platform/upstream/gst-plugins-good.git
13 years agoflacparse: tell baseparse the duration in samples for better accuracy
Tim-Philipp Müller [Mon, 23 May 2011 12:50:46 +0000 (13:50 +0100)]
flacparse: tell baseparse the duration in samples for better accuracy

Tell GstBaseParse the duration in samples instead of time, so that
a duration query in DEFAULT format will return the correct number
of samples without rounding errors. Baseparse will convert this
into time itself when needed.

https://bugzilla.gnome.org/show_bug.cgi?id=650785

13 years agoflacdec: also try upstream first for duration query in DEFAULT format
Tim-Philipp Müller [Mon, 23 May 2011 12:25:44 +0000 (13:25 +0100)]
flacdec: also try upstream first for duration query in DEFAULT format

https://bugzilla.gnome.org/show_bug.cgi?id=650785

13 years agoflacparse: make conversion from TIME to DEFAULT format (samples) work
Tim-Philipp Müller [Mon, 23 May 2011 12:23:21 +0000 (13:23 +0100)]
flacparse: make conversion from TIME to DEFAULT format (samples) work

Fix copy'n'paste error in the previous commit.

13 years agoflacparse: Implement conversions between TIME and DEFAULT format
Sebastian Dröge [Mon, 23 May 2011 09:36:36 +0000 (11:36 +0200)]
flacparse: Implement conversions between TIME and DEFAULT format

Fixes bug #650785.

13 years agoflacparse: don't error out on invalid minimum_blocksize value in streaminfo header
Tim-Philipp Müller [Sun, 22 May 2011 17:50:51 +0000 (18:50 +0100)]
flacparse: don't error out on invalid minimum_blocksize value in streaminfo header

We don't use it, so may just as well accept an invalid value
of 0 here, which is likely inconsequential anyway.

https://bugzilla.gnome.org/show_bug.cgi?id=650691

13 years agortp: fix static array overruns in a nicer way
Stefan Kost [Fri, 20 May 2011 07:34:47 +0000 (10:34 +0300)]
rtp: fix static array overruns in a nicer way

Use G_N_ELEMENTS instead of hard-coding the array size.

13 years agortp: fix static array overruns
Stefan Kost [Thu, 19 May 2011 21:53:44 +0000 (00:53 +0300)]
rtp: fix static array overruns

Yes array[10] has elements from 0...9.

13 years agodocs: update plugin introspection data
Stefan Kost [Thu, 19 May 2011 20:31:19 +0000 (23:31 +0300)]
docs: update plugin introspection data

Now more files are merged and produced in a canonical fashion, which hopefully
creates less or no delta in the future.

13 years agoAutomatic update of common submodule
Stefan Kost [Thu, 19 May 2011 19:57:15 +0000 (22:57 +0300)]
Automatic update of common submodule

From 9e5bbd5 to 69b981f

13 years agoqtdemux: add missing break
Stefan Kost [Thu, 19 May 2011 15:21:33 +0000 (18:21 +0300)]
qtdemux: add missing break

13 years agodeinterlace: Add support for deinterlacing using buffer caps/flags
Robert Swain [Mon, 8 Nov 2010 13:06:15 +0000 (14:06 +0100)]
deinterlace: Add support for deinterlacing using buffer caps/flags

When not using the fieldanalysis element immediately upstream of deinterlace,
behaviour should remain unchanged. fieldanalysis will set the caps and flags on
the buffers such that they can be interpreted and acted upon to produce
progressive output.

There are two main modes of operation:

- Passive pattern locking
  Passive pattern locking is a non-blocking, low-latency mode of operation that
  is suitable for close-to-live usage. Initially a telecine stream will be
  output as variable framerate with naïve timestamp adjustment. With each
  incoming buffer, an attempt is made to lock onto a pattern. When a lock is
  obtained, the src pad and output buffer caps will reflect the pattern and
  timestamps will be accurately interpolated between pattern repeats. This
  means that initially and at pattern transitions there will be short periods
  of inaccurate timestamping.

