (21:15 +0900)]
rtpgsmpay: fix accidental garbage data before actual payload

Do not allocate payload size outbuf if appending payload buffer.

The commit 137672ff1824948bda4b1b1967de8c24a0055b67 attached payload
to the output buffer but forgot to remove payload allocation.  That
effectively doubled payload size and add zero'ed or random bytes.

Makes the following pipeline work again:

gst-launch-1.0 -v audiotestsrc wave=2 ! gsmenc ! rtpgsmpay ! rtpgsmdepay ! gsmdec ! autoaudiosink

https://bugzilla.gnome.org/show_bug.cgi?id=784616

7 years agomatroskademux: segment seek position is expressed in buffer time
Mark Nauwelaerts [Sat, 1 Jul 2017 16:57:47 +0000 (18:57 +0200)]
matroskademux: segment seek position is expressed in buffer time

... so it need not be corrected again for stream start

7 years agoavidemux: provide average bitrate tag
Mark Nauwelaerts [Sun, 9 Jul 2017 08:54:27 +0000 (10:54 +0200)]
avidemux: provide average bitrate tag

7 years agoexamples: v4l2: fix wrong initializations brought by 4e8ad583022671c5
Reynaldo H. Verdejo Pinochet [Sat, 8 Jul 2017 06:49:44 +0000 (23:49 -0700)]
examples: v4l2: fix wrong initializations brought by 4e8ad583022671c5

https://bugzilla.gnome.org/show_bug.cgi?id=682770

7 years agoexamples: v4l2: add example for v4l2src renegotiation
Nicolas Dufresne [Fri, 27 Feb 2015 16:03:42 +0000 (13:03 -0300)]
examples: v4l2: add example for v4l2src renegotiation

Based on work from Thiago Santos <thiagoss@osg.samsung.com>

https://bugzilla.gnome.org/show_bug.cgi?id=682770

7 years agomeson: find python3 via python3 module
Tim-Philipp Müller [Fri, 7 Jul 2017 10:58:10 +0000 (11:58 +0100)]
meson: find python3 via python3 module

https://bugzilla.gnome.org/show_bug.cgi?id=783198

7 years agotests: rtpbin: fix build in uninstalled setup
Tim-Philipp Müller [Wed, 5 Jul 2017 13:44:41 +0000 (14:44 +0100)]
tests: rtpbin: fix build in uninstalled setup

7 years agortpsession: Send EOS if all internal sources sent bye
Olivier Crête [Tue, 4 Jul 2017 21:42:25 +0000 (17:42 -0400)]
rtpsession: Send EOS if all internal sources sent bye

The ones which are not internal should not matter, and we should
wait for all sources to have sent their BYEs.

And add unit test

https://bugzilla.gnome.org/show_bug.cgi?id=773218

7 years agortpsession: Only send EOS if all sources have been marked bye
Olivier Crête [Tue, 4 Jul 2017 16:24:41 +0000 (12:24 -0400)]
rtpsession: Only send EOS if all sources have been marked bye

Now that multiple sender RTPSource can share the same RTPSession, we
must not send an EOS unless they're all marked bye.

7 years agocaca: Do not include, unused, sys/time.h
Thibault Saunier [Tue, 4 Jul 2017 15:49:29 +0000 (11:49 -0400)]
caca: Do not include, unused, sys/time.h

Which moreover makes building on windows (mingw/msvc) fail:
 https://ci.appveyor.com/project/thiblahute/gst-build-ge9m5

7 years agortprtxreceive: Add memory and boudary checks
Nicolas Dufresne [Mon, 3 Jul 2017 15:47:13 +0000 (11:47 -0400)]
rtprtxreceive: Add memory and boudary checks

This element was not checking if mapping the RTP buffer and the payload
worked, and was not checking if the RTX payload was large enough.

https://bugzilla.gnome.org/show_bug.cgi?id=784484

7 years agosouphttpsrc: Unset limit on the number of connection if soup session sharing is used
Seungha Yang [Tue, 4 Jul 2017 05:58:00 +0000 (14:58 +0900)]
souphttpsrc: Unset limit on the number of connection if soup session sharing is used

