platform/upstream/gstreamer.git
10 years agohlsdemux: Prefer to use nettle for decryption but fall-back to libgcrypt
Sebastian Dröge [Sun, 9 Feb 2014 17:49:49 +0000 (18:49 +0100)]
hlsdemux: Prefer to use nettle for decryption but fall-back to libgcrypt

nettle is used by newer versions of gnutls, while older versions of gnutls
used libgcrypt. Support both for now as not every distro has nettle yet.

nettle is preferred as it is more efficient to use and much smaller.

10 years agohlsdemux: Update to non-deprecated GLib thread API
Sebastian Dröge [Sun, 9 Feb 2014 17:19:33 +0000 (18:19 +0100)]
hlsdemux: Update to non-deprecated GLib thread API

10 years agohlsdemux: Use libgcrypt directly instead of going through gnutls
Sebastian Dröge [Sun, 9 Feb 2014 17:09:36 +0000 (18:09 +0100)]
hlsdemux: Use libgcrypt directly instead of going through gnutls

gnutls is also just wrapping gcrypt, but we don't need any of
the TLS related functionality. We just need to be able to decrypt
AES128-CBC.

10 years agoopus: Remove unused variable from unit test
Sebastian Dröge [Sat, 8 Feb 2014 19:08:29 +0000 (20:08 +0100)]
opus: Remove unused variable from unit test

10 years agokate: Remove unused variable from unit test
Sebastian Dröge [Sat, 8 Feb 2014 19:07:57 +0000 (20:07 +0100)]
kate: Remove unused variable from unit test

10 years agoassrender: Fix compiler warning in test
Sebastian Dröge [Sat, 8 Feb 2014 19:07:16 +0000 (20:07 +0100)]
assrender: Fix compiler warning in test

error: suggest braces around initialization of subobject [-Werror,-Wmissing-braces]

10 years agosrtpdec: Fix comparison compiler warning
Sebastian Dröge [Sat, 8 Feb 2014 18:59:43 +0000 (19:59 +0100)]
srtpdec: Fix comparison compiler warning

error: comparison of constant -1 with expression of type
'GstSrtpCipherType' is always false

10 years agomotioncells: Remove unused private fields
Sebastian Dröge [Sat, 8 Feb 2014 18:29:33 +0000 (19:29 +0100)]
motioncells: Remove unused private fields

10 years agosegmentation: Fix integer underflow check
Sebastian Dröge [Sat, 8 Feb 2014 18:28:26 +0000 (19:28 +0100)]
segmentation: Fix integer underflow check

error: comparison of unsigned expression < 0 is
always false [-Werror,-Wtautological-compare]

10 years agodfbvideosink: Fix unitialized variable compiler warning
Sebastian Dröge [Sat, 8 Feb 2014 17:58:38 +0000 (18:58 +0100)]
dfbvideosink: Fix unitialized variable compiler warning

10 years agobz2: Fix typo in header include guards
Sebastian Dröge [Sat, 8 Feb 2014 17:57:30 +0000 (18:57 +0100)]
bz2: Fix typo in header include guards

10 years agovdpau: Fix comparison compiler warning
Sebastian Dröge [Sat, 8 Feb 2014 17:56:06 +0000 (18:56 +0100)]
vdpau: Fix comparison compiler warning

error: comparison of constant -1 with expression
of type 'const GstVideoFormat' is always false

10 years agouvch264src: Fix unitialized variable compiler warnings
Sebastian Dröge [Sat, 8 Feb 2014 17:54:58 +0000 (18:54 +0100)]
uvch264src: Fix unitialized variable compiler warnings

10 years agoid3tag: Fix uninitialized variable compiler warning
Sebastian Dröge [Sat, 8 Feb 2014 17:48:42 +0000 (18:48 +0100)]
id3tag: Fix uninitialized variable compiler warning

error: variable 'image_type' is used uninitialized whenever
'if' condition is false [-Werror,-Wsometimes-uninitialized]

10 years agodvdspu: Remove unused variable
Sebastian Dröge [Sat, 8 Feb 2014 17:45:10 +0000 (18:45 +0100)]
dvdspu: Remove unused variable

10 years agompegts: pmt: Detect descriptor length correct
Jesper Larsen [Sat, 8 Feb 2014 17:20:22 +0000 (18:20 +0100)]
mpegts: pmt: Detect descriptor length correct

