Also move some stuff from vlc_config.h to vlc_aout*.h

Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>

This is required for PulseAudio synchronization and to pause/resume
with low-latency. This should also be useful for other buffered
 audio outputs such as ALSA or CoreAudio.

This keeps the VLC value in sync with the PulseAudio one. For some
reason, the Qt4 interface does not pick up the correct default though.

Notes:

- The variable value represents the PulseAudio sink index.
  It is not a member of the AOUT_VAR_ enumeration.

- There is no configuration item. PulseAudio normally remembers the
  last sink VLC used. So I think there is no need value in writing
  code to duplicate that effort.

- I am not sure if we need more locking inside the variable callback.
  This needs to be clarified with PulseAudio guys.

This improves mute flag handling in the PulseAudio output:
We do not blindly reset the mute flag.

This enables use of PulseAudio flat volume, if enabled in PulseAudio.

Note that, in that case, VLC will multiply its own volume by the output
device ("sink" in PulseAudio parliance) base volume. That effectively
maps 100% volume in VLC to +0dB in PulseAudio. To me, that seems more
intuitive and backward compatible with older VLC versions.

However, PulseAudio people seem to think that 100% should be mapped to
the maximum hardware amplification without software amplification,
that is to say PA_VOLUME_NORM.

Known limitations:

* The initial volume is determined by PulseAudio. However VLC shows 100%
regardless, which is not necessarily correct.

* If the VLC stream/sink input volume is changed outside of VLC (by a
mixer application), the VLC user interface will not see the updated
value.

This should fix practically glitches. The resulting A/V synchronization
is not too bad for me inspite of the long buffer (1/2 seconds).
Your mileage may vary.

This suppresses most hiccups. But it also decreases precision to
slightly noticeable levels: about +/-150ms here. For reference, our
audio output core target is +/-40ms.

This should fix the deadlocks and memory starvation problems.

 - preserve sample format from VLC, if possible,
 - preserve physical channels mapping,
 - remove a useless callback,
 - fix small race conditions with pa_*_wait() on startup,
 - cosmetic changes.

pulse: let pulseaudio define prebuffer/minrequest and use that as buffersize, also change maxlength to 100ms

It's more along pulseaudio recommendations, and we don't really need to define those
ourself. Maxlength is defined to keep latency in someway in limits. Please tell if
this acts for you.

Signed-off-by: Ilkka Ollakka <ileoo@videolan.org>

Also compare against RUNNING and not DONE on waiting loop. Also lock mainloop
when modifying values that are used in callback-code, and don't signal without lock.

Signed-off-by: Ilkka Ollakka <ileoo@videolan.org>

Certain LibVLC applications fail to call XInitThreads() before they
call XOpenDisplay(). Then VLC calls XInitThreads(). Then the
applications call XCloseDisplay(), which raises a segmentation fault.
In this case, Xlib tries acquire lock that was never created as the
Display ppinter was created before threaded Xlib mode was enabled.

These applications can now pass --no-xlib to libvlc_new(). This will
prevent any VLC Xlib-based plugin from being used. Currently, this
affects interfaces (not really an issue), PulseAudio (until Colin's
patch is merged upstream) and GLX (Xlib-based by design).

This will be necessary to address #3662.

This reverts commit 8b40fb0f.
Reading the PulseAudio source code, #2538 seems wrong.

Conflicts:

	bindings/python