platform/upstream/gst-plugins-good.git
13 years agodeinterlace: analyse RFF fields in correct order
David Schleef [Tue, 30 Nov 2010 23:28:50 +0000 (15:28 -0800)]
deinterlace: analyse RFF fields in correct order

Code was repeating the second field, not the first.
Fixes: #636179.

13 years agortspsrc: handle stale digest authentication session data
Mark Nauwelaerts [Mon, 29 Nov 2010 14:32:40 +0000 (15:32 +0100)]
rtspsrc: handle stale digest authentication session data

In particular, handle Unauthorized server response when trying to convey
keep-alive.

Fixes #635532.

13 years agortph264depay: fix segfault on empty payload
Thijs Vermeir [Fri, 26 Nov 2010 14:00:29 +0000 (15:00 +0100)]
rtph264depay: fix segfault on empty payload

https://bugzilla.gnome.org/show_bug.cgi?id=635843

13 years ago0.10.25.5 pre-release
Tim-Philipp Müller [Thu, 18 Nov 2010 00:45:29 +0000 (00:45 +0000)]
0.10.25.5 pre-release

13 years agopo: update translations
Tim-Philipp Müller [Thu, 18 Nov 2010 00:44:45 +0000 (00:44 +0000)]
po: update translations

14 years agodeinterlace: fix reference leak
Tim-Philipp Müller [Sun, 14 Nov 2010 00:18:16 +0000 (00:18 +0000)]
deinterlace: fix reference leak

14 years agodeinterlace: Flush QoS and history before applying segment
Jan Schmidt [Fri, 12 Nov 2010 12:59:06 +0000 (23:59 +1100)]
deinterlace: Flush QoS and history before applying segment

When handling newsegment, flush out the buffer history in the
existing segment, not the new one. Fixes playback in some DVD
cases.

Partially fixes #633294

14 years agodeinterlace: improve event logging
Tim-Philipp Müller [Fri, 12 Nov 2010 12:20:16 +0000 (12:20 +0000)]
deinterlace: improve event logging

14 years agodeinterlace: Implement field history flushing
Robert Swain [Fri, 5 Nov 2010 16:00:15 +0000 (17:00 +0100)]
deinterlace: Implement field history flushing

In a number of cases it is necessary to flush the field history by
performing 'degraded' deinterlacing - that is, using the user-chosen
method for as many fields as possible, then using vfir for as long as
there are >= 2 fields remaining in the history, then using linear for
the last field.

This should avoid losing fields being kept for history for example at
EOS.

This may address part of #633294

14 years agodeinterlace: Refactor chain function
Robert Swain [Fri, 5 Nov 2010 14:44:35 +0000 (15:44 +0100)]
deinterlace: Refactor chain function

This is needed to be able to output a frame from outside the chain
function, i.e. in the following commit that adds flushing of the field
history.

14 years agoconfigure: we still require Gtk+ >= 2.14.0 when compiling against 2.0
Tim-Philipp Müller [Fri, 5 Nov 2010 17:17:56 +0000 (17:17 +0000)]
configure: we still require Gtk+ >= 2.14.0 when compiling against 2.0

The check for the minor version was dropped in the previous commit.

14 years agoconfigure: add --with-gtk option and default to Gtk+ 2.0 while the 3.0 API is still...
Tim-Philipp Müller [Fri, 5 Nov 2010 16:24:42 +0000 (16:24 +0000)]
configure: add --with-gtk option and default to Gtk+ 2.0 while the 3.0 API is still in flux

https://bugzilla.gnome.org/show_bug.cgi?id=634014

14 years agoicydemux: fix use-after-free of taglist
Jonathan Matthew [Thu, 4 Nov 2010 06:42:07 +0000 (16:42 +1000)]
icydemux: fix use-after-free of taglist

Broken by commit 4c2f5333 (bug #630205).

https://bugzilla.gnome.org/show_bug.cgi?id=633970

14 years ago0.10.25.4 pre-release
Tim-Philipp Müller [Mon, 1 Nov 2010 17:29:01 +0000 (17:29 +0000)]
0.10.25.4 pre-release