10 years agompegts: Correctly check for PAT,PMT,CAT,TSDT
Jesper Larsen [Sat, 8 Feb 2014 16:43:03 +0000 (17:43 +0100)]
mpegts: Correctly check for PAT,PMT,CAT,TSDT

10 years agompegts: Fix some packetizing bugs
Jesper Larsen [Sat, 8 Feb 2014 12:08:02 +0000 (13:08 +0100)]
mpegts: Fix some packetizing bugs

- Length of NIT stream descriptors was not detected correct
- Reserved bits was not set according to EN 300 468, ISO/IEC 13818-1
- Also set output data size if the section was previously packetized

https://bugzilla.gnome.org/show_bug.cgi?id=723892

10 years agompegtspacketizer: Remove or move variable assignments
Edward Hervey [Fri, 7 Feb 2014 10:54:07 +0000 (11:54 +0100)]
mpegtspacketizer: Remove or move variable assignments

To where they're needed

10 years agodvb-descriptor: Remove unneeded increment
Edward Hervey [Fri, 7 Feb 2014 10:53:38 +0000 (11:53 +0100)]
dvb-descriptor: Remove unneeded increment

10 years agodvb-descriptor: Fix comparision
Edward Hervey [Fri, 7 Feb 2014 10:42:51 +0000 (11:42 +0100)]
dvb-descriptor: Fix comparision

we want to stop if one OR the other is NULL.

10 years agompegts: Updated docs with new API
Jesper Larsen [Thu, 6 Feb 2014 12:53:49 +0000 (13:53 +0100)]
mpegts: Updated docs with new API

10 years agompegtsmux: Use mpeg-ts lib for PAT/PMT sections
Jesper Larsen [Tue, 19 Nov 2013 11:21:31 +0000 (12:21 +0100)]
mpegtsmux: Use mpeg-ts lib for PAT/PMT sections

Rewrite of the PAT/PMT section handling to use the mpeg-ts library

10 years agompegtsmux: Add support for muxing SI tables
Jesper Larsen [Wed, 20 Nov 2013 10:14:46 +0000 (11:14 +0100)]
mpegtsmux: Add support for muxing SI tables

The muxer is now able to include DVB sections in the transport stream.

The si-interval property will determine how often the SI tables are
muxed into the stream.

The section is handled by the mpeg-ts library. Below is a small example
that will include a Netork Information Table with a Network Name
descriptor in the stream.

GstMpegTsNIT *nit;
GstMpegTsDescriptor *descriptor;
GstMpegTsSection *section;
GstElement *mpegtsmux;

gst_mpegts_initialize ();

nit = gst_mpegts_section_nit_new ();
nit->actual_network = TRUE;

descriptor = gst_mpegts_descriptor_from_dvb_network_name ("Network name");
g_ptr_array_add (nit->descriptors, descriptor);

section = gst_mpegts_section_from_nit (nit);

// mpegtsmux should be retrieved from the pipeline
gst_mpegts_section_send_event (section, mpegtsmux);
gst_mpegts_section_unref (section);

10 years agompegts: Support registration and custom descriptor
Jesper Larsen [Tue, 19 Nov 2013 10:30:33 +0000 (11:30 +0100)]
mpegts: Support registration and custom descriptor

Support for registration descriptor (0x05)
Add function to create a descriptor with custom tag and data

10 years agompegts: Add creation of DVB Subtitling descriptor
Jesper Larsen [Tue, 19 Nov 2013 10:15:52 +0000 (11:15 +0100)]
mpegts: Add creation of DVB Subtitling descriptor

Descriptor tag is 0x59

10 years agompegts: Support parsing of DVB Teletext descriptor
Jesper Larsen [Tue, 19 Nov 2013 09:50:30 +0000 (10:50 +0100)]
mpegts: Support parsing of DVB Teletext descriptor

Descriptor tag is 0x56

10 years agompegts: Add support for creating PAT/PMT
Jesper Larsen [Wed, 20 Nov 2013 10:06:03 +0000 (11:06 +0100)]
mpegts: Add support for creating PAT/PMT

10 years agompegts: Add program_number to GstMpegTsPMT
Jesper Larsen [Wed, 20 Nov 2013 10:04:58 +0000 (11:04 +0100)]
mpegts: Add program_number to GstMpegTsPMT

10 years agompegts: Add network name descriptor construction
Jesper Larsen [Mon, 28 Oct 2013 13:49:08 +0000 (14:49 +0100)]
mpegts: Add network name descriptor construction

Add function to create a Network Name DVB descriptor.

