Only I-frames are decoded for now.
Signed-off-by: Paul B Mahol <onemda@gmail.com>

Cherry-picked from Libav d30719e6.
Signed-off-by: wm4 <nfxjfg@googlemail.com>

Makes sure the output can be mapped as OpenGL texture.
This is what at least video players normally want.

If AVVideotoolboxContext.cv_pix_fmt_type is set to 0, don't set the
kCVPixelBufferPixelFormatTypeKey value on the VT decoder.

This makes VT output its native format, which can be much faster on
some hardware iterations (if the native format does not match with
the requested format, it will be converted, which is slow).

The default is still forcing nv12.

Public fields were added after the private fields (negating the entire
point of this). New private fields go into AVStreamInternal anyway.

The new marker was set by guessing which fields are supposed to be
private and wshich not. recommended_encoder_configuration is accessed by
ffserver_config.c directly, and is supposed to use the public API.

ffmpeg.c accesses AVStream.cur_dts, even though it's a private field,
but that seems to be an older error.

Allow all struct fields to be accessed directly, as long as they're
public.

Before this change, many fields were "public", but could be accessed via
AVOption only. This meant they were effectively not public, but were
present for documentation purposes, which was incredibly confusing at
best.

Similar code is used elsewhere in vp56 to force a more complete reinit in the future.
Fixes null pointer dereference
Fixes: 707/clusterfuzz-testcase-4717453097566208

Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/targets/ffmpegSigned-off-by: Michael Niedermayer <michael@niedermayer.cc>

Fixes undefined shift, all callers should be changed to check the value
they use with wp_exp2() or its return value.

Fixes: 692/clusterfuzz-testcase-5757381516460032

Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/targets/ffmpegSigned-off-by: Michael Niedermayer <michael@niedermayer.cc>

This matches ff_h263_decode_motion() both functions error codes are interpreted by the same common code

Fixes: 690/clusterfuzz-testcase-4744944981901312

Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/targets/ffmpegSigned-off-by: Michael Niedermayer <michael@niedermayer.cc>

Fixes: 689/clusterfuzz-testcase-6029352737177600

Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/targets/ffmpegSigned-off-by: Michael Niedermayer <michael@niedermayer.cc>

Fixes: asan_bug_leak
Found-by: Thomas Guilbert <tguilbert@google.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

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

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

Removes explicit reference to symbols and fixes dereferencing issue.
Signed-off-by: Rick Kern <kernrj@gmail.com>

Fixes timeout with 686/clusterfuzz-testcase-5853946876788736

this shortcuts (i.e. speeds up) the error and
return-to-user when decoding a truncated frame

Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/targets/ffmpeg
Previous version reviewed by: "Ronald S. Bultje" <rsbultje@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

Fixes: 682/clusterfuzz-testcase-4799120021651456

Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/targets/ffmpegSigned-off-by: Michael Niedermayer <michael@niedermayer.cc>

Fixes timeout
Fixes: 501/clusterfuzz-testcase-5672752870588416

Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/targets/ffmpegSigned-off-by: Michael Niedermayer <michael@niedermayer.cc>

Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>

qmin and qmax are not necessary for nvenc vbr.
Enforcing this constraint, doesn't allow user to use vbr 2 pass mode without explicity setting the qmin and qmax options
Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>

Fixes ticket #5546.

Regression since 2c8d876d

Rodger Combs will be added to the ffmpeg-security alias when this patch is applied
Reviewed-by: Paul B Mahol <onemda@gmail.com>
Reviewed-by: Clément Bœsch <u@pkh.me>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

wm4 will be added to the ffmpeg-security alias when this patch is applied
Reviewed-by: Paul B Mahol <onemda@gmail.com>
Reviewed-by: Clément Bœsch <u@pkh.me>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

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

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

Fixes: 680/clusterfuzz-testcase-5416627266912256
Fixes: 681/clusterfuzz-testcase-5013323462475776

Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/targets/ffmpegSigned-off-by: Michael Niedermayer <michael@niedermayer.cc>

Fixes integer overflow
Fixes: 675/clusterfuzz-testcase-6722971232108544

Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/targets/ffmpegSigned-off-by: Michael Niedermayer <michael@niedermayer.cc>

Fixes: 674/clusterfuzz-testcase-6713275880308736

Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/targets/ffmpegSigned-off-by: Michael Niedermayer <michael@niedermayer.cc>

Fixes: 677/clusterfuzz-testcase-6635120628858880

Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/targets/ffmpegReviewed-by: Steven Liu <lingjiujianke@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

One byte less is read in case of small width.
Closes #6194.
Signed-off-by: Paul B Mahol <onemda@gmail.com>

Fixes SIGFPE, closes #6196.
Signed-off-by: Paul B Mahol <onemda@gmail.com>

Between 1.00 and 1.16 times faster on Intel Yorkfield Core 2 Quad.
Between 1.11 and 1.39 times faster on Intel Kaby Lake Pentium.

~1.37x faster (147 vs. 108 cycles) compared to mmxext function

~1.10x faster (69 vs. 63 cycles) compared to mmxext function

~1.14x faster (90 vs 78 cycles) compared with mmxext

~1.21x faster (68 vs. 56 cycles) compared with mmxext function

~1.14x faster (93 vs. 81 cycles) compared with mmxext function