- Active pattern locking
  Active pattern locking is a blocking, high-latency mode of operation that is
  targeted at use-cases where timestamp accuracy is paramount. Buffers will be
  queued until enough are present to make a lock. When locked, timestamps will
  be accurately interpolated between pattern repeats. Orphan fields can be
  dropped or deinterlaced. If no lock can be obtained, a single field might be
  pushed through to be deinterlaced.

Locking can also be disabled or 'auto' chooses between passive and active
locking modes depending on whether upstream is live.

13 years agoconfigure: Remove config script check for caca
David Schleef [Tue, 10 May 2011 23:25:40 +0000 (16:25 -0700)]
configure: Remove config script check for caca

13 years agortp: Fix segmentation fault processing payload buffers
Jose Antonio Santos Cadenas [Wed, 18 May 2011 10:36:40 +0000 (12:36 +0200)]
rtp: Fix segmentation fault processing payload buffers

This commit checks if the value returned by
gst_rtp_buffer_get_payload_buffer and
gst_rtp_buffer_get_payload_subbuffer is NULL before using it.

13 years agoAutomatic update of common submodule
Stefan Kost [Wed, 18 May 2011 13:10:07 +0000 (16:10 +0300)]
Automatic update of common submodule

From fd35073 to 9e5bbd5

13 years agoavidemux: ensure 0-padding when correcting dubious list size
Mark Nauwelaerts [Wed, 18 May 2011 10:52:31 +0000 (12:52 +0200)]
avidemux: ensure 0-padding when correcting dubious list size

13 years agoAutomatic update of common submodule
Stefan Kost [Wed, 18 May 2011 09:24:25 +0000 (12:24 +0300)]
Automatic update of common submodule

From 46dfcea to fd35073

13 years agortspsrc: use EINVAL for missing url parameter
Stefan Kost [Wed, 18 May 2011 07:22:27 +0000 (10:22 +0300)]
rtspsrc: use EINVAL for missing url parameter

Fixes gcc warning about using uninitialized variable 'res'.

13 years agovarious: fix author tag in element details
Stefan Kost [Thu, 28 Apr 2011 12:37:40 +0000 (15:37 +0300)]
various: fix author tag in element details

13 years agoauparse: implement seeking
Chris E Jones [Wed, 20 Apr 2011 19:25:58 +0000 (15:25 -0400)]
auparse: implement seeking

Implement seeking and seeking query. Fixes #644512

13 years agortspsrc: also allow PAUSE to be interrupted
Mark Nauwelaerts [Wed, 6 Apr 2011 14:05:55 +0000 (16:05 +0200)]
rtspsrc: also allow PAUSE to be interrupted

... as it is on the way out to NULL.

See #632504.

13 years agortspsrc: ensure proper closing and cleanup
Mark Nauwelaerts [Wed, 6 Apr 2011 13:51:49 +0000 (15:51 +0200)]
rtspsrc: ensure proper closing and cleanup

... since the TEARDOWN sequence might not have had a chance to even start,
but at least connections should be closed (synchronously) and state cleaned up.

See #632504.

13 years agortspsrc: fix and improve async handling
Mark Nauwelaerts [Wed, 6 Apr 2011 13:49:01 +0000 (15:49 +0200)]
rtspsrc: fix and improve async handling

Simplify the command handling; passing a command to thread means we really
want it to get the message, which means to always flush provided the command
can handle being interrupted.  Command thread indicates whether command
allows interruption and ensure non-flushing connection as it subsequently
needs it.

In particular, this also makes the TEARDOWN sequence interruptable
and also prevents races where _loop_ could miss a command and would
continue receiving (or at least trying to).

See #632504.

13 years agortspsrc: tweak post-seek loop handling
Mark Nauwelaerts [Wed, 6 Apr 2011 12:53:27 +0000 (14:53 +0200)]
rtspsrc: tweak post-seek loop handling

13 years agortspsrc: open on play and pause when not done yet
Wim Taymans [Mon, 10 Jan 2011 11:46:37 +0000 (12:46 +0100)]
rtspsrc: open on play and pause when not done yet

With the async state changes, it is possible that we need to open the stream
before play and pause.
Also make sure we remember a previous open failure so that we don't keep trying
again.