10 years agompegts: Add support for creating a NIT section
Jesper Larsen [Mon, 28 Oct 2013 13:48:13 +0000 (14:48 +0100)]
mpegts: Add support for creating a NIT section

Functions that will enable user to create Network Information Tables.

10 years agompegts: Add network_id to GstMpegTsNIT
Jesper Larsen [Mon, 28 Oct 2013 13:26:39 +0000 (14:26 +0100)]
mpegts: Add network_id to GstMpegTsNIT

The network_id is stored in the subtable extension. Make a field
in the GstMpegTsNIT for better code readability

10 years agompegts: Add functions to send sections as events
Jesper Larsen [Thu, 31 Oct 2013 12:47:23 +0000 (13:47 +0100)]
mpegts: Add functions to send sections as events

Added function that enables the user to send a GstMpegTsSection as
an event to a GstElement. (i.e. mpegtsmux)

10 years agompegts: Add functions to packetize section
Jesper Larsen [Mon, 28 Oct 2013 13:44:13 +0000 (14:44 +0100)]
mpegts: Add functions to packetize section

Sections needs to be packetized for use in i.e. mpegtsmux.
These functions handles the generic common parts of a GstMpegTsSection

10 years agompegts: Add init functions for descriptor/section
Jesper Larsen [Mon, 28 Oct 2013 13:39:49 +0000 (14:39 +0100)]
mpegts: Add init functions for descriptor/section

10 years agompegts: Add function to encode UTF8 strings
Jesper Larsen [Wed, 30 Oct 2013 09:43:21 +0000 (10:43 +0100)]
mpegts: Add function to encode UTF8 strings

This can be used to create descriptors with appropriate character maps

10 years agompegts: Use big endian UCS-2 coding
Jesper Larsen [Wed, 30 Oct 2013 09:42:07 +0000 (10:42 +0100)]
mpegts: Use big endian UCS-2 coding

Parsing of control codes requires a big endian character map

10 years agompegts: Add support for two-way iconv conversions
Jesper Larsen [Wed, 30 Oct 2013 09:33:18 +0000 (10:33 +0100)]
mpegts: Add support for two-way iconv conversions

To use in conversions from UTF-8 to another character map

10 years agoapplemedia: Fix libtool usage
Edward Hervey [Thu, 6 Feb 2014 15:17:39 +0000 (16:17 +0100)]
applemedia: Fix libtool usage

--tag=CC is needed for static build

10 years agompegtsdemux: Update plugin license field
Edward Hervey [Thu, 6 Feb 2014 13:21:52 +0000 (14:21 +0100)]
mpegtsdemux: Update plugin license field

The original code (old mpegtsparse) from which this plugin was based on
was dual-licensed. This allowed usage of the code under any of the
licenses (which including LGPL):

"""
* Alternatively, the contents of this file may be used under the terms of
* the GNU Lesser General Public License Version 2 or later (the "LGPL"),
* in which case the provisions of the LGPL are applicable instead
* of those above. If you wish to allow use of your version of this file only
* under the terms of the LGPL, and not to allow others to
* use your version of this file under the terms of the MPL, indicate your
* decision by deleting the provisions above and replace them with the notice
* and other provisions required by the LGPL. If you do not delete
* the provisions above, a recipient may use your version of this file under
* the terms of the MPL or the LGPL.
"""

When refactored (leading to the creation of this new plugin), I chose all
new code to be LGPL-only (which was allowed for pre-existing code) by removing
the MPL sections.

The headers were all updated, but not the plugin license field. This commit
fixes this.

10 years agotests: insertbin: update tests to 1.x scenario
Thiago Santos [Tue, 4 Feb 2014 17:15:40 +0000 (14:15 -0300)]
tests: insertbin: update tests to 1.x scenario

The thread that calls the success/failure callback can be the
same that is adding/removing the element as the IDLE probe can
happen instantly if the pad is not 'busy'.