Soup allows only up to two connections per host in a session,
if we use default value. When session sharing is used, however,
more connections might be required in a session.
(e.g., multi-audio adaptive streaming case)

https://bugzilla.gnome.org/show_bug.cgi?id=784495

7 years agoimagefreeze: fix use-after-free on seek event
Tim-Philipp Müller [Mon, 3 Jul 2017 19:27:29 +0000 (20:27 +0100)]
imagefreeze: fix use-after-free on seek event

Get seqnum before unreffing the seek event.

https://bugzilla.gnome.org/show_bug.cgi?id=784486

7 years agoqtmux: robustify time tracking for sparse subtitle stream
Mark Nauwelaerts [Sat, 1 Jul 2017 16:59:14 +0000 (18:59 +0200)]
qtmux: robustify time tracking for sparse subtitle stream

7 years agoqtmux: correctly track chunk size of subtitle stream
Mark Nauwelaerts [Sat, 1 Jul 2017 16:59:07 +0000 (18:59 +0200)]
qtmux: correctly track chunk size of subtitle stream

... thereby ensuring correct chunk offset tracking for all streams.

7 years agortpstats: fix assertion 'denom > 0' failed
Julien Isorce [Tue, 27 Jun 2017 14:59:18 +0000 (15:59 +0100)]
rtpstats: fix assertion 'denom > 0' failed

gst_util_uint64_scale_int takes a gint as denom parameter
whereas ctx->clock_rate is a guint32.

It happens when gst_rtp_packet_rate_ctx_reset set clock_rate
to -1.

So just define clock_rate as gint like it is done in rtpsource.h

https://bugzilla.gnome.org/show_bug.cgi?id=784250

7 years agov4l2: Block recursive calls to resurect_buffer
Matt Fischer [Wed, 28 Jun 2017 19:05:27 +0000 (14:05 -0500)]
v4l2: Block recursive calls to resurect_buffer

When resurrecting a buffer, the subsequent free call can result
in the group-released handler being called again, which causes
a recursive loop.  This patch blocks the signal handler during
the time that it executes, ensuring that the loop will not occur.

https://bugzilla.gnome.org/show_bug.cgi?id=759292

7 years agotests: souphttpsrc: Avoid deprecated ssl-ca-file property
Jan Alexander Steffens (heftig) [Tue, 20 Jun 2017 14:39:36 +0000 (16:39 +0200)]
tests: souphttpsrc: Avoid deprecated ssl-ca-file property

SoupSession's ssl-ca-file property is deprecated. Use the recommended
tls-database property.

This is a bit more complex as it requires creating a GTlsFileDatabase
object for an absolute (!) path to the CA certificates file.

https://bugzilla.gnome.org/show_bug.cgi?id=784005

7 years agotests: souphttpsrc: Avoid deprecated server ssl properties
Jan Alexander Steffens (heftig) [Tue, 20 Jun 2017 14:37:55 +0000 (16:37 +0200)]
tests: souphttpsrc: Avoid deprecated server ssl properties

The ssl-cert-file and ssl-key-file properties are deprecated. Use the
soup_server_set_ssl_cert_file function to load the files.

https://bugzilla.gnome.org/show_bug.cgi?id=784005

7 years agotests: souphttpsrc: Make ssl_cert/key_file static
Jan Alexander Steffens (heftig) [Tue, 20 Jun 2017 14:34:41 +0000 (16:34 +0200)]
tests: souphttpsrc: Make ssl_cert/key_file static

Just a bit of cleanup.

https://bugzilla.gnome.org/show_bug.cgi?id=784005

7 years agotests: souphttpsrc: Update test-cert.pem
Jan Alexander Steffens (heftig) [Tue, 20 Jun 2017 14:28:35 +0000 (16:28 +0200)]
tests: souphttpsrc: Update test-cert.pem

Recent GnuTLS disregards the Common Name and only looks at the Subject
Alternative Name extension. Since our test-cert has no SAN extension,
validation fails.