13 years agortspsrc: improve async handling
Wim Taymans [Mon, 10 Jan 2011 10:45:03 +0000 (11:45 +0100)]
rtspsrc: improve async handling

Simplify the command handling, only continue looping when we have not received
another command or when the previous loop was successfull.
Avoid looping on a disconnected socket.

13 years agortspsrc: rework reconnect code
Wim Taymans [Fri, 7 Jan 2011 17:02:49 +0000 (18:02 +0100)]
rtspsrc: rework reconnect code

Use the same async code path to implement reconnects.
Make sure we only post progress messages when doing async things.

13 years agortspsrc: small cleanups
Wim Taymans [Fri, 7 Jan 2011 16:19:59 +0000 (17:19 +0100)]
rtspsrc: small cleanups

Make sure we cancel the previous task when queuing a new one.
Move the messages to a central place so we can more easily post them.

13 years agortspsrc: don't post errors when interrupting
Wim Taymans [Fri, 7 Jan 2011 14:15:49 +0000 (15:15 +0100)]
rtspsrc: don't post errors when interrupting

13 years agortspsrc: implement more async handling
Wim Taymans [Fri, 7 Jan 2011 12:43:06 +0000 (13:43 +0100)]
rtspsrc: implement more async handling

Remove some old locks.
Make sure we never go into the loop function when flushing.

13 years agortspsrc: first attempt at async implementation
Wim Taymans [Fri, 7 Jan 2011 10:40:32 +0000 (11:40 +0100)]
rtspsrc: first attempt at async implementation

13 years agortspsrc: small header cleanups
Wim Taymans [Fri, 7 Jan 2011 10:40:11 +0000 (11:40 +0100)]
rtspsrc: small header cleanups

13 years agossrcdemux: Fix uninitialized variable compiler warning for (pre-) releases too
Sebastian Dröge [Tue, 17 May 2011 08:47:32 +0000 (10:47 +0200)]
ssrcdemux: Fix uninitialized variable compiler warning for (pre-) releases too

13 years agov4l2objects: Only allow mpeg-ts on source objects
Edward Hervey [Thu, 28 Apr 2011 13:57:04 +0000 (15:57 +0200)]
v4l2objects: Only allow mpeg-ts on source objects

Ugly fix for #648312

13 years agortpssrcdemux: Fix uninitialized variable compiler warning
Sebastian Dröge [Tue, 17 May 2011 07:24:08 +0000 (09:24 +0200)]
rtpssrcdemux: Fix uninitialized variable compiler warning

13 years agossrcdemux: Implement iterate internal links for sink pads
Olivier Crête [Fri, 6 May 2011 23:09:17 +0000 (19:09 -0400)]
ssrcdemux: Implement iterate internal links for sink pads

https://bugzilla.gnome.org/show_bug.cgi?id=649617

13 years agortpssrcdemux: iterate pad function is only valid for src pads
Olivier Crête [Fri, 6 May 2011 22:41:01 +0000 (18:41 -0400)]
rtpssrcdemux: iterate pad function is only valid for src pads

The iterate function is only used for src pads, so mark it as such and remove
dead code.

https://bugzilla.gnome.org/show_bug.cgi?id=649617

13 years agortpssrcdemux: Release lock before emitting signal
Olivier Crête [Fri, 6 May 2011 22:12:53 +0000 (18:12 -0400)]
rtpssrcdemux: Release lock before emitting signal

If the lock is not released before emitting a signal, it may cause a deadlock
if any other function in the element is called.

Also removed an unused timestamp parameter

https://bugzilla.gnome.org/show_bug.cgi?id=649617

13 years agomatroskaparse: calculate segment duration after parsing all the IDs
Debarshi Ray [Sun, 15 May 2011 20:25:15 +0000 (23:25 +0300)]
matroskaparse: calculate segment duration after parsing all the IDs

Since the segment duration is given in terms of the
GST_MATROSKA_ID_TIMECODESCALE we should only convert it into
nanoseconds when we are sure that any scale specified in the file has
been read.

https://bugzilla.gnome.org/show_bug.cgi?id=650258

13 years agomatroskademux: additional lock safety
Mark Nauwelaerts [Wed, 4 May 2011 09:55:21 +0000 (11:55 +0200)]
matroskademux: additional lock safety

Fixes #619590.