This required moving some checks for the callback counter around
as well as removing some pad pushes from the main test thread as
they were made useless after the IDLE pad probe was fixed in core
by commit 0324358ebc8122a41ab499a3ef7973b61c3e621b

10 years agotests: camerabin: avoid warning if test is failing
Thiago Santos [Tue, 4 Feb 2014 02:39:35 +0000 (23:39 -0300)]
tests: camerabin: avoid warning if test is failing

In case the test fails the message won't be returned and it
will lead to an assertion that might be misleading to those
debugging the issue.

10 years agowrappercamerabinsrc: stop source to do internal reconfiguration safely
Thiago Santos [Tue, 4 Feb 2014 02:10:03 +0000 (23:10 -0300)]
wrappercamerabinsrc: stop source to do internal reconfiguration safely

In order to be able to change the caps on multiple capsfilters the
source element needs to be stopped, otherwise it will get a few
reconfigure events and might try to renegotiate while the bin
is still transitioning its caps, leading to a not-negotiated failure
and the image capture won't happen because the source will be
unusable.

The solution is to keep the source in paused while the caps are being
changed in the bin, and then bring the element back to playing once
it is done. Unfortunately this increases the image capture latency,
but it should always work.

A possible improvement to reduce the latency is to add another signal
to be called before 'start-capture': 'prepare-capture'. At this step
the camera source should set all caps it needs and get the source
ready for doing the capture as soon as 'start-capture' is called.
This can be done on a future commit

10 years agocamerabin: remove unused variable
Thiago Santos [Tue, 4 Feb 2014 02:09:28 +0000 (23:09 -0300)]
camerabin: remove unused variable

Probably a left over from 0.10 -> 1.x porting

10 years agompg123: improved error report and checks
Carlos Rafael Giani [Tue, 4 Feb 2014 16:22:27 +0000 (17:22 +0100)]
mpg123: improved error report and checks

Signed-off-by: Carlos Rafael Giani <dv@pseudoterminal.org>
10 years agouridownloader: Fix race condition between EOS handling and downloading a range
Duncan Palmer [Tue, 4 Feb 2014 11:52:25 +0000 (12:52 +0100)]
uridownloader: Fix race condition between EOS handling and downloading a range

https://bugzilla.gnome.org/show_bug.cgi?id=723134

10 years agoaiffparse: adaptive buffer size
Vincent Penquerc'h [Tue, 4 Feb 2014 10:46:16 +0000 (05:46 -0500)]
aiffparse: adaptive buffer size

Copied from wavparse, helps with CPU usage on high bitrate
files.

10 years agoSince b1a756fd, applemedia fails to build on osx 10.7, which lacks VideoToolbox.h...
Dan Kegel [Mon, 3 Feb 2014 02:17:06 +0000 (18:17 -0800)]
Since b1a756fd, applemedia fails to build on osx 10.7, which lacks VideoToolbox.h... so disable it on OSX if that header is missing.

https://bugzilla.gnome.org/show_bug.cgi?id=723508

10 years agoaiffparse: fix data start offset calculation not counting COMM size
Vincent Penquerc'h [Sun, 2 Feb 2014 13:50:53 +0000 (13:50 +0000)]
aiffparse: fix data start offset calculation not counting COMM size

In particular, this fixes seeking back to the start reading data
out of sync with sample start granularity, yielding swapped channels.

10 years agotests: gdpdepay: update to 1.x events
Thiago Santos [Sat, 1 Feb 2014 02:43:32 +0000 (23:43 -0300)]
tests: gdpdepay: update to 1.x events

Tests needed stream start at beginning and segment after the
caps event

10 years agogdp: add DTS to headers
Thiago Santos [Fri, 31 Jan 2014 21:38:44 +0000 (18:38 -0300)]
gdp: add DTS to headers

10 years agogdppay: update to 1.x reality
Thiago Santos [Fri, 31 Jan 2014 15:15:49 +0000 (12:15 -0300)]
gdppay: update to 1.x reality

* stream-start-id is mandatory at the beginning, so add that to the
gdp headers
* caps must be sent before new segment, invert the order from legacy
0.10 code

And fix the tests as a ref is now kept for those buffers that compose
the header

10 years agofisheye: Fix documentation
Colin Macdonald [Fri, 31 Jan 2014 21:58:16 +0000 (16:58 -0500)]
fisheye: Fix documentation

https://bugzilla.gnome.org/show_bug.cgi?id=723398

10 years agoh264parse: Default framerate to 0/1 when no informations found
Thibault Saunier [Wed, 29 Jan 2014 15:32:30 +0000 (16:32 +0100)]
h264parse: Default framerate to 0/1 when no informations found

It is not perfect but it allows us to be sure that the mandatory 'framerate'
field is present in the caps.