Generate a new certificate with SAN. In addition to 127.0.0.1, for good
measure make it valid for localhost and ::1, too.

https://bugzilla.gnome.org/show_bug.cgi?id=784005

7 years agosouphttpsrc: Allow any type of proxy
Nicolas Dufresne [Thu, 29 Jun 2017 19:22:39 +0000 (15:22 -0400)]
souphttpsrc: Allow any type of proxy

Currently we only allowed HTTP proxy. Don't filter for the scheme, just check
if it looks like an URI. Soup will warn if the URI is invalid or if
proxy protocol is not supported. This enables using SOCKS 4/5 which is
directly implemented into GIO.

https://bugzilla.gnome.org/show_bug.cgi?id=783012

7 years agov4l2: increase by one the number of allocated buffers
Guillaume Desmottes [Wed, 24 May 2017 13:07:51 +0000 (15:07 +0200)]
v4l2: increase by one the number of allocated buffers

Increasing this number fix a buffer starvation problem I'm hitting
with a "v4l2src ! kmssink" pipeline.

kmssink requests 2 buffer as it keeps a reference on the last rendered
one. So we were allocating 3 buffers for the pipeline.
Once the first 2 buffers have been pushed we ended up with:
- one buffer queued in v4l2
- one being pushed
- one kept as last rendered

If this 3rd buffer is released after that v4l2 used the first one to
capture we end up with a buffer starvation problem as no buffer is currently
queued in v4l2 for capture.

Fixing this by adding one extra buffer to the pipeline so when one
buffer is being pushed downstream the other can already be queued to
capture the next frame.

We were already adding 3 buffers if downstream didn't reply to the
allocation query. I reduced this number to 2 to compensate the extra
buffer which is now always added.

https://bugzilla.gnome.org/show_bug.cgi?id=783049

7 years agortspsrc: Create send/recv mutexes once, not on every connect()
Sebastian Dröge [Thu, 29 Jun 2017 15:59:58 +0000 (18:59 +0300)]
rtspsrc: Create send/recv mutexes once, not on every connect()

Also fixes a crash caused by freeing an uninitialized mutex in an error
case.

https://bugzilla.gnome.org//show_bug.cgi?id=784282

7 years agov4l2allocator: Fix memory leak with dmabuf
Matt Fischer [Tue, 27 Jun 2017 23:20:17 +0000 (18:20 -0500)]
v4l2allocator: Fix memory leak with dmabuf

This patch fixes a memory leak that is caused if the dmabuf file
descriptor dup fails.  Previously, _cleanup_failed_alloc() would
not unref the memory because mems_allocated had not yet been
incremented.

https://bugzilla.gnome.org/show_bug.cgi?id=784302

7 years agoqtdemux: specify '_swr' atom as a container atom
Mark Nauwelaerts [Wed, 28 Jun 2017 17:46:04 +0000 (19:46 +0200)]
qtdemux: specify '_swr' atom as a container atom

... so it is parsed as an mp4 style metadata atom as written by muxer

7 years agoqtmux: initialize mdhd language code as undefined
Mark Nauwelaerts [Tue, 27 Jun 2017 18:14:57 +0000 (20:14 +0200)]
qtmux: initialize mdhd language code as undefined

7 years agortpjitterbuffer: Add a faststart-min-packets property
Nicolas Dufresne [Thu, 22 Jun 2017 19:34:42 +0000 (15:34 -0400)]
rtpjitterbuffer: Add a faststart-min-packets property

When set this property will allow the jitterbuffer to start delivering
packets as soon as N most recent packets have consecutive seqnum. A
faststart-min-packets of zero disables this feature. This heuristic is
also used in rtpsource which implements the probation mechanism and a
similar heuristic is used to handle long gaps.

https://bugzilla.gnome.org/show_bug.cgi?id=769536

7 years agomeson: Allow using glib as a subproject
Thibault Saunier [Fri, 23 Jun 2017 20:18:57 +0000 (16:18 -0400)]
meson: Allow using glib as a subproject