13 years agoqtdemux: also check for bitrate info in caps
Mark Nauwelaerts [Tue, 26 Apr 2011 14:06:56 +0000 (16:06 +0200)]
qtdemux: also check for bitrate info in caps

13 years agoqtdemux: guess bitrate if only one stream's bitrate is unknown
Arun Raghavan [Mon, 24 May 2010 19:34:43 +0000 (01:04 +0530)]
qtdemux: guess bitrate if only one stream's bitrate is unknown

If the bitrates for all but one audio/video streams are known, and the
total stream size and duration can be determined, this calculates the
unkown bitrate as (stream size / duration) - (sum of known bitrates).
While this is not guaranteed to be very accurate, it should be good
enough for most purposes.

For example, this is useful for H.263 + AAC streams where no 'btrt' atom
is available for the video portion.

https://bugzilla.gnome.org/show_bug.cgi?id=619548

13 years agoqtdemux: Export max bitrate for AMR-NB/-WB streams
Arun Raghavan [Mon, 31 May 2010 18:29:59 +0000 (23:59 +0530)]
qtdemux: Export max bitrate for AMR-NB/-WB streams

This parses the 'damr' atom if present, and exports the maximum bitrate
of the stream using the mode set field to determine the highest bitrate
frame type that might be present.

https://bugzilla.gnome.org/show_bug.cgi?id=620186

13 years agopulse: Define PATH_MAX if it isn't defined
Pino Toscano [Mon, 16 May 2011 07:04:31 +0000 (09:04 +0200)]
pulse: Define PATH_MAX if it isn't defined

GNU Hurd for example doesn't define it.

13 years agomatroskademux: calculate segment duration after parsing all the IDs
Debarshi Ray [Sun, 15 May 2011 20:25:15 +0000 (23:25 +0300)]
matroskademux: calculate segment duration after parsing all the IDs

Since the segment duration is given in terms of the
GST_MATROSKA_ID_TIMECODESCALE we should only convert it into
nanoseconds when we are sure that any scale specified in the file has
been read.

https://bugzilla.gnome.org/show_bug.cgi?id=650258

13 years agoflvmux: Add support for mpegversion 2, which is also AAC
Andoni Morales Alastruey [Mon, 9 May 2011 17:00:45 +0000 (19:00 +0200)]
flvmux: Add support for mpegversion 2, which is also AAC

13 years agoflacdec: Send EOS when seeking after the end of file instead of failing
Sebastian Dröge [Wed, 11 May 2011 08:25:15 +0000 (10:25 +0200)]
flacdec: Send EOS when seeking after the end of file instead of failing

Fixes bug #649780.

13 years agowavenc: Set fixedcaps getcaps function on the sinkpad
Sebastian Dröge [Fri, 29 Apr 2011 06:59:20 +0000 (08:59 +0200)]
wavenc: Set fixedcaps getcaps function on the sinkpad

wavenc does not allow to change the caps during playback
and always returning the template caps is just wrong.

13 years agowavenc: Allow setcaps to be called after a format was negotiated if it's compatible
Sebastian Dröge [Fri, 29 Apr 2011 06:55:19 +0000 (08:55 +0200)]
wavenc: Allow setcaps to be called after a format was negotiated if it's compatible

Otherwise wavenc will fail if upstream decides to set equivalent caps or caps
with additional information later.

Thanks to Alexander Schremmer for finding this bug.