14 years agopo: update translations
Tim-Philipp Müller [Mon, 1 Nov 2010 17:28:36 +0000 (17:28 +0000)]
po: update translations

14 years agoconfigure: fix --disable-external
Tim-Philipp Müller [Mon, 1 Nov 2010 16:04:20 +0000 (16:04 +0000)]
configure: fix --disable-external

14 years agortph264depay: only set delta unit on all-non-key units
Wim Taymans [Mon, 1 Nov 2010 13:56:28 +0000 (14:56 +0100)]
rtph264depay: only set delta unit on all-non-key units

Only set the delta flag when all of the units in the packet are delta units.
Based on patch from Olivier Crête <olivier.crete@collabora.co.uk>

Fixes #632945

14 years agogoom: Return not-negotiated when bps is unknown
Thiago Santos [Tue, 26 Oct 2010 18:44:37 +0000 (15:44 -0300)]
goom: Return not-negotiated when bps is unknown

If caps weren't negotiated, goom should return not-negotiated
from its chain functions instead of using bps unitialized, which
leads to a division by 0

https://bugzilla.gnome.org/show_bug.cgi?id=633212

14 years agoAutomatic update of common submodule
Jan Schmidt [Wed, 27 Oct 2010 12:16:54 +0000 (13:16 +0100)]
Automatic update of common submodule

From 7bbd708 to 011bcc8

14 years agovideoflip: Forward src pad events upstream.
Jan Schmidt [Tue, 26 Oct 2010 15:54:11 +0000 (16:54 +0100)]
videoflip: Forward src pad events upstream.

Fix passing navigation and other events upstream by actually sending them.

Fixes: #633205

14 years agoqtdemux: fix deadlock in error code path
Tim-Philipp Müller [Sun, 24 Oct 2010 17:50:30 +0000 (18:50 +0100)]
qtdemux: fix deadlock in error code path

GST_ELEMENT_ERROR must not be called with the object lock held,
since it will call gst_object_get_parent() internally, which
takes the object lock as well.

14 years agomatroskademux: Remove useless clearing of send_xiph_headers for Dirac
Philip Jägenstedt [Wed, 20 Oct 2010 08:21:48 +0000 (10:21 +0200)]
matroskademux: Remove useless clearing of send_xiph_headers for Dirac

This looks like a mistake when copy-pasting the Theora code.

https://bugzilla.gnome.org/show_bug.cgi?id=632682

14 years agomatroskademux: don't crash if vorbis/theora codec data is missing
Philip Jägenstedt [Wed, 20 Oct 2010 11:28:28 +0000 (13:28 +0200)]
matroskademux: don't crash if vorbis/theora codec data is missing

Error out properly in this case instead of crashing.

https://bugzilla.gnome.org/show_bug.cgi?id=632682

14 years ago0.10.25.3 pre-release
Tim-Philipp Müller [Fri, 22 Oct 2010 17:11:46 +0000 (18:11 +0100)]
0.10.25.3 pre-release

14 years agortspsrc: fix duration reporting
Mark Nauwelaerts [Tue, 19 Oct 2010 14:45:51 +0000 (16:45 +0200)]
rtspsrc: fix duration reporting

Init segment prior to storing duration info in it.

Fixes #632548.

14 years agogconf: Don't install schemas when GConf is disabled
Bastien Nocera [Tue, 19 Oct 2010 13:21:53 +0000 (14:21 +0100)]
gconf: Don't install schemas when GConf is disabled

https://bugzilla.gnome.org/show_bug.cgi?id=632553

14 years agowin32: set GST_PACKAGE_RELEASE_DATETIME also in win32 config.h
Tim-Philipp Müller [Sat, 16 Oct 2010 14:43:53 +0000 (15:43 +0100)]
win32: set GST_PACKAGE_RELEASE_DATETIME also in win32 config.h

14 years ago0.10.25.2 pre-release
Tim-Philipp Müller [Sat, 16 Oct 2010 00:33:52 +0000 (01:33 +0100)]
0.10.25.2 pre-release

14 years agopo: update translations
Tim-Philipp Müller [Sat, 16 Oct 2010 00:26:01 +0000 (01:26 +0100)]
po: update translations

