On each picture the video decoder wrapper did query XDM_GETSTATUS for decoder status,
but error information is alread available in p_sys->out_args in member 'extendedError'.
This patch saves one syscall every picture (except the first). The first picture defines
the output properties of the decoded pictures and these must be stored in p_dec->fmt_out.video.*
to obtain correctly sized pictures (picture_t).

No functional changes, but just cleaning up.

Remove inefficient memset() on the entire output picture. Instead the
framebuffer is expected to clear the display properly on initialization
and cleanup. This patch requires commit id: 09976bb9.

The framebuffer video output cleared the display in multiple places and it only
worked properly for RGB displays. This patch removes several memset() on the entire
framebuffer.
Clearing of the framebuffer is done in p_vout->pf_init() and p_vout->pf_end() taking
the chroma of the video output into account. Only RGB2, RV15, RV16, RV24, RV32, UYVY,
YUNV and Y422 chromas are currently supported.

Update the DaVinci resizer implementation modules/codec/davinci to latest
state of Rafael Carre his work. This introduces one new file 'resizer_coefs.c'.

The file modules/codec/davinci/fb.c used a hack to bypass VLC video
output chain in order to improve playback speed. By doing so it was
able to eliminate a lot of memory copies from and to different buffer
types.

Improvements for readability:
- Renamed p_sys->e to p_sys_>engine
- Renamed p_sys->d to p_sys->handle

Commit-id: 91ebe760 makes sure that
CERuntime_Init() and CERuntimeExit() are called only once. This commit
uses the wrappers that ensure that for audio and video decoder.

Optimize for the case where fullscreen is not requested. In that case just
copy the lines into a black destination picture and let the TV do the scaling.
If your TV does not do scaling, then you will see the video in original format.

If the source and destination picture are not of the same chroma (YVUV), then
use the resizer to do chroma conversion and scale at the same time.

Write directly into the special DSP buffer and only copy the remainder to a temporarily buffer.

- encoder.c: separate XDM 0.9 and XDM 1.0 code into separate files
- fourcc.c : indicate which XDM API to use for this codec

- remove hackish code for direct framebuffer output

Specifying a revision for ffmpeg has no influence on the libswscale. This will be
the current head of libswscale which might conflict with ffmpeg's HEAD. The
suggested revision to use is 29968, which must be forced by removing libswscale and
checking out: svn co -r 29968 svn://svn.ffmpeg.org/mplayer/trunk/libswscale libswscale

Forcing the encoder to output an key frame at every --sout-davinci-keyint setting improves the generated stream random access feature. A client picks up the stream every keyint frames now. While previous it didn't.

Capture audio in a seperate thread. It ensures that --input-slave oss:// runs independent of the video capturing and resulting in higher throughput.

This is a workaround for the Neuros OSD 2 linux kernel 2.6.23. Since vlc does a poll before a read it is guaranteed that there is data, thus using O_NONBLOCK does not really add any advantage here.

This reverts commit 3e1c59ce.

This reverts commit f86d0e82.

Fixes:
- compiler warnings on unused variables
- comparing correct types
- optimized calculations

The framerate calculation previously done in v4l2 access input was always late. Reworked the calculation to move closer to the requested framerate by setting a new frame grab deadline based on grab-date + duration + late (deadline miss). On a fast CPU it will not make much difference, but on a slow CPU the algorithm will grab frames based on FPS settings.