13 years agoBack to development
Tim-Philipp Müller [Sat, 14 May 2011 09:02:22 +0000 (10:02 +0100)]
Back to development

13 years agoRelease 0.10.29 RELEASE-0.10.29
Tim-Philipp Müller [Tue, 10 May 2011 09:04:28 +0000 (10:04 +0100)]
Release 0.10.29

Highlights:

  - amrparse, aacparse, ac3parse, flacparse, mpegaudioparse, dcaparse audio parsers (moved from -bad)
  - muxers now mux based on running time
  - ISO MP4 muxers: mp4mux/3gppmux/qtmux/mj2mux (moved from -bad)
  - new matroskaparse element
  - new v4l2radio element
  - rtpsession: support RTCP Early Feedback (the AVPF profile)
  - orc 0.4.14 or newer recommended
  - many other fixes and improvements

13 years agoqtmux: Fix signed floating point values writing
Edward Hervey [Thu, 5 May 2011 11:24:23 +0000 (13:24 +0200)]
qtmux: Fix signed floating point values writing

You would end up on some architectures with 0 being written out
instead of the proper value.

https://bugzilla.gnome.org/show_bug.cgi?id=649449

13 years agomatroskamux: avoid building index when streamable
Mark Nauwelaerts [Wed, 4 May 2011 10:04:15 +0000 (12:04 +0200)]
matroskamux: avoid building index when streamable

... as it will not be written anyway.

Fixes #648937 (?).

13 years agobuild: add old qtdemux/quicktime directories to CRUFT_DIRS and CRUFT_FILES
Tim-Philipp Müller [Mon, 2 May 2011 11:09:02 +0000 (12:09 +0100)]
build: add old qtdemux/quicktime directories to CRUFT_DIRS and CRUFT_FILES

13 years agoflvmux: don't overwrite metadata tag with duration in streaming mode
Tom Janiszewski [Sun, 1 May 2011 04:04:03 +0000 (00:04 -0400)]
flvmux: don't overwrite metadata tag with duration in streaming mode

A duration tag gets inserted only for streamable=false, so only
update/write the duration later if we actually inserted that tag,
otherwise we write garbage into other tags.

https://bugzilla.gnome.org/show_bug.cgi?id=649060

13 years ago0.10.28.4 pre-release
Tim-Philipp Müller [Sat, 30 Apr 2011 17:16:36 +0000 (18:16 +0100)]
0.10.28.4 pre-release

13 years agoquicktime: rename plugin to isomp4
Tim-Philipp Müller [Sat, 30 Apr 2011 16:46:36 +0000 (17:46 +0100)]
quicktime: rename plugin to isomp4

https://bugzilla.gnome.org/show_bug.cgi?id=648004

13 years ago0.10.28.3 pre-release
Tim-Philipp Müller [Wed, 27 Apr 2011 11:45:51 +0000 (12:45 +0100)]
0.10.28.3 pre-release

13 years agortpgstpay: fix buffer leak
Wim Taymans [Tue, 26 Apr 2011 13:58:12 +0000 (15:58 +0200)]
rtpgstpay: fix buffer leak

13 years agojpegdec: documentation typo "jpegddec"
Philip Jägenstedt [Mon, 25 Apr 2011 08:04:52 +0000 (10:04 +0200)]
jpegdec: documentation typo "jpegddec"

https://bugzilla.gnome.org/show_bug.cgi?id=648589

13 years agoavimux,matroskamux: Add stream-format to h264 caps
David Schleef [Sun, 24 Apr 2011 23:45:07 +0000 (16:45 -0700)]
avimux,matroskamux: Add stream-format to h264 caps

Fixes #606662.

13 years agopngdec: Remove temporary code
David Schleef [Sun, 20 Feb 2011 20:13:49 +0000 (12:13 -0800)]
pngdec: Remove temporary code

Now that we depend on (what will be) -base-0.10.33.