14 years agotests: Don't dist generated orc files
David Schleef [Fri, 15 Oct 2010 20:22:03 +0000 (13:22 -0700)]
tests: Don't dist generated orc files

14 years agoUpdate generated orc code
David Schleef [Fri, 15 Oct 2010 21:02:19 +0000 (14:02 -0700)]
Update generated orc code

14 years agoconfigure: bump Orc requirement to 0.4.11
Tim-Philipp Müller [Fri, 15 Oct 2010 17:00:10 +0000 (18:00 +0100)]
configure: bump Orc requirement to 0.4.11

14 years agortpbin: Use the right constant to define the "use-pipeline-clock" property
Olivier Crête [Thu, 14 Oct 2010 21:41:30 +0000 (17:41 -0400)]
rtpbin: Use the right constant to define the "use-pipeline-clock" property

The wrong #define was being used, now use the correct one.

14 years agoAutomatic update of common submodule
David Schleef [Thu, 14 Oct 2010 19:31:48 +0000 (12:31 -0700)]
Automatic update of common submodule

From 5a668bf to 7bbd708

14 years agoac3: demuxers provide framed output
Mark Nauwelaerts [Thu, 14 Oct 2010 15:26:14 +0000 (17:26 +0200)]
ac3: demuxers provide framed output

14 years agomatroskamux: reduce newsegment event spam and set discont flag where needed
Tim-Philipp Müller [Wed, 13 Oct 2010 23:11:27 +0000 (00:11 +0100)]
matroskamux: reduce newsegment event spam and set discont flag where needed

Only send newsegment events with new positions downstream when actually
needed, instead of sending multiple newsegment events with new seek
positions in a row. Also set the discont flag on buffers after a
discontinuity.

14 years agomatroskamux: set correct buffer offsets after seeks
Tim-Philipp Müller [Wed, 13 Oct 2010 22:46:02 +0000 (23:46 +0100)]
matroskamux: set correct buffer offsets after seeks

Re-use the existing 'pos' field maintained by ebml writer to set
buffer offsets. This also makes sure that we set the right offsets
on buffers after a seek (e.g. when writing an index at the end).

14 years agomatroskamux: don't forward tag events downstream
Tim-Philipp Müller [Wed, 13 Oct 2010 23:22:03 +0000 (00:22 +0100)]
matroskamux: don't forward tag events downstream

Don't forward stream-specific tag events downstream (esp. not
before any newsegment event).x

14 years agoqtdemux: handle another mp4v variation
Mark Nauwelaerts [Wed, 13 Oct 2010 15:15:25 +0000 (17:15 +0200)]
qtdemux: handle another mp4v variation

... including the glbl atom containing codec-data.

14 years agovarious (gst): add a missing G_PARAM_STATIC_STRINGS flags
Stefan Kost [Wed, 13 Oct 2010 14:21:23 +0000 (17:21 +0300)]
various (gst): add a missing G_PARAM_STATIC_STRINGS flags

14 years agovarious (sys): add a missing G_PARAM_STATIC_STRINGS flags
Stefan Kost [Wed, 13 Oct 2010 14:13:04 +0000 (17:13 +0300)]
various (sys): add a missing G_PARAM_STATIC_STRINGS flags

14 years agovarious (ext): add a missing G_PARAM_STATIC_STRINGS flags
Stefan Kost [Wed, 13 Oct 2010 13:25:15 +0000 (16:25 +0300)]
various (ext): add a missing G_PARAM_STATIC_STRINGS flags

14 years agovarious: wrap property registration and add a single fixme for long desc.
Stefan Kost [Wed, 13 Oct 2010 13:34:09 +0000 (16:34 +0300)]
various: wrap property registration and add a single fixme for long desc.

14 years agoh264depay: always mark the codec_data as keyframe
Wim Taymans [Wed, 13 Oct 2010 09:46:58 +0000 (11:46 +0200)]
h264depay: always mark the codec_data as keyframe

We need to mark the codec_data as a keyframe or else downstream decoders might
decide to skip it, waiting for a keyframe.

