Signed-off-by: Martin Storsjö <martin@martin.st>

Favor draw_mouse option. Should simplify parsing, and make it more
robust.

In particular, do not upcase first word, do not use final dot, use a verb
to specify what the option does, sort entries by name, apply random
vertical align.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Avoid fastidious creation and immediate destroy of the display in case
one of the dither options fails to be set.

Also add NULL checks in write_trailer(), which are now necessary since
the function can be called when not all the components have been
initialized.

More consistent with libcaca function names. Should not harm since the
device was added less than one day ago.

v4l2_fd_open is already called at the end of v4l2_open;
calling a second time leaks structures inside libv4l2.

Fix trac ticket #1560.

An help message tells what an option does, not what its argument is.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Stefano Sabatini <stefasab@gmail.com>

It was deprecated since a long time and removed after the 2->3 major
bump.

The list was set to avfilter_all_channel_layouts,
which is really all common channel layouts.
Not setting it leaves an empty list,
which now means really all.

Signed-off-by: Mans Rullgard <mans@mansr.com>

We generally do not declare variables within for statements and
there are compilers that choke on such constructs.

Fixes Bugzilla #279.

Signed-off-by: Diego Biurrun <diego@biurrun.de>

The rounding was wrong due to incorrect ue of floats,
changed to rationals and av_rescale.

The results were not properly passed to SDL.

The size passed to SDL_CreateYUVOverlay is the size
of the video in pixels, it must not take the aspect ratio
into account.

It allows vsync to be automatically set to passthrough.

avdevice is just a part of lavf and uses lavf API/ABI/structures

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

The planar/packed switch and the packing_formats list is no longer
required, since the planar/packed information is now stored in the sample
format enum.

This is technically a major API break, possibly it should be not too
painful as we marked the audio filtering API as unstable.

Also remove one pointless zero initialization in rangecoder.c.

Found-by: ubitux
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

When using "-f x11grab -i :0.0" valgrind reports a definitely lost
memory block with this message:

==31544== 5 bytes in 1 blocks are definitely lost in loss record 1 of 2
==31544==    at 0x4026E68: memalign (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==31544==    by 0x4026F17: posix_memalign (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==31544==    by 0x60D399A: av_malloc (in /usr/lib/x86_64-linux-gnu/libavutil.so.51.22.1)
==31544==    by 0x60D3A70: av_strdup (in /usr/lib/x86_64-linux-gnu/libavutil.so.51.22.1)
==31544==    by 0x4A2BE58: ??? (in /usr/lib/x86_64-linux-gnu/libavdevice.so.53.2.0)
==31544==    by 0x506D29E: avformat_open_input (in /usr/lib/x86_64-linux-gnu/libavformat.so.53.21.0)
==31544==    by 0x400A80: main (in /home/ao2/WIP/am7xxx-play/tests/a.out)

The 5 bytes lost are the ones from param = av_strdup(":0.0"), so let's
free param in the exit path.

Also check the av_strdup() return value.

Note: calling av_free(param) even when av_strdup() fails and param is
NULL is OK and keeps the code simpler without adding another label to
skip av_free().

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>

The period argument is supposed to be the number of samples since
the last update.

Initially found and designed by Michael Niedermayer.

The feedback factors for the timefilter are directly computed from
the expected period. This commit changes the init function to accept
the period itself and compute the feedback factors internally,
rather than having all client code duplicate the formulas.

This commit also actually fixes the formulas: the current code had
sqrt(2*o), but the correct formula, both theoretically and according
to experimental testing, is sqrt(2)*o.

Furthermore, it adds an exponential to feedback factors larger than
1 with large periods.