Fixes clang warning "floating point absolute value function 'fabsf'
when argument is of integer type [-Wabsolute-value]".

Reported-By: infer

Reported-By: infer

Reported-By: infer

CC: libav-stable@libav.org
Sample-Id: test_bref.mp4
Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>

Untangle eamad from the mpeg12.h header and speed up rebuilding.

Having this macro in an header only facilitates the use of such header.
The code increase is minimal and files have one less dependency
on mpegvideo.h.

Mark it as static.

It is only used in one place, and it is small enough that it might be
called directly.

display_matrix_size is only initialized when av_stream_get_side_data()
returns a side data pointer. The code is safe since the only effect this
has is setting the display_matrix pointer to NULL which it was already
anyway.

Bug-Id: 865
CC: libav-stable@libav.org

The first check is done without the AVIOContext, so alloc it only if said check succeeds
Signed-off-by: James Almer <jamrial@gmail.com>
Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Fixes stack buffer overflow errors detected by address sanitizer in
fate-imc.

CC: libav-stable@libav.org

Fixes stack buffer overflow errors detected by address sanitizer in
various fate tests.

CC: libav-stable@libav.org

Fixes stack buffer overflow errors detected by address sanitizer in
various fate tests.

CC: libav-stable@libav.org

CC: libav-stable@libav.org
Bug-Id: CID 1257774

CC: libav-stable@libav.org
Bug-Id: CID 1292518

CC: libav-stable@libav.org
Bug-Id: CID 1292520

CC: libav-stable@libav.org
Bug-Id: CID 609333

CC: libav-stable@libav.org
Bug-Id: CID 1302711

It is necessary to avoid circular header dependencies.

The main ff_alloc_picture() function is made more generic with all the
parameters necessary as arguments. This will allows to move most of the
related functions to a separate file later.

Right now wrappers are provided to try and minimize the number of
changes in the code.

Use more generic arguments and remove its static attribute since it will
be moved to a separate file.

The C runtime C99 compatibility had been improved a lot and it now
rejects some of the compatibility defines provided for the older
versions.

Many thanks to Ray for the time spent testing.

Bug-Id: 864
CC: libav-stable@libav.org

The channel configuration can be delivered only by the PCE,
try to parse it first and not try to decode until a channel
configuration is set.

CC: libav-stable@libav.org

These are defined in ISO/IEC 14496-3:2009/PDAM 4 for 6.1 and 7.1.
It also defines another 7.1 layout with configuration 14, that one
is not added here for now.

11: 3/3.1    FC FL+FR BL+BR BC LFE
12: 3/2/2.1  FC FL+FR SiL+SiR BL+BR LFE
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>

FDK AAC encoder outputs SCE(front)+CPE(front)+CPE(back)+CPE(back) on
MODE_7_1_REAR_SURROUND configuration.
Since decoder couldn't properly map 4 back channels, decoding failed
unless -request_channel_layout 0x8000000000000000 has been specified.
Now we treat first CPE(back) as CPE(side) on channel mapping.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>

Support only streams with Content-Length.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>

The decoder assigns channels using default channel configuration
for 5.1ch when it parses an ADTS frame header using consecutive
channel ids.

When a PCE comes, it reassigns channels using PCE configuration
using directly the ids provided. They can be arbitrary.

Always use consecutive channel ids to avoid decoding glitches due
spurious reconfigurations due the channel ids mismatch between the
two otherwise-identical channel maps.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>