Code done by Laurent Aimar

This is work done by Rémi Denis-Courmont

Code is Pierre Ynard, aka linkfanel

Missing a "Lesser" in the legal part

We need PutImage meaning the Input and the Image bits in type mask.
Output adapters are rather rare (the _exotic_ Xorg V4L2 driver is one).

Cool hints:
 - If your header doesn't match "Lesser" 3 times, you are not good
 - If you use LGPLv2, you are not good
 - If you use the old FSF address, you are not good

Main author is Keary Griffin

Contributors are:
 - Rémi Denis-Courmont

Indirect contributors (mostly includes and parenthesis) are:
 - Laurent Aimar
 - Sam Hocevar
 - Antoine Cellerier
 - Jérome Decoodt
 - Clément Stenac
 - Rémi Duraffort
 - Jean-Paul Saman
 - Christophe Massiot
 - Gildas Bazin

And reset the rtsp-host variable accordingly. This was broken since
you couldn't set a base path without passing garbage as the host name.

--rtsp-host was used to set vod configuration including the base path.
Restore its parsing to fix this, and reset httpd configuration
variables.

The HTTP output parses the dst parameter to set the http-host and
http-port variables before calling the httpd, as discussed and approved
by Rémi in http://mailman.videolan.org/pipermail/vlc-devel/2011-September/081916.html

This will allow normal variable inheritance

This makes up for the lack of pa_context_rttime_free() in libpulse.

This should make life easier for *BSD support, and less error-prone on
Linux. Unfortunately, this does not work for enumerations.

Fix warning

Works-around #5396

This is no longer necessary now, when the omxil module keeps
track of the number of initializations of the OMX Core.
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

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

This avoids deinitializing the global OMX Core if a second
instance of the module is initialized (e.g. for testing whether
it can handle decoding another stream) and deinitialized while
the first instance still is active.
Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>