13 years agoconfigure: don't pass -Waddress to ObjC compiler on OSX when compiling osxvideosink
Tim-Philipp Müller [Sun, 24 Apr 2011 13:03:56 +0000 (14:03 +0100)]
configure: don't pass -Waddress to ObjC compiler on OSX when compiling osxvideosink

Temporary workaround until we fix this properly and check for
the ObjC warning/error flags instead of just passing CFLAGS to the
ObjC compiler.

https://bugzilla.gnome.org/show_bug.cgi?id=643939

13 years agoquicktime: rename plugin filename from *qtdemux* to *quicktime*
Tim-Philipp Müller [Sun, 24 Apr 2011 12:29:32 +0000 (13:29 +0100)]
quicktime: rename plugin filename from *qtdemux* to *quicktime*

https://bugzilla.gnome.org/show_bug.cgi?id=648004

13 years agoAutomatic update of common submodule
Tim-Philipp Müller [Sun, 24 Apr 2011 13:03:41 +0000 (14:03 +0100)]
Automatic update of common submodule

From c3cafe1 to 46dfcea

13 years agodocs: add various qtmux variants to documentation
Tim-Philipp Müller [Thu, 21 Apr 2011 22:30:26 +0000 (23:30 +0100)]
docs: add various qtmux variants to documentation

13 years agoquicktime: register 3gppmux element in addition to the misnamed gppmux
Tim-Philipp Müller [Thu, 21 Apr 2011 21:51:52 +0000 (22:51 +0100)]
quicktime: register 3gppmux element in addition to the misnamed gppmux

13 years agortpsession: Remove incomplete support for RTCP FIR
Olivier Crête [Mon, 18 Apr 2011 22:08:30 +0000 (18:08 -0400)]
rtpsession: Remove incomplete support for RTCP FIR

Remove bits that were meant to suppport RTCP FIR

https://bugzilla.gnome.org/show_bug.cgi?id=648160

13 years agotests: add generic set_index test
Tim-Philipp Müller [Tue, 19 Apr 2011 13:33:25 +0000 (14:33 +0100)]
tests: add generic set_index test

13 years agoflvdemux: fix deadlock on setting index on flvdemux
Tim-Philipp Müller [Tue, 19 Apr 2011 13:33:42 +0000 (14:33 +0100)]
flvdemux: fix deadlock on setting index on flvdemux

13 years agotests: add index-setting test for baseparse/flacparse
Tim-Philipp Müller [Tue, 19 Apr 2011 13:16:11 +0000 (14:16 +0100)]
tests: add index-setting test for baseparse/flacparse

https://bugzilla.gnome.org/show_bug.cgi?id=646811

13 years agowavpack: Remove bus GSource to prevent a valgrind warning
Sebastian Dröge [Mon, 18 Apr 2011 09:29:15 +0000 (11:29 +0200)]
wavpack: Remove bus GSource to prevent a valgrind warning

13 years agowavenc: Remove bus GSource to prevent a valgrind warning
Sebastian Dröge [Mon, 18 Apr 2011 09:14:32 +0000 (11:14 +0200)]
wavenc: Remove bus GSource to prevent a valgrind warning

13 years agotagschecking: Remove bus GSource to prevent a valgrind warning
Sebastian Dröge [Mon, 18 Apr 2011 09:11:53 +0000 (11:11 +0200)]
tagschecking: Remove bus GSource to prevent a valgrind warning

13 years agoimagefreeze: Remove bus GSource to prevent a valgrind warning
Sebastian Dröge [Mon, 18 Apr 2011 09:10:01 +0000 (11:10 +0200)]
imagefreeze: Remove bus GSource to prevent a valgrind warning

13 years agoavidemux: fix 'variable may be used uninitialized' warnings caused by -DG_DISABLE_ASSERT
Tim-Philipp Müller [Sun, 17 Apr 2011 00:29:01 +0000 (01:29 +0100)]
avidemux: fix 'variable may be used uninitialized' warnings caused by -DG_DISABLE_ASSERT

13 years ago0.10.28.2 pre-release
Tim-Philipp Müller [Sat, 16 Apr 2011 17:50:11 +0000 (18:50 +0100)]
0.10.28.2 pre-release

13 years agogst: update disted orc backup code
Tim-Philipp Müller [Sat, 16 Apr 2011 17:49:27 +0000 (18:49 +0100)]
gst: update disted orc backup code

13 years agodocs: update for pre-release
Tim-Philipp Müller [Sat, 16 Apr 2011 17:29:45 +0000 (18:29 +0100)]
docs: update for pre-release

