The latter is 1 cycle faster on a cortex-53 and since the operands are
bytewise (or larger) bitmask (impossible to overflow to zero) both are
equivalent.

This is cherrypicked from libav commit
e7ae8f7a.
Signed-off-by: NMichael Niedermayer <michael@niedermayer.cc>

Since aarch64 has enough free general purpose registers use them to
branch to the appropiate storage code. 1-2 cycles faster for the
functions using loop_filter 8/16, ... on a cortex-a53. Mixed results
(up to 2 cycles faster/slower) on a cortex-a57.

This is cherrypicked from libav commit
d7595de0.
Signed-off-by: NMichael Niedermayer <michael@niedermayer.cc>

Signed-off-by: NMichael Bradshaw <mjbshaw@google.com>
Signed-off-by: NMichael Niedermayer <michael@niedermayer.cc>

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

Fixes ticket #6068.

The duotone file is interpreted as gray
Signed-off-by: NMichael Niedermayer <michael@niedermayer.cc>

Signed-off-by: NMichael Niedermayer <michael@niedermayer.cc>

Fixes part of #5918.
Signed-off-by: NPaul B Mahol <onemda@gmail.com>

Seemingly ff_clear_block_sse assumed that the block array is aligned,
so make sure it is.

Fixes ticket #6079
Signed-off-by: NJames Almer <jamrial@gmail.com>

Signed-off-by: NJames Almer <jamrial@gmail.com>

Signed-off-by: NJames Almer <jamrial@gmail.com>

Signed-off-by: NJames Almer <jamrial@gmail.com>

Signed-off-by: NJames Almer <jamrial@gmail.com>

Signed-off-by: NJames Almer <jamrial@gmail.com>

Several codecs other than huffyuv use them.
Signed-off-by: NJames Almer <jamrial@gmail.com>

because the oc have been  potint to hls->avf or hls->vtt_avf
here is not needed point once again
Signed-off-by: NSteven Liu <lq@chinaffmpeg.org>

when hlsenc use flag second_level_segment_index,
second_level_segment_size and second_level_segment_duration,
the rename is ok but the output filename always use the old filename
so move the rename operation after the close the ts file and
before open new segment
Reported-by: NChristian Johannesen <chrisjohannesen@gmail.com>
Reviewed-by: NBodecs Bela <bodecsb@vivanet.hu>
Signed-off-by: NSteven Liu <lq@chinaffmpeg.org>

CID: 1396852
check the devices_list alloc status,
and release the devices_list when alloc devices error
Reviewed-by: NMichael Niedermayer <michael@niedermayer.cc>
Signed-off-by: NSteven Liu <lq@chinaffmpeg.org>

Signed-off-by: NThomas Turner <thomastdt@googlemail.com>
Signed-off-by: NMichael Niedermayer <michael@niedermayer.cc>

Fixes compilation with hardcoded tables after eaff1aa0
and e71b8119Reviewed-by: NTimo Rothenpieler <timo@rothenpieler.org>
Signed-off-by: NJames Almer <jamrial@gmail.com>

Fixes ticket #6075.

Hopefully fix compilation with suncc.

Fixes decoding the sample from ticket #6072 with ffmpeg.

avfilter_graph_request_oldest() does work that should be done by
either the filter or the application.

The principle of this function, calling ff_request_frame() from
outside the filter was always shaky. This version is less elegant
since it requires making special cases for each filter, but it
is more robust since it no longer calls ff_request_frame()
directly without notifying the filter.

Eventually, avfilter_graph_request_oldest() will be deprecated
for a function to just run the graph.

ff_request_frame_to_filter() and ff_filter_frame_to_filter()
are only used in avfilter.c.

Having two different functions allows to have stricter tests
and detect errors earlier.

It does not change anything for the existing filters and makes
better code fatrorization when future code will use the utility
functions.