As soon as some information is found in the stream, that will be
updated.

https://bugzilla.gnome.org/show_bug.cgi?id=723243

10 years agotests: gdppay: refactor and add missing bits
Thiago Santos [Thu, 30 Jan 2014 22:45:08 +0000 (19:45 -0300)]
tests: gdppay: refactor and add missing bits

Most of the tests weren't updated after the sticky events order
and stream start. Fix that and refactor those tests check that
are the same to some common functions.

Those functions still don't actually test the content but at
least now they are in a single place and can be improved
without replication

10 years agotests: gdppay: Do not forget to send stream-start event on test
Thiago Santos [Thu, 30 Jan 2014 21:17:33 +0000 (18:17 -0300)]
tests: gdppay: Do not forget to send stream-start event on test

Even if we don't want to push the caps event, a stream-start must
be present before any data in 1.x

10 years agogdppay: push the stream-start event
Thiago Santos [Thu, 30 Jan 2014 21:09:13 +0000 (18:09 -0300)]
gdppay: push the stream-start event

instead of just swallowing upstream's stream-start into its
payload, push it as elements expect a stream-start before data

10 years agotsdemux: We do not support the SKIP seek flag but that is not a problem
Sebastian Dröge [Thu, 30 Jan 2014 19:28:04 +0000 (20:28 +0100)]
tsdemux: We do not support the SKIP seek flag but that is not a problem

10 years agoAutomatic update of common submodule
Edward Hervey [Thu, 30 Jan 2014 09:44:18 +0000 (10:44 +0100)]
Automatic update of common submodule

From d48bed3 to 1a07da9

10 years agotests: camerabin: improve preview handling in tests
Thiago Santos [Thu, 30 Jan 2014 06:28:45 +0000 (03:28 -0300)]
tests: camerabin: improve preview handling in tests

Always erase the preview image data to allow the next check for
it to operate properly

10 years agouridownloader: Always release the object lock in gst_uri_downloader_fetch_uri_with_ra...
Duncan Palmer [Wed, 29 Jan 2014 19:38:57 +0000 (20:38 +0100)]
uridownloader: Always release the object lock in gst_uri_downloader_fetch_uri_with_range()

even if downloader->priv->urisrc is not set. This avoids deadlock under some
some conditions on cancel.

https://bugzilla.gnome.org/show_bug.cgi?id=723127

10 years agoopencv: The plugin works with 2.4.8 too
Nicola Murino [Wed, 29 Jan 2014 12:25:37 +0000 (13:25 +0100)]
opencv: The plugin works with 2.4.8 too

https://bugzilla.gnome.org/show_bug.cgi?id=723230

10 years agocodecparsers: Don't use gst_ as prefix for private symbols
Edward Hervey [Wed, 29 Jan 2014 09:50:40 +0000 (10:50 +0100)]
codecparsers: Don't use gst_ as prefix for private symbols

Otherwise they get exposed/accessible in the resulting .so

And make check-exports will complain

10 years agotsdemux: Fix flags comparison operator precedence
Amaury Medeiros [Tue, 28 Jan 2014 02:55:57 +0000 (23:55 -0300)]
tsdemux: Fix flags comparison operator precedence

Fix operator precedence, so the whole operation is
negated instead of only flags variable.

https://bugzilla.gnome.org/show_bug.cgi?id=722096

10 years agompegts: install but nodist for generated header
Mark Nauwelaerts [Sun, 26 Jan 2014 13:42:52 +0000 (14:42 +0100)]
mpegts: install but nodist for generated header

10 years agoandroidmedia: Don't handle FLUSHING or NOT_LINKED as error
Sebastian Dröge [Sat, 25 Jan 2014 16:40:46 +0000 (17:40 +0100)]
androidmedia: Don't handle FLUSHING or NOT_LINKED as error

And also just stop the task for FLUSHING.