Fixes #631996

14 years agomatroskamux: make buffer offsets a byte count rather than a buffer count
Zaheer Abbas Merali [Wed, 13 Oct 2010 06:16:47 +0000 (07:16 +0100)]
matroskamux: make buffer offsets a byte count rather than a buffer count

14 years agoext, gst: canonicalise property names where this wasn't the case
Tim-Philipp Müller [Thu, 7 Oct 2010 20:12:48 +0000 (21:12 +0100)]
ext, gst: canonicalise property names where this wasn't the case

ie. "foo_bar" -> "foo-bar"

14 years agortpmpvpay: fix timestamping of rtp buffers
Thijs Vermeir [Tue, 12 Oct 2010 13:02:42 +0000 (15:02 +0200)]
rtpmpvpay: fix timestamping of rtp buffers

Incomming buffer is only pushed on the adapter at the end of the
handle_buffer function. But duration/timestamp of this buffer is already
taken into account for the current data in the adapter. This leads to
wrong rtp timestamps and extra latency.

14 years agoexamples: Fix build with GTK+ 3.0
Sebastian Dröge [Tue, 12 Oct 2010 09:37:40 +0000 (11:37 +0200)]
examples: Fix build with GTK+ 3.0

14 years agortspsrc: mark as a source
Wim Taymans [Mon, 11 Oct 2010 13:12:00 +0000 (15:12 +0200)]
rtspsrc: mark as a source

Mark the rtspsrc element as a source.
Requires 0.10.31.1 now

14 years agoautodetect: Set GST_ELEMENT_IS_SOURCE flag on sources
Sebastian Dröge [Mon, 11 Oct 2010 12:24:13 +0000 (14:24 +0200)]
autodetect: Set GST_ELEMENT_IS_SOURCE flag on sources

14 years agoswitchsrc: Set the GST_ELEMENT_IS_SOURCE flag
Sebastian Dröge [Mon, 11 Oct 2010 12:21:07 +0000 (14:21 +0200)]
switchsrc: Set the GST_ELEMENT_IS_SOURCE flag

14 years agoconfigure: Require core 0.10.30.1
Sebastian Dröge [Mon, 11 Oct 2010 12:17:33 +0000 (14:17 +0200)]
configure: Require core 0.10.30.1

14 years agomatroskamux: set offsets on outgoing buffers
Zaheer Abbas Merali [Sun, 10 Oct 2010 13:43:58 +0000 (14:43 +0100)]
matroskamux: set offsets on outgoing buffers

14 years agov4l2sink: Only get/set overlay params if needed
IOhannes m zmölnig [Sat, 9 Oct 2010 12:14:27 +0000 (14:14 +0200)]
v4l2sink: Only get/set overlay params if needed

it's perfectly ok for a video output device to not have overlay capabilities.
this patch removes the need to get/set the overlay parameters if the user
does not explicitely request one of the overlay properties

14 years agov4l2sink: Protect against NULL-pointer access
IOhannes m zmölnig [Thu, 30 Sep 2010 13:28:23 +0000 (15:28 +0200)]
v4l2sink: Protect against NULL-pointer access

gst_v4l2sink_change_state() would free the pool without checking whether there
was a valid pool...

14 years agoAutomatic update of common submodule
David Schleef [Fri, 8 Oct 2010 19:43:51 +0000 (12:43 -0700)]
Automatic update of common submodule

From c4a8adc to 5a668bf

14 years agoAutomatic update of common submodule
Sebastian Dröge [Fri, 8 Oct 2010 10:53:33 +0000 (12:53 +0200)]
Automatic update of common submodule

From 5e3c9bf to c4a8adc

14 years agodeinterlace: Fix required fields logic
Robert Swain [Wed, 6 Oct 2010 09:29:55 +0000 (11:29 +0200)]
deinterlace: Fix required fields logic

Both history_count and fields_required count from 1. As per the while loop
condition that follows this code, to perform the deinterlacing method, we need
history_count >= fields_required fields in the history. Therefore if we have
history_count < fields_required (not fields_required + 1), we need more fields.