7 years agomeson: build examples
Tim-Philipp Müller [Mon, 26 Jun 2017 10:09:48 +0000 (11:09 +0100)]
meson: build examples

https://bugzilla.gnome.org/show_bug.cgi?id=784134

7 years agomeson: fix with-package-name option
Tim-Philipp Müller [Mon, 26 Jun 2017 08:47:55 +0000 (09:47 +0100)]
meson: fix with-package-name option

https://bugzilla.gnome.org/show_bug.cgi?id=784082

7 years agomeson: tests: icles: simplify build file
Tim-Philipp Müller [Mon, 26 Jun 2017 08:38:46 +0000 (09:38 +0100)]
meson: tests: icles: simplify build file

7 years agomeson: build tests/icles/
Tim-Philipp Müller [Sun, 25 Jun 2017 23:22:05 +0000 (00:22 +0100)]
meson: build tests/icles/

https://bugzilla.gnome.org/show_bug.cgi?id=784134

7 years agoqtmux: correctly calculate overall first_ts to ensure stream sync
Mark Nauwelaerts [Mon, 19 Jun 2017 19:13:42 +0000 (21:13 +0200)]
qtmux: correctly calculate overall first_ts to ensure stream sync

... by minding and compensating for the dts_adjustment that may have
been introduced in the PTS timeline.

7 years agomatroskademux: track highest known cluster position and time
Mark Nauwelaerts [Sat, 10 Jun 2017 13:14:41 +0000 (15:14 +0200)]
matroskademux: track highest known cluster position and time

... to use as a fallback initial duration estimate and to provide for
interpolation when scanning for position.

7 years agomatroskademux: improve and simplify searching for cluster and position
Mark Nauwelaerts [Sat, 10 Jun 2017 11:46:20 +0000 (13:46 +0200)]
matroskademux: improve and simplify searching for cluster and position

... avoiding inefficiency proportional to file size

7 years agomatroskademux: increase chunk size when scanning for cluster
Mark Nauwelaerts [Thu, 8 Jun 2017 14:55:29 +0000 (16:55 +0200)]
matroskademux: increase chunk size when scanning for cluster

7 years agomatroskademux: maintain variable state when searching for position
Mark Nauwelaerts [Thu, 8 Jun 2017 14:39:06 +0000 (16:39 +0200)]
matroskademux: maintain variable state when searching for position

... so skipping to next cluster happens efficiently

7 years agomeson: build raw1394 plugin
Tim-Philipp Müller [Fri, 23 Jun 2017 23:21:00 +0000 (00:21 +0100)]
meson: build raw1394 plugin

https://bugzilla.gnome.org/show_bug.cgi?id=784134

7 years agomeson: build aalib plugin
Tim-Philipp Müller [Fri, 23 Jun 2017 22:50:00 +0000 (23:50 +0100)]
meson: build aalib plugin

https://bugzilla.gnome.org/show_bug.cgi?id=784134

7 years agomeson: build caca plugin
Tim-Philipp Müller [Fri, 23 Jun 2017 22:38:27 +0000 (23:38 +0100)]
meson: build caca plugin

https://bugzilla.gnome.org/show_bug.cgi?id=784134

7 years agodocs: update for git master
Tim-Philipp Müller [Fri, 23 Jun 2017 19:01:59 +0000 (20:01 +0100)]
docs: update for git master

7 years agosys: remove sunaudio plugin
Tim-Philipp Müller [Fri, 23 Jun 2017 18:52:04 +0000 (19:52 +0100)]
sys: remove sunaudio plugin

Even though hooked up to the build system, it's clear that no one
has ever built or used this with GStreamer 1.x. It wants to link
against libgstinterfaces, which no longer exists. And uses 0.10-style
raw audio caps. And the last meaningful change was done in 2009.
Let's just remove it.