10 years agomxfdemux: Don't go into pull mode when the sequential flag is set
Sjoerd Simons [Fri, 24 Jan 2014 20:41:25 +0000 (21:41 +0100)]
mxfdemux: Don't go into pull mode when the sequential flag is set

When the scheduling query results has GST_SCHEDULING_FLAG_SEQUENTIAL set
in its flags don't go into pull mode to prevent over-eager seeking.

https://bugzilla.gnome.org/show_bug.cgi?id=722934

10 years agotests: h264parse: fix tests after leading 0x00 removal
Thiago Santos [Fri, 24 Jan 2014 15:36:59 +0000 (12:36 -0300)]
tests: h264parse: fix tests after leading 0x00 removal

Commit 6af387cd5ab2c946025e5499903e75ee87b063a9 made h264parse
strip a leading 0x00 byte from some output scenarios. This broke
tests as bs_to_nal test expects one more byte on the output.

Fix this by comparing the output with the expected stripped version,
too.

10 years agocodecparsers: refactor common nal parsing to nalutils
Thiago Santos [Fri, 24 Jan 2014 04:07:45 +0000 (01:07 -0300)]
codecparsers: refactor common nal parsing to nalutils

Moves common code from h264 and h265 to a separate file

10 years agoh264parse: Fix multiple SEI messages in one SEI RBSP parsing.
Aurélien Zanelli [Fri, 3 Jan 2014 08:44:28 +0000 (09:44 +0100)]
h264parse: Fix multiple SEI messages in one SEI RBSP parsing.

An SEI RBSP could contains more than one SEI message as specified in
7.4.2.3.1.

This commit change the parser API: the gst_h264_parser_parse_sei()
function now create and fill a GArray containing GstH264SEIMessage.

https://bugzilla.gnome.org/show_bug.cgi?id=721715

10 years agoh265parser: remove trailling 0x00 bytes as the spec doesn't allow them
Thiago Santos [Wed, 22 Jan 2014 13:26:32 +0000 (10:26 -0300)]
h265parser: remove trailling 0x00 bytes as the spec doesn't allow them

Just like h264, the spec doesn't allow the last bytes of a NAL to be 0x00.
So remove it as it is probably just padding.

This is related to https://bugzilla.gnome.org/show_bug.cgi?id=721384

10 years agoh264parser: remove trailling 0x00 bytes as the spec doesn't allow them
Thiago Santos [Mon, 20 Jan 2014 20:24:54 +0000 (17:24 -0300)]
h264parser: remove trailling 0x00 bytes as the spec doesn't allow them

The spec states that the last byte of a NAL 'shall not' be 0x00
and it is allowed for byte-stream format to add padding 0x00 for
alignment.

So our parser should strip any trailling 0x00.

https://bugzilla.gnome.org/show_bug.cgi?id=721384

10 years agompegtsmux: don't leak pad name when treating events
Justin Joy [Tue, 21 Jan 2014 14:42:58 +0000 (23:42 +0900)]
mpegtsmux: don't leak pad name when treating events

https://bugzilla.gnome.org/show_bug.cgi?id=722699

10 years agompegpsmux: update last_ts to mux correctly
Wim Taymans [Tue, 21 Jan 2014 09:27:20 +0000 (10:27 +0100)]
mpegpsmux: update last_ts to mux correctly

We use last_ts to decide what buffer to mux next so make sure that it
always contains the last known valid timestamp.

10 years agoh265parser: Initialize pointer correctly that is never assigned but freed in error...
duhui.lee [Tue, 21 Jan 2014 01:58:35 +0000 (10:58 +0900)]
h265parser: Initialize pointer correctly that is never assigned but freed in error cases

Fixes crash on broken streams.

https://bugzilla.gnome.org/show_bug.cgi?id=722669

10 years agod3dvideosink: First destroy the window, then unregister the class
Alexey Chernov [Tue, 21 Jan 2014 08:44:00 +0000 (09:44 +0100)]
d3dvideosink: First destroy the window, then unregister the class

It's impossible to create another pipeline with d3dvideosink after disposing
the previous one due to some problem in d3dvideosink. The message is: "Unable
to register Direct3D hidden window class".

