This fixes depending on implementation defined behavior

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

This fixes depending on implementation defined behavior

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

This fixes depending on implementation defined behavior

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

This fixes depending on implementation defined behavior

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

Fixes Ticket4332

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

When precision is fixed and volume is 0, filter_frame does not
perform any operation on the output buffer. This works if the
output buffer has been allocated and zeroed with ff_get_audio_buffer
but not if the input buffer is used as output buffer.

Fix this by not using the input buffer as output buffer if
precision is fixed and volume is 0.

Signed-off-by: Gilles Chanteperdrix <gilles.chanteperdrix@xenomai.org>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

As far as matrix coefficients are concerned, BT470BG and SMPTE170M
match BT601. This is similar to the check performed in vf_scale.

Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>

See previous commit for a rationale.

This makes the sorting of the colors along an axis (r, g or b)
predictible, and thus testable under FATE. The performance is not really
an issue here since the function is called only once at the end and will
need to sort very small number of entries, so an alternative would be to
make the sorting functions (see DECLARE_CMP_FUNC()) fallback on another
axis in case of equality. This approach was actually simpler.

I don't know if there is any advantage in using a multidimensional sort,
but it will affect the final palette one way or another.

Also drop initialization since it is initialized later on before any
usage.

Fixes CID1271048
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

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

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

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

(or 32k to 256k in 32-bit)

This is similar to a00bab3475

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

Fixes CID1270820
Fixes CID1270818

Fixes CID1270819

Fixes CID1270822

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

Fixes CID1270821

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

This might help coverity

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

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

CC: libav-stable@libav.org
Bug-Id: CID 1267891

(or 32k to 256k in 32-bit)

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

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

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

Signed-off-by: Paul B Mahol <onemda@gmail.com>

i7-4770K luma 21% faster, chroma 18% faster A10-7850K luma 42% faster, chroma 37% faster on 1920x1080 res

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

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

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

This code looks clumsy, and an assert would probably be more welcome.

ceilf() can only work if the reminder of the division is not 0.

This fixes memory errors with for instance:
  ffmpeg -f lavfi -i testsrc=s=800x500 -threads 3 -vf dctdnoiz -frames:v 1 -f null -

Signed-off-by: Eejya Singh <singh.eejya@gmail.com>
Signed-off-by: Clément Bœsch <u@pkh.me>

Fixes warning in `make checkheaders`.

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

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

This also avoids droping the frame which is closest to the target timestamp

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