14 years agoflvmux: resend onMetada tag when tags changes in streamable mode
Andoni Morales Alastruey [Mon, 20 Sep 2010 17:43:45 +0000 (19:43 +0200)]
flvmux: resend onMetada tag when tags changes in streamable mode

14 years agoqtdemux: AAC codec_data can be > 2 bytes long
Arun Raghavan [Tue, 5 Oct 2010 18:40:50 +0000 (19:40 +0100)]
qtdemux: AAC codec_data can be > 2 bytes long

This fixes the assumption that DecoderSpecificInfo must be 2 bytes long
for AAC files. The specification allows HE-AAC to be explicitly
signalled in a backward compatible way. This is done by means of an
additional information after the regular AAC header. It is expected that
decoders that can play AAC but not HE-AAC will parse the header normally
and ignore extended bits, much as they do for the HE-AAC specific payload
in the actual stream.

https://bugzilla.gnome.org/show_bug.cgi?id=612313

14 years agomatroskademux: only unref buffer when no longer needed for cluster scanning
Mark Nauwelaerts [Tue, 5 Oct 2010 14:01:19 +0000 (16:01 +0200)]
matroskademux: only unref buffer when no longer needed for cluster scanning

Fixes #629047.

14 years agomatroskademux: avoid infinite cluster scanning
Mark Nauwelaerts [Tue, 5 Oct 2010 14:00:45 +0000 (16:00 +0200)]
matroskademux: avoid infinite cluster scanning

14 years agogoom: take duration into account when doing QoS
Wim Taymans [Tue, 5 Oct 2010 10:20:52 +0000 (12:20 +0200)]
goom: take duration into account when doing QoS

Take the duration of the frames into account so that we don't drop frames that
are only partially past the QoS deadline.

14 years agogoom: use adapter for timestamping
Wim Taymans [Tue, 5 Oct 2010 08:40:15 +0000 (10:40 +0200)]
goom: use adapter for timestamping

Use the adapter timestamp code to get more accurate timestamps.
Fix latency calculation, we add our own latency in the worst case.

14 years agoraw1394: Don't compile hdv1394src if libiec61883 isn't available
Edward Hervey [Mon, 4 Oct 2010 20:31:32 +0000 (22:31 +0200)]
raw1394: Don't compile hdv1394src if libiec61883 isn't available

Fixes #629896

14 years agoicydemux: forward tag events
Andoni Morales Alastruey [Mon, 20 Sep 2010 17:44:09 +0000 (19:44 +0200)]
icydemux: forward tag events

https://bugzilla.gnome.org/show_bug.cgi?id=630205

14 years agogoom2k1: report our latency correctly
Wim Taymans [Mon, 4 Oct 2010 17:00:45 +0000 (19:00 +0200)]
goom2k1: report our latency correctly

Fixes #631303

14 years agogoom2k1: add defines for default width/height/fps
Wim Taymans [Mon, 4 Oct 2010 16:56:15 +0000 (18:56 +0200)]
goom2k1: add defines for default width/height/fps

Add some defines for the default width/height/fps instead of using different
values in different places.

14 years agogoom: add latency compensation code.
Wim Taymans [Mon, 4 Oct 2010 16:52:14 +0000 (18:52 +0200)]
goom: add latency compensation code.

Implement a latency query and report how much latency we will add to the
stream.
Alse make some defaults for the default width/height/framerate

Fixes #631303

14 years agotest: add python version of the audio sender
Wim Taymans [Mon, 4 Oct 2010 15:56:57 +0000 (17:56 +0200)]
test: add python version of the audio sender

Add a python version of the audio sender pipeline.

Ported by Sp4rc on IRC.

14 years agotests: Add python RTP client example
Wim Taymans [Mon, 4 Oct 2010 15:52:22 +0000 (17:52 +0200)]
tests: Add python RTP client example

Add a python version of the PCMA client app.

Ported by Sp4rc on IRC.

14 years agortp: Fix unitialized compiler warnings on OS X build bot
Sebastian Dröge [Mon, 4 Oct 2010 07:39:59 +0000 (09:39 +0200)]
rtp: Fix unitialized compiler warnings on OS X build bot

These warnings are wrong though, the variables are only used in
the cases where they *are* initialized by the bit reader.