13 years agopo: update translations
Tim-Philipp Müller [Sat, 16 Apr 2011 17:27:54 +0000 (18:27 +0100)]
po: update translations

13 years agoqtmux: refuse incomplete legacy h264 caps
Tim-Philipp Müller [Sat, 16 Apr 2011 17:17:01 +0000 (18:17 +0100)]
qtmux: refuse incomplete legacy h264 caps

Refuse h264 caps without stream-format and codec_data fields for
now, to avoid creating broken files. This might cause some pipelines
that worked previously to fail. However, the move from -bad to -good
is our only chance to fix this up, so make it strict for now. We can
always change it back to be less strict in future.

https://bugzilla.gnome.org/show_bug.cgi?id=647919

13 years agov4l2sink: fix another unused-but-set-variable warning
Tim-Philipp Müller [Sat, 16 Apr 2011 17:16:11 +0000 (18:16 +0100)]
v4l2sink: fix another unused-but-set-variable warning

13 years agopulse, speexenc, rtpgsmpay: don't use g_assert() for error handling
Tim-Philipp Müller [Sat, 16 Apr 2011 17:10:24 +0000 (18:10 +0100)]
pulse, speexenc, rtpgsmpay: don't use g_assert() for error handling

Don't use g_assert() for error handling, even if they're highly unlikely.
Either we *know* that something can't happen, in which case we
should just not handle it, or we think something can happen, but it is
very very unlikely that it will ever happen, in which case we should
handle it like any other error instead of asserting.

g_assert() is best left for conditions we have control of, like checking
internal consistency of our code, not checking return values of external
code.

Fixes a bunch of warnings when compiling with -DG_DISABLE_ASSERT:
gstrtpgsmpay.c: In function 'gst_rtp_gsm_pay_handle_buffer':
gstrtpgsmpay.c:130:17: warning: variable 'rtpgsmpay' set but not used
gstspeexenc.c: In function 'gst_speex_enc_encode':
gstspeexenc.c:904:19: warning: variable 'written' set but not used
pulsesink.c: In function 'gst_pulsesink_change_state':
pulsesink.c:2725:9: warning: variable 'res' set but not used
pulsesrc.c: In function 'gst_pulsesrc_change_state':
pulsesrc.c:1253:7: warning: variable 'e' set but not used

13 years agoexamples: fix some warnings in rtp example
Tim-Philipp Müller [Sat, 16 Apr 2011 17:07:35 +0000 (18:07 +0100)]
examples: fix some warnings in rtp example

Caused by -DG_DISABLE_ASSERT

13 years agoexamples: don't put code with side-effects into g_assert()
Tim-Philipp Müller [Sat, 16 Apr 2011 16:57:32 +0000 (17:57 +0100)]
examples: don't put code with side-effects into g_assert()

Otherwise things won't work too well when compiling with
-DG_DISABLE_ASSERT (as we do for pre-releases and releases).

13 years agodeinterlace, matroska: fix two variable-may-be-used-uninitialized compiler warnings
Tim-Philipp Müller [Sat, 16 Apr 2011 15:51:32 +0000 (16:51 +0100)]
deinterlace, matroska: fix two variable-may-be-used-uninitialized compiler warnings

We use -DG_DISABLE_ASSERT for the pre-releases, which makes these
warnings pop up in cases that were previously covered by g_assert_not_reached()
and the like:
tvtime/greedyh.c:801:14: warning: 'scanline' may be used uninitialized in this function
matroska-mux.c:501:19: warning: 'context' may be used uninitialized in this function

13 years agojack: fix unused-but-set-variable warnings with gcc-4.6
Tim-Philipp Müller [Sat, 16 Apr 2011 12:33:45 +0000 (13:33 +0100)]
jack: fix unused-but-set-variable warnings with gcc-4.6

13 years agoexamples: fix 'control reaches end of non-void function' warning in cairo example
Tim-Philipp Müller [Sat, 16 Apr 2011 12:23:50 +0000 (13:23 +0100)]
examples: fix 'control reaches end of non-void function' warning in cairo example

13 years agov4l2src: Address unused but set variable
Robert Swain [Fri, 15 Apr 2011 13:47:24 +0000 (15:47 +0200)]
v4l2src: Address unused but set variable

The v4l2object formats list was being obtained into a local variable and
then still used from the context. Make use of the local variable.