I've evaluated the problem and it's that UnregisterClass() in working thread is
called before DestroyWindow() and UnregisterClass() does nothing.

https://bugzilla.gnome.org/show_bug.cgi?id=722622

10 years agompegpsmux: fix timestamp handling
Wim Taymans [Mon, 20 Jan 2014 15:25:51 +0000 (16:25 +0100)]
mpegpsmux: fix timestamp handling

If the first buffer that we handle for a stream has no timestamp, we
would never consider this pad again for muxing which causes queues to
fill up and pipelines to stall. Instead, try to mux pads with -1
timestamps as soon as possible.

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=722330

10 years agompegpsmux: return GST_FLOW_EOS when we pushed EOS
Wim Taymans [Mon, 20 Jan 2014 15:23:49 +0000 (16:23 +0100)]
mpegpsmux: return GST_FLOW_EOS when we pushed EOS

Return GST_FLOW_EOS from the collected function when we pushed EOS or
else it simply keeps on calling our collected function.

10 years agoapplemedia: Actually the enc/decbins are not used anywhere currently
Sebastian Dröge [Mon, 20 Jan 2014 09:38:16 +0000 (10:38 +0100)]
applemedia: Actually the enc/decbins are not used anywhere currently

Code is commented out.

10 years agoapplemedia: Fix build on iOS and probably also on OSX
Sebastian Dröge [Mon, 20 Jan 2014 09:34:57 +0000 (10:34 +0100)]
applemedia: Fix build on iOS and probably also on OSX

The video toolbox is not available on iOS, and also on OSX
we should probably build all source files that are used for
it.

https://bugzilla.gnome.org/show_bug.cgi?id=722590

10 years agoh265parse: Update caps when receive VPS/SPS/PPS
duhui.lee [Mon, 20 Jan 2014 06:21:42 +0000 (15:21 +0900)]
h265parse: Update caps when receive VPS/SPS/PPS

https://bugzilla.gnome.org/show_bug.cgi?id=722240

10 years agoh265parser: Fix segfault when parsing HRD parameter
duhui.lee [Mon, 20 Jan 2014 08:03:09 +0000 (17:03 +0900)]
h265parser: Fix segfault when parsing HRD parameter

https://bugzilla.gnome.org/show_bug.cgi?id=722240

10 years agoopenexrdec: Don't start scanning for the header from the start again each time
Sebastian Dröge [Sat, 18 Jan 2014 12:56:28 +0000 (13:56 +0100)]
openexrdec: Don't start scanning for the header from the start again each time

This will be incredible slow if the upstream block size is very small. Instead
continue scanning for the header where we previously stopped.

For the standard filesrc block-size this made decoding a file about
3 times faster.

https://bugzilla.gnome.org/show_bug.cgi?id=719890

10 years agosfdec: This is a decoder, not a demuxer
Sebastian Dröge [Sat, 18 Jan 2014 12:06:29 +0000 (13:06 +0100)]
sfdec: This is a decoder, not a demuxer

Don't confuse autopluggers!

https://bugzilla.gnome.org/show_bug.cgi?id=722316

10 years agotsdemux: Fix leak of PCROffsetGroup
Andrey Utkin [Fri, 17 Jan 2014 23:19:36 +0000 (01:19 +0200)]
tsdemux: Fix leak of PCROffsetGroup

https://bugzilla.gnome.org/show_bug.cgi?id=722462

10 years agoaudiomixer: Remove some useless assignments and skip the current part of the buffer...
Sebastian Dröge [Thu, 16 Jan 2014 11:18:45 +0000 (12:18 +0100)]
audiomixer: Remove some useless assignments and skip the current part of the buffer if the pad is muted

10 years agoresindvd: Merge changes across from master mpegdemux
Jan Schmidt [Thu, 16 Jan 2014 09:16:47 +0000 (20:16 +1100)]
resindvd: Merge changes across from master mpegdemux

Merge various changes and fixes from the master mpegdemux
Performance improvement from the way streams are organised,
return flow combining, language tag event generation,
adjustments and fixes in debug output, and things like that.

10 years agompegdemux: Merge some changes from the resindvd copy
Jan Schmidt [Thu, 16 Jan 2014 09:14:48 +0000 (20:14 +1100)]
mpegdemux: Merge some changes from the resindvd copy

One bug fix, some debug changes and other inconsequential changes
just to reduce the diff between the 2 copies a bit.