14 years agortpg722pay: Fix uninitialized variable compiler warning
Sebastian Dröge [Sun, 3 Oct 2010 21:49:08 +0000 (23:49 +0200)]
rtpg722pay: Fix uninitialized variable compiler warning

The clock rate is always 8000 Hz according to the RFC and
the sampling rate must always be 16000 Hz.

14 years agortpjitterbuffer: improve article reference in comment block
Vladimir Eremeev [Fri, 1 Oct 2010 09:59:10 +0000 (13:59 +0400)]
rtpjitterbuffer: improve article reference in comment block

https://bugzilla.gnome.org/show_bug.cgi?id=631082

14 years agoqtdemux: Use pbutils for H.264 profile/level extraction
Arun Raghavan [Fri, 30 Apr 2010 15:30:31 +0000 (21:00 +0530)]
qtdemux: Use pbutils for H.264 profile/level extraction

The functions used to extract this data have been moved to gstpbutils to
facilitate reuse.

https://bugzilla.gnome.org/show_bug.cgi?id=617318

14 years agomatroskademux: Use pbutils for H.264 profile/level extraction
Arun Raghavan [Fri, 30 Apr 2010 15:30:31 +0000 (21:00 +0530)]
matroskademux: Use pbutils for H.264 profile/level extraction

The functions used to extract this data have been moved to gstpbutils to
facilitate reuse.

https://bugzilla.gnome.org/show_bug.cgi?id=617318

14 years agoqtdemux: Export MPEG-4 video profile and level in stream caps
Arun Raghavan [Thu, 22 Apr 2010 14:09:47 +0000 (19:39 +0530)]
qtdemux: Export MPEG-4 video profile and level in stream caps

This uses gstpbutils to extract the profile and level from the video
object sequence and adds this to stream caps. This can be used as
metadata and for fine-grained decoder selection.

https://bugzilla.gnome.org/show_bug.cgi?id=616521

14 years agoqtdemux: fix aac channel override based on codec data for 7.1 case
Tim-Philipp Müller [Thu, 30 Sep 2010 11:44:52 +0000 (12:44 +0100)]
qtdemux: fix aac channel override based on codec data for 7.1 case

14 years agoqtdemux: Export AAC profile and level in caps
Arun Raghavan [Fri, 30 Apr 2010 08:36:27 +0000 (14:06 +0530)]
qtdemux: Export AAC profile and level in caps

This exports the AAC profile and level in caps for use as metadata and
(eventually) for more fine-grained selection of decoders at
caps-negotiation time. (Doesn't work for HE-AAC yet though.)

https://bugzilla.gnome.org/show_bug.cgi?id=612313

14 years agortp: add G722 pay and depayloader
Wim Taymans [Thu, 30 Sep 2010 16:34:04 +0000 (18:34 +0200)]
rtp: add G722 pay and depayloader

14 years agortpjitterbuffer: update link to documentation
Thijs Vermeir [Thu, 30 Sep 2010 10:08:49 +0000 (12:08 +0200)]
rtpjitterbuffer: update link to documentation

14 years agoexamples: fix indentation on rtp client example
Thijs Vermeir [Thu, 30 Sep 2010 09:34:56 +0000 (11:34 +0200)]
examples: fix indentation on rtp client example

14 years agoexamples: fix typo in port of rtp examples
Thijs Vermeir [Thu, 30 Sep 2010 09:33:24 +0000 (11:33 +0200)]
examples: fix typo in port of rtp examples

14 years agowavenc: miniscule code clean-up
Tim-Philipp Müller [Wed, 29 Sep 2010 12:20:22 +0000 (13:20 +0100)]
wavenc: miniscule code clean-up

GST_CLOCK_TIME_NONE is not something that should be used in connection with
GST_FORMAT_BYTES.

14 years agoavidemux: reverse playback; prevent overlap of subsequent fragments
Mark Nauwelaerts [Wed, 29 Sep 2010 08:34:36 +0000 (10:34 +0200)]
avidemux: reverse playback; prevent overlap of subsequent fragments

14 years agortspsrc: fix missing null-terminator in protocols array
René Stadler [Tue, 28 Sep 2010 13:21:48 +0000 (16:21 +0300)]
rtspsrc: fix missing null-terminator in protocols array

Fixes random crash regression from commit ae84ae.

