Pointed-out-by: Pedro Ribeiro <pedrib@gmail.com>

libavcodec version >= 54.92.100
libavutil  version >= 52.10.100

The first time befor a surface is used it will return VASurfaceReady.
However whenever it has been used it status is ALWAYS VASurfaceRendering,
therefor it is unusable as a guard against selecting a surface that is
being used for rendering or displaying.

Use p_va->pix_fmt to decide what to do.

(cherry picked from commit 105af824e6e5d298fa797bee2515a4fd121f1b39)
Signed-off-by: Jean-Paul Saman <jpsaman@videolan.org>

(cherry picked from commit 82fdabda9df936231519720e055c7033d5a209b0)
Signed-off-by: Jean-Paul Saman <jpsaman@videolan.org>

Make it possible to select a specific 'hw decoder' module. Currently
only: vaapi, dxva2 and vdpau are supported choices.

(cherry picked from commit 501bf022f353922c14f1d718874a3bc06931d3fc)
Signed-off-by: Jean-Paul Saman <jpsaman@videolan.org>

This should render vdpau safe to enable with profiles not supported
by the hardware (such as high depth or high chroma). Software fallback
will be used automatically.
(cherry picked from commit 1b83ab71d0f2787be4070b1a43235dc99d3d5ad2)
Signed-off-by: Jean-Paul Saman <jpsaman@videolan.org>

(cherry picked from commit 6559a56bd9da9e6c7f88e6d99cdd9c5e2a795340)
Signed-off-by: Jean-Paul Saman <jpsaman@videolan.org>

(cherry picked from commit f902d8b15b1d1e527d7bb5e9adf715775ee68bbb)
Signed-off-by: Jean-Paul Saman <jpsaman@videolan.org>

(cherry picked from commit a0e61ed659cde4db1778001e0543df8785de1261)
Signed-off-by: Jean-Paul Saman <jpsaman@videolan.org>

This is mostly useful if the hw accelerator cannot handle the stream.
The software decoder fallback will then use frame multithreading if it
supports it.
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>
(cherry picked from commit b72986e83bad854b2db4a0e7e1d4e54be5140179)
Signed-off-by: Jean-Paul Saman <jpsaman@videolan.org>

(cherry picked from commit eb86cfe7de7acb16d437238cf572d7d63b1f474e)
Signed-off-by: Jean-Paul Saman <jpsaman@videolan.org>

(cherry picked from commit c5d9ac73f6d09b2307c32c43cda342bfd2de09a7)
Signed-off-by: Jean-Paul Saman <jpsaman@videolan.org>

This requires bleeding-edge libavcodec.
(cherry picked from commit ae1e8b4ac3167981f7f46c83afeaa9423330fc18)

Conflicts:
	modules/codec/Modules.am
	po/POTFILES.in

(cherry picked from commit 3d29910e66214abb38d72a5d8172c4402144eb0d)

Conflicts:
	modules/codec/avcodec/video.c

The acceleration plugin now returns the expected pixel format instead
of checking it. That way, the avcodec module can check it directly.

This works as long as each acceleration plugin supports only one pixel
format (otherwise a table would be required).
(cherry picked from commit b667100bcf0cbd916db4fbb6ce6f8a51ff971ad8)

Conflicts:
	modules/codec/avcodec/dxva2.c
	modules/codec/avcodec/va.h
	modules/codec/avcodec/vaapi.c
	modules/codec/avcodec/vda.c
	modules/codec/avcodec/video.c

(cherry picked from commit e7eb250a251766658a3ac5d8d8e7da9a3d792b4f)

Conflicts:
	modules/codec/Modules.am
	modules/codec/avcodec/avcodec.c
	modules/codec/avcodec/dxva2.c
	modules/codec/avcodec/va.h
	modules/codec/avcodec/vaapi.c
	modules/codec/avcodec/vda.c
	modules/codec/avcodec/video.c

(cherry picked from commit 93fd69f494343adb8871501729801c49c7651130)

Conflicts:
	modules/codec/avcodec/dxva2.c
	modules/codec/avcodec/va.h
	modules/codec/avcodec/vaapi.c
	modules/codec/avcodec/vda.c
	modules/codec/avcodec/video.c

This is a manual backport of commit-id: 1962cd3095652e676cea7830da863cc6f2b621d6

It is really 2.0.7 now.

Remove the dependency on video_output/opengl.{c,h} and cleanup an unused variable.

This reverts commit 91f15717.

Conflicts:
	modules/codec/avcodec/vaapi.c
	src/video_output/video_output.c

Rename functions and structures.

Reuse as much code as possible from vaapi-xcb implementation. Most of
the routines can be shared and are placed inside common.{c,h}. The
video output has been moved to modules/video_output/vaapi/, but is
still build from modules/codec/avcodec/Modules.am. This is because
the video output shares the same libva connection as the decoder.
Since this is mandatory for some GPU implementations such as AMD Fusion.

Ignore wholly translucent rows at the top and bottom of the subtitle page

Only create an overlay large enough for the remaining rows and only only ask
zvbi to render those rows.

One happy side effect is that now --vbi-text no longer outputs header text
for some teletext streams (although --vbi-text is still unusable imho).

Reducing the size of the overlay allows stutter free playback of teletext
subtitles on significantly slower hardware.