7 years agomeson: build oss4 plugin
Tim-Philipp Müller [Fri, 23 Jun 2017 18:35:28 +0000 (19:35 +0100)]
meson: build oss4 plugin

https://bugzilla.gnome.org/show_bug.cgi?id=784134

7 years agomeson: build oss plugin
Tim-Philipp Müller [Fri, 23 Jun 2017 18:23:52 +0000 (19:23 +0100)]
meson: build oss plugin

https://bugzilla.gnome.org/show_bug.cgi?id=784134

7 years agortspsrc: Actually use the receive lock when receiving, not the send lock
Sebastian Dröge [Thu, 22 Jun 2017 08:38:56 +0000 (11:38 +0300)]
rtspsrc: Actually use the receive lock when receiving, not the send lock

7 years agoflactag: Fix warning with the newly added GstStateChange values
Thibault Saunier [Thu, 15 Jun 2017 15:46:54 +0000 (11:46 -0400)]
flactag: Fix warning with the newly added GstStateChange values

https://bugzilla.gnome.org/show_bug.cgi?id=783798

7 years agortspsrc: do not checksum the stream id
Mathieu Duponchelle [Thu, 15 Jun 2017 17:09:26 +0000 (19:09 +0200)]
rtspsrc: do not checksum the stream id

https://bugzilla.gnome.org/show_bug.cgi?id=783307

7 years agoqtmux: add support for muxing PNG
Tim-Philipp Müller [Thu, 15 Jun 2017 22:31:24 +0000 (23:31 +0100)]
qtmux: add support for muxing PNG

Demuxer already supported it.

7 years agortspsrc: Use a mutex for protecting against concurrent send/receives
Sebastian Dröge [Thu, 15 Jun 2017 07:40:51 +0000 (10:40 +0300)]
rtspsrc: Use a mutex for protecting against concurrent send/receives

We currently send data to the RTSP connection from multiple threads:
whenever a command is to be handled and whenever RTCP is generated. This
can cause data corruption or worse if both happen at the same time.

As such, protect gst_rtsp_connection_send() and gst_rtsp_connection_receive()
calls with a mutex. While this means that we hold a mutex during the IO
operation, this is not actually a problem as the IO operation can be
interrupted (gst_rtsp_connection_flush()) at any time and is blocking by
itself anyway.

7 years agoqtmux: Un-merge the last two stsc entries after serializing
Sebastian Dröge [Thu, 15 Jun 2017 08:50:44 +0000 (11:50 +0300)]
qtmux: Un-merge the last two stsc entries after serializing

The last entry will most likely get new samples added to it in "robust"
muxing mode, changing the samples_per_chunk and thus making it wrong to
keep the last two entries merged. It will run into an assertion later
when adding a new sample to the chunk.

Thanks to gdiener@cardinalpeak.com for the analysis of the bug and
proposal for a solution.

7 years agowavparse: Actually clip to upstream size instead of size of the data chunk
Sebastian Dröge [Tue, 13 Jun 2017 21:09:25 +0000 (00:09 +0300)]
wavparse: Actually clip to upstream size instead of size of the data chunk

There might be other chunks after the data chunk, so clipping the chunk
size with the data size can lead to a negative number and all following
calculations go wrong and cause crashes or worse.

This was introduced in 3ac119bbe2c360e28c087cf3852ea769d611b120.

https://bugzilla.gnome.org/show_bug.cgi?id=783760

7 years agosplitmux: Drop allocation queries
Vivia Nikolaidou [Tue, 13 Jun 2017 14:40:19 +0000 (17:40 +0300)]
splitmux: Drop allocation queries

They can cause us to deadlock, while we're waiting for a new frame and
upstream is waiting for the allocation query to be answered before
sending a frame

https://bugzilla.gnome.org/show_bug.cgi?id=783753

7 years agortspsrc: uniquify stream ids
Mathieu Duponchelle [Thu, 1 Jun 2017 00:03:27 +0000 (02:03 +0200)]
rtspsrc: uniquify stream ids

https://bugzilla.gnome.org/show_bug.cgi?id=783307

7 years agomeson: Do not use path separator in test names
Thibault Saunier [Wed, 7 Jun 2017 16:47:59 +0000 (12:47 -0400)]
meson: Do not use path separator in test names

Avoiding warnings like:

    WARNING: Target "elements/audioamplify" has a path separator in its name.