14 years agortspsrc: don't add /UDP in the transport, it's the default
Wim Taymans [Fri, 24 Sep 2010 14:26:20 +0000 (16:26 +0200)]
rtspsrc: don't add /UDP in the transport, it's the default

don't add the default UDP lower-transport, some servers don't seem to like it.

Fixes #630500

14 years agortpmanager: packet lost should not be a warning. It happens all the time...
Pascal Buhler [Fri, 25 Jun 2010 15:08:03 +0000 (17:08 +0200)]
rtpmanager: packet lost should not be a warning. It happens all the time...

14 years agortpbin: Make cleaning up sources in rtp_session_on_timeout MT safe
Pascal Buhler [Fri, 24 Sep 2010 13:33:40 +0000 (15:33 +0200)]
rtpbin: Make cleaning up sources in rtp_session_on_timeout MT safe

Using _foreach_remove on the hashtable, while releasing the lock protecting
that table inside the callback is not a good idea. The hashtable might
then change (a source removed or added) while signals like on_timeout
are being sent.

This solution makes a copy of the table, performs the _foreach without
actually removing any sources, but marks them for removal on a second
iteration with the real list, but this time not letting go of the lock.

Fixes #630452

14 years agoid3demux: Sanitize id3 frame names
Edward Hervey [Fri, 24 Sep 2010 13:19:15 +0000 (15:19 +0200)]
id3demux: Sanitize id3 frame names

This is similar to what is done in qtdemux. Avoids providing invalid
structure/tags names

14 years agoapedemux: Skip empty tags
Edward Hervey [Fri, 24 Sep 2010 12:59:45 +0000 (14:59 +0200)]
apedemux: Skip empty tags

Avoid creating bogus string tags. Also added logging of the string
values of the tag name and value.

14 years agosoup: init debug category before using it
Tim-Philipp Müller [Fri, 24 Sep 2010 07:56:36 +0000 (08:56 +0100)]
soup: init debug category before using it

14 years agortpbin: Handle rysnc of iterator when looking for free pad name
Pascal Buhler [Mon, 12 Apr 2010 07:49:14 +0000 (09:49 +0200)]
rtpbin: Handle rysnc of iterator when looking for free pad name

If a new pad was added while iterating then a pad could be
returned that was already in use.

Fixes #630451

14 years agortpsession: fix compilation
Wim Taymans [Fri, 24 Sep 2010 12:09:12 +0000 (14:09 +0200)]
rtpsession: fix compilation

14 years agortpbin: Unlock before adding pad in new_payload_found
Trond Andersen [Wed, 7 Apr 2010 13:31:52 +0000 (15:31 +0200)]
rtpbin: Unlock before adding pad in new_payload_found

Holding internal locks while potentially calling out is a source
of deadlocks, and in this case the application might subscribe to the
pad-added signal.

Fixes #630449

14 years agortpsession: relax third-party collision detection
Havard Graff [Mon, 31 Aug 2009 16:37:40 +0000 (18:37 +0200)]
rtpsession: relax third-party collision detection

If the source has been inactive for some time, we assume that it has
simply changed its transport source address. Hence, there is no true
third-party collision - only a simulated one.

Fixes #630447

14 years agortpsource: whitespace fixes
Wim Taymans [Fri, 24 Sep 2010 11:50:02 +0000 (13:50 +0200)]
rtpsource: whitespace fixes

14 years agortpsource: simplify the rate estimation some more
Wim Taymans [Fri, 24 Sep 2010 11:48:50 +0000 (13:48 +0200)]
rtpsource: simplify the rate estimation some more

14 years agortpmanager: provide additional statistics
Havard Graff [Mon, 31 Aug 2009 16:34:08 +0000 (18:34 +0200)]
rtpmanager: provide additional statistics

14 years agoconfigure: set plugin release datetime
Tim-Philipp Müller [Thu, 23 Sep 2010 23:01:05 +0000 (00:01 +0100)]
configure: set plugin release datetime