13 years agooss4: Address unused but set variables
Robert Swain [Fri, 15 Apr 2011 13:17:34 +0000 (15:17 +0200)]
oss4: Address unused but set variables

GCC 4.6.x complains about such variable usage. Unused but set variables
were removed except that gst_oss4_mixer_slider_set_mute () now returns
the value from the call to gst_oss4_mixer_set_control_val ().

13 years agojpegenc: pulsesink: raw1394: Address unused but set variables
Robert Swain [Fri, 15 Apr 2011 13:14:13 +0000 (15:14 +0200)]
jpegenc: pulsesink: raw1394: Address unused but set variables

GCC 4.6.x spits warnings about such usage of variables. The variables in
raw1394 were marked with G_GNUC_UNUSED as this seemed omre appropriate.
The others were removed.

13 years agoy4mencode: shapewipe: Address unused but set variables
Robert Swain [Fri, 15 Apr 2011 13:12:44 +0000 (15:12 +0200)]
y4mencode: shapewipe: Address unused but set variables

GCC 4.6.x complains about such usage.

13 years agotests: Address unused but set variables
Robert Swain [Fri, 15 Apr 2011 13:11:35 +0000 (15:11 +0200)]
tests: Address unused but set variables

GCC 4.6.x spits warnings about such usage of variables.

13 years agovideomixer: Fix argb/rgba overlay orc code
Robert Swain [Fri, 15 Apr 2011 13:36:41 +0000 (15:36 +0200)]
videomixer: Fix argb/rgba overlay orc code

Remove some redundant operations (convubw) and use the correct variable,
t2, in the orc_overlay_bgra function.

13 years agovideomixer: address unused but set variables
Robert Swain [Fri, 15 Apr 2011 13:33:35 +0000 (15:33 +0200)]
videomixer: address unused but set variables

GCC 4.6.x spits warnings about variables that are set but unused. Such
variables have been removed in blend, collectpads2 and videomixer2.

13 years agortp, rtpmanager: Address unused but set variables
Robert Swain [Fri, 15 Apr 2011 12:57:20 +0000 (14:57 +0200)]
rtp, rtpmanager: Address unused but set variables

GCC 4.6.x spits warnings about variables that are unused but set. Such
variables have been removed where trivial but with comments left behind
for informational purposes in some cases.

gst_rtp_session_chain_recv_rtcp () was changed in commit 490113d4
to always return GST_FLOW_OK instead of the return value of
rtp_session_process_rtcp (), so we'll keep it that way.

13 years agoquicktime: Remove unused but set variables
Robert Swain [Fri, 15 Apr 2011 09:29:30 +0000 (11:29 +0200)]
quicktime: Remove unused but set variables

GCC 4.6.x spits warnings about such variable usage. Note that some
calculations are left as comments for informative purposes.

13 years agomatroska: Remove unused but set variables
Robert Swain [Fri, 15 Apr 2011 09:23:38 +0000 (11:23 +0200)]
matroska: Remove unused but set variables

GCC 4.6.x spits warnings about such variable usage.

13 years agoimagefreeze: Remove unused but set duration variable
Robert Swain [Fri, 15 Apr 2011 09:19:26 +0000 (11:19 +0200)]
imagefreeze: Remove unused but set duration variable

GCC 4.6.x spits warnings about such variable usage.

13 years agoflxdemux: Remove unused but set keyframe variables
Robert Swain [Fri, 15 Apr 2011 09:18:19 +0000 (11:18 +0200)]
flxdemux: Remove unused but set keyframe variables

The FIXMEs about the keyframe flag never being used are left for later
fixing, at which point the keyframe variables could be added back.

13 years agoedgetv: Remove unused but set height variable
Robert Swain [Fri, 15 Apr 2011 09:16:42 +0000 (11:16 +0200)]
edgetv: Remove unused but set height variable

GCC 4.6.x spits warnings about such variables.