The code would block on the Audio Capture Thread that did not finish,
when the recording was stopped. By explicitly checking for the module
to be alive in the Audio Capture Thread this problem is resovled.

The codec format 'araẃ' is a vlc internal codec that selects the
araw audio decoder. It does not specify the bits per sample.

Silence overrun messages and add notify vlc core of timestamp.

The alsa provided 'arecord' application tries to read first before doing
a snd_pcm_wait() (aka poll). Doing it the other way results in loss of
capture buffers when the system is under load. This is either a bug in
alsa implementation of snd_pcm_wait() or something else. The disadvantage
is that there seems to be a lot more overruns now.

Make fourcc available for access alsa input without dragging in
vlc_aout.h.

On some systems the latency for audio capture when used with input-slave
is too big, resulting in poll timeouts and no audio. Running the audio
capture in a separate thread solves this latency issue.

If allocating fails, then graciously fail instead of crash.

Graciously fail if snd_pcm_start() returns an error.

The framerate as specified by the user or detected by the v4l2 access input was
overwritten by the rawvideo decoder. The module did not specify:

	es_fmt.video.i_frame_rate
	es_fmt.video.i_frame_rate_base

Their default values are 0. The rawvideo decoder detects that there is no framerate
specified and resorts to its own default (fps 25.0). Through setting framerate and
framerate_base in the v4l2 es format the correct fps is communicated to vlc core.

The framerate reduction code is not needed. If the user did not set
the requested framerate, then the v4l2 module asks the driver for its
default value. When fps is specified by the user then that one is used.

here I rename „VLC grotuvas“ to „VLC leistuvė“ according to
recommendations (because „grotuvas“ is better word for technical device
and „leistuvė“ is better for software application)
Signed-off-by: Christophe Mutricy <xtophe@chewa.net>

Signed-off-by: Christophe Mutricy <xtophe@chewa.net>

Signed-off-by: Christophe Mutricy <xtophe@chewa.net>

(cherry picked from commit 82f06bd0e7d0112a8448914e2fddec771f35792d)

Kludge beurk beurk
(cherry picked from commit 26edb0ca8af568cbba61b0e1b3b3b6fffa680b8f)
Signed-off-by: Pierre Ynard <linkfanel@yahoo.fr>

Fix a busy loop and socket leak
(cherry picked from commit d16bdb968e2d1ef40e1cc109b79a4a0f3322b61f)
Signed-off-by: Pierre Ynard <linkfanel@yahoo.fr>

(cherry picked from commit 90000056f6c533170b92dd37fe1655e25f85c317)
Signed-off-by: Pierre Ynard <linkfanel@yahoo.fr>

(cherry picked from commit 415dab3b66f17c808045d41577c0b97d7732db33)
Signed-off-by: Pierre Ynard <linkfanel@yahoo.fr>

Fix a busy loop and socket leak. Thanks to Santiago Gimeno for
pointing it out (Debian bug #607869)
(cherry picked from commit 8e49d55a31f2e96d56e0c65e49b25c4c45589ea6)
Signed-off-by: Pierre Ynard <linkfanel@yahoo.fr>

This reverts commit a1b3978d280993a87033ee3de22b08d6dad9544f.

It doesn't fix anything and it breaks new connections
Signed-off-by: Pierre Ynard <linkfanel@yahoo.fr>
(cherry picked from commit c4e65e386a753aef5d9954b2dc280056b8682e4b)
Signed-off-by: Pierre Ynard <linkfanel@yahoo.fr>

Signed-off-by: Christophe Mutricy <xtophe@chewa.net>

Signed-off-by: Christophe Mutricy <xtophe@chewa.net>

Signed-off-by: Christophe Mutricy <xtophe@chewa.net>

Signed-off-by: Christophe Mutricy <xtophe@chewa.net>

Signed-off-by: Christophe Mutricy <xtophe@chewa.net>

Signed-off-by: Christophe Mutricy <xtophe@chewa.net>

Signed-off-by: Christophe Mutricy <xtophe@chewa.net>

This reverts commit 1b29a131.
This reverts commit 79f476d6.