7 years agoFix v4l2 example
Nicolas Dufresne [Tue, 6 Jun 2017 15:29:29 +0000 (11:29 -0400)]
Fix v4l2 example

7 years agoqtdemux: remove not needed code
Jimmy Ohn [Mon, 5 Jun 2017 07:55:13 +0000 (16:55 +0900)]
qtdemux: remove not needed code

remove not needed code about res variable.

https://bugzilla.gnome.org/show_bug.cgi?id=783422

7 years agov4l2videoenc: Make sure min_buffers is valid
Nicolas Dufresne [Fri, 2 Jun 2017 18:01:17 +0000 (14:01 -0400)]
v4l2videoenc: Make sure min_buffers is valid

When upstream does no use the v4l2videoenc pool, we need to activate
that internal pool. Though, we relied the driver to provide a minimum
required buffer, which Qualcomm Venus driver don't currently provide.

https://bugzilla.gnome.org/show_bug.cgi?id=783361

7 years agortph265depay: fix caps leak
Tim-Philipp Müller [Fri, 2 Jun 2017 10:30:15 +0000 (11:30 +0100)]
rtph265depay: fix caps leak

7 years agortph264depay: simplify buffer accumulation control flow
Tim-Philipp Müller [Fri, 26 May 2017 15:30:06 +0000 (16:30 +0100)]
rtph264depay: simplify buffer accumulation control flow

There is no difference between pushing out a buffer directly
with gst_rtp_base_depayload_push() and returning it from the
process function. The base class will just call _depayload_push()
on the returned buffer as well.

So instead of marshalling buffers through three layers and back,
just push them from one place in handle_nal() and always return
NULL from the process vfunc. This simplifies the code a little.

Also rename _push_fragmentation_unit() to _finish_fragmentation_unit()
for clarity. Push sounds like it means being pushed out, whereas
it might just be pushed into an adapter.

This change has the side-effect that multiple NALs in a single STAP
(such as SPS/PPS) may no longer be pushed out as a single buffer if
we output NALs in byte-stream format (i.e. not aggregate AUs), but
that shouldn't really make any difference to anyone.

7 years agortpsession: print value of unknown RTCP Payload Type
Juan Navarro [Tue, 30 May 2017 20:23:10 +0000 (22:23 +0200)]
rtpsession: print value of unknown RTCP Payload Type

This adds printing the actual value of any unknown RTCP PT
to the already existing WARNING log message.

https://bugzilla.gnome.org/show_bug.cgi?id=783248

7 years agov4l2videoenc: Don't leak VideoCodecState
Edward Hervey [Fri, 26 May 2017 15:52:19 +0000 (17:52 +0200)]
v4l2videoenc: Don't leak VideoCodecState

CID #1409852

7 years agodvdemux: Remove un-needed variable check
Edward Hervey [Fri, 26 May 2017 15:48:01 +0000 (17:48 +0200)]
dvdemux: Remove un-needed variable check

if pad wasn't present by now everything would have broken before

CID #1409854

7 years agopo: update POTFILES
Piotr Drąg [Thu, 25 May 2017 13:26:37 +0000 (15:26 +0200)]
po: update POTFILES

https://bugzilla.gnome.org/show_bug.cgi?id=783093

7 years agov4l2videoenc: Remove unused function
Nicolas Dufresne [Wed, 24 May 2017 18:19:27 +0000 (14:19 -0400)]
v4l2videoenc: Remove unused function

7 years agov4l2: Don't redefine __bitwise if already set
Nicolas Dufresne [Sun, 21 May 2017 13:29:11 +0000 (15:29 +0200)]
v4l2: Don't redefine __bitwise if already set

https://bugzilla.gnome.org/show_bug.cgi?id=728438

7 years agov4l2: Add Video Encoder support
Ayaka [Tue, 23 May 2017 18:40:56 +0000 (14:40 -0400)]
v4l2: Add Video Encoder support

This implements H264 encoding support using generic V4L2 interface. It is
reported to work with Samsung MFC driver, IXM.6 CODA driver and
Qualcomm mainline Venus driver. Other platform should be supported as
none of this work is platform specific.