10 years agoaudiomixer: Fix and simplify overlap calculation
Sebastian Dröge [Wed, 15 Jan 2014 22:30:28 +0000 (23:30 +0100)]
audiomixer: Fix and simplify overlap calculation

10 years agoresindvdsrc: fix minor memory leak when creating title info message
Tim-Philipp Müller [Tue, 14 Jan 2014 23:55:53 +0000 (23:55 +0000)]
resindvdsrc: fix minor memory leak when creating title info message

10 years agoamcvideodec: Add the divx variant for mpeg4 video
Jorge Zapata [Wed, 9 Oct 2013 15:49:06 +0000 (17:49 +0200)]
amcvideodec: Add the divx variant for mpeg4 video

Looks like all the mpeg4 video decoders also accept the divx
variants. So we better add those caps too

https://bugzilla.gnome.org/show_bug.cgi?id=711163

10 years agoconfigure: fix (again) check for DB API: 5.4 is Not Recent Enough
Vincent Penquerc'h [Tue, 14 Jan 2014 17:08:36 +0000 (17:08 +0000)]
configure: fix (again) check for DB API: 5.4 is Not Recent Enough

Maybe testing the version is clearer, but testing for < 5 is not
enough, my version is 5.4 and does not yet have those new enums.

If you git blame to this and have a version > 5.4 that does not
either, please feel free to join along and bump the version.

10 years agoh265parser: Fix segfault when parsing VPS
duhui.lee [Tue, 14 Jan 2014 14:21:25 +0000 (23:21 +0900)]
h265parser: Fix segfault when parsing VPS

https://bugzilla.gnome.org/show_bug.cgi?id=722081

10 years agofaceblur: set maximum feature size to 0x0
Sebastian Dröge [Tue, 14 Jan 2014 09:38:37 +0000 (10:38 +0100)]
faceblur: set maximum feature size to 0x0

Previously faces would only be detected if they were at least 30x30 pixels
large and at most 32x32 pixels. We keep the minimum setting (maybe needs
a property as in facedetect) but disable the maximum feature size.

See https://bugzilla.gnome.org/show_bug.cgi?id=722158

10 years agofacedetect: set maximum feature size to 0x0
Kipp Cannon [Tue, 14 Jan 2014 06:06:02 +0000 (01:06 -0500)]
facedetect: set maximum feature size to 0x0

This disables the "max feature size" feature. The current configuration
is totally busted: The max feature size is hard-coded to 2 pixels more
than the user-supplied min feature size which pretty much means you need
to guess the size of the person's face to within a few pixels to get the
code to find it.

https://bugzilla.gnome.org/show_bug.cgi?id=722158

10 years agotsdemux: fix memleak of PCROffsetCurrent
Andrey Utkin [Tue, 14 Jan 2014 00:19:12 +0000 (02:19 +0200)]
tsdemux: fix memleak of PCROffsetCurrent

https://bugzilla.gnome.org/show_bug.cgi?id=722145

10 years agoh265parser: prevent to overrun chroma_weight_l0_flag
Justin Joy [Mon, 13 Jan 2014 12:50:16 +0000 (21:50 +0900)]
h265parser: prevent to overrun chroma_weight_l0_flag

The index of elements cannot exceed or equal to the size of elements.

https://bugzilla.gnome.org/show_bug.cgi?id=722101

10 years agoschroenc: set buffer offset to buffer timestamp
Vincent Penquerc'h [Mon, 13 Jan 2014 10:01:43 +0000 (10:01 +0000)]
schroenc: set buffer offset to buffer timestamp

oggmux expects this to order incoming buffers

https://bugzilla.gnome.org/show_bug.cgi?id=722078

10 years agodvb: Use DVB_API_VERSION to know if we have recent enough version
Edward Hervey [Fri, 10 Jan 2014 07:58:27 +0000 (08:58 +0100)]
dvb: Use DVB_API_VERSION to know if we have recent enough version

We end up with the same end-result but it's more comprehensible

10 years agoconfigure: check for good enough dvb kernel headers
Stefan Sauer [Thu, 9 Jan 2014 15:38:13 +0000 (16:38 +0100)]
configure: check for good enough dvb kernel headers

Recent changes are using new dvb api. Don't built the plugin, if we can't.
Fixes #721869