The implementation consist of a GstV4l2VideoEnc base class, which
implements the core streaming functionality. This base class is implemented
by GstV4l2H264Enc class that implements the caps negotiation specific to
H264 profiles and level. This implementation supports hardware with multiple
H264 encoder. Though, to make it simplier to use, the first discovered H264
encoder will be named v4l2h264enc. Other encoder found during discovery will
have a unique name like v4l2video0h264enc.

This work is the combined work of multiple developpers in the last 3
years. Thanks to all of the contributors:

  Ayaka <ayaka@soulik.info>
  Frédéric Sureau <frederic.sureau@vodalys.com>
  Jean-Michel Hautbois <jean-michel.hautbois@veo-labs.com>
  Nicolas Dufresne <nicolas.dufresne@collabora.com>
  Pablo Anton <pablo.anton@vodalys-labs.com>

https://bugzilla.gnome.org/show_bug.cgi?id=728438

7 years agov4l2videodec: Remove unused forward declaration
Nicolas Dufresne [Tue, 23 May 2017 18:36:37 +0000 (14:36 -0400)]
v4l2videodec: Remove unused forward declaration

https://bugzilla.gnome.org/show_bug.cgi?id=728438

7 years agov4l2pool: Fix wrong error message
Ayaka [Mon, 5 Oct 2015 15:30:46 +0000 (16:30 +0100)]
v4l2pool: Fix wrong error message

https://bugzilla.gnome.org/show_bug.cgi?id=728438

7 years agov4l2: increase pre-allocated encoded buffer size
Ayaka [Mon, 5 Oct 2015 15:20:19 +0000 (16:20 +0100)]
v4l2: increase pre-allocated encoded buffer size

As of today, the MFC encoder often need to exceed that 1 MB
size for encoded buffer we fixed earlier for decoding.

https://bugzilla.gnome.org/show_bug.cgi?id=728438

7 years agortpopusdepay: minor perf improvements
Tim-Philipp Müller [Wed, 24 May 2017 15:32:30 +0000 (16:32 +0100)]
rtpopusdepay: minor perf improvements

Use the ::process_rtp_packet() vfunc to avoid mapping the
RTP buffer twice.

gst_rtp_buffer_get_payload_buffer() returns a new sub-buffer
which will always be writable, so no need to make it writable.

7 years agortp: opus: use existing utility funcs for copying/dropping metas
Tim-Philipp Müller [Wed, 24 May 2017 15:14:54 +0000 (16:14 +0100)]
rtp: opus: use existing utility funcs for copying/dropping metas

We had our own copies of those while the code was in -bad, but now
we can use the existing utility functions instead of re-implementing
them.

7 years agortp: cache meta tag quarks and add more utility functions for metas
Tim-Philipp Müller [Wed, 24 May 2017 11:57:10 +0000 (12:57 +0100)]
rtp: cache meta tag quarks and add more utility functions for metas

Every g_quark_from_static_string() is a hash table lookup serialised
on the global quark lock in GLib. Let's just look up the two quarks
we need once and cache them locally for future use. While we're at it,
add new utility functions for the two most commonly used tags
(audio + video). Make first argument a gpointer so we don't have to
cast and make the code ugly. These are used for logging purposes
only anyway.

7 years agoaacparse : Fix, Caps were not set while reusing aacparse
vijay [Wed, 24 May 2017 06:03:05 +0000 (11:33 +0530)]
aacparse : Fix, Caps were not set while reusing aacparse

While reusing aacparse caps were not set.This fix enables aacparse to reuse in same pipeline.

https://bugzilla.gnome.org/show_bug.cgi?id=783027

7 years agomeson: don't need config.h.meson any longer
Tim-Philipp Müller [Sun, 21 May 2017 16:45:34 +0000 (17:45 +0100)]
meson: don't need config.h.meson any longer

7 years agoshout2send: use non-blocking I/O and a configurable network operations timeout
George Kiagiadakis [Mon, 24 Apr 2017 13:55:22 +0000 (16:55 +0300)]
shout2send: use non-blocking I/O and a configurable network operations timeout

This allows timing out on network errors much earlier
(currently it takes ~15min to timeout) and we can still
unlock and change state in the meantime.

https://bugzilla.gnome.org/show_bug.cgi?id=571722

7 years agomeson: make C++ compiler optional
Tim-Philipp Müller [Sun, 21 May 2017 09:37:19 +0000 (10:37 +0100)]
meson: make C++ compiler optional

It's only needed for the taglib plugin which is optional.

7 years agomultifile: remove some cruft
Tim-Philipp Müller [Sun, 21 May 2017 09:33:43 +0000 (10:33 +0100)]
multifile: remove some cruft

7 years agoosxaudio: fixes playback of mono streams with no channel-mask field in caps
Josep Torra [Sat, 20 May 2017 15:09:52 +0000 (17:09 +0200)]
osxaudio: fixes playback of mono streams with no channel-mask field in caps

Fixes a negotiation error seen when trying to playback of a .MOV file with
a mono AAC audio stream decoded by avcdec_aac that doesn't set channel-mask
field but sink was requiring channel-mask=0x3.

7 years agodvdemux: Push tag event to both pads
Ravi Kiran K N [Sun, 6 Sep 2015 15:19:59 +0000 (20:49 +0530)]
dvdemux: Push tag event to both pads

Tags are pushed to "videosrcpad"/"audiosrcpad" in
gst_dvdemux_add_pad() method, however they will be NULL
in this method, hence tags are not pushed.
Instead, send tag event to "pad" created gst_dvdemux_add_pad().

Signal no-more-pads when both pads are created

https://bugzilla.gnome.org/show_bug.cgi?id=743657

7 years agomeson: add options to set package name and origin
Tim-Philipp Müller [Sat, 20 May 2017 13:53:42 +0000 (14:53 +0100)]
meson: add options to set package name and origin

https://bugzilla.gnome.org/show_bug.cgi?id=782172

7 years agomultifilesink: fix property name in example pipeline
Luis de Bethencourt [Sat, 20 May 2017 10:40:33 +0000 (11:40 +0100)]
multifilesink: fix property name in example pipeline

Since the move from CVS the property name of the documentation example
has been filename instead of location. Users trying the gst-launch
command as is will get:
no property name "filename" in element

Fixing it.

7 years agoosxvideo: fix macOS 10.12 deprecation warnings
Josep Torra [Sat, 20 May 2017 09:13:40 +0000 (11:13 +0200)]
osxvideo: fix macOS 10.12 deprecation warnings

Add #defines to allow older versions of macOS to use the new constant names.

7 years agoisomp4: Safely ignore [skip] atoms
Edward Hervey [Sat, 13 May 2017 07:05:57 +0000 (09:05 +0200)]
isomp4: Safely ignore [skip] atoms

Instead of warning about them

7 years agosouphttpsrc: Make session sharing thread-safe on our side
Sebastian Dröge [Thu, 18 May 2017 12:10:30 +0000 (15:10 +0300)]
souphttpsrc: Make session sharing thread-safe on our side

https://bugzilla.gnome.org/show_bug.cgi?id=780140

7 years agoFix up package name and origin in some plugins
Tim-Philipp Müller [Thu, 18 May 2017 09:43:19 +0000 (10:43 +0100)]
Fix up package name and origin in some plugins

7 years agogst: Clear floating flag in constructor of all GstObject subclasses that are not...
Sebastian Dröge [Mon, 15 May 2017 16:51:47 +0000 (19:51 +0300)]
gst: Clear floating flag in constructor of all GstObject subclasses that are not owned by any parent

https://bugzilla.gnome.org/show_bug.cgi?id=743062

7 years ago1394: Sink the clock reference in the constructor
Sebastian Dröge [Mon, 15 May 2017 11:22:34 +0000 (14:22 +0300)]
1394: Sink the clock reference in the constructor

This is now needed as GstClock does not do that internally anymore,
because that broke bindings.

https://bugzilla.gnome.org/show_bug.cgi?id=743062