Currently the frame pool used by the default get_buffer2()
implementation is a single struct, allocated when opening the decoder.
A pointer to it is simply copied to each frame thread and we assume that
no thread attempts to modify it at an unexpected time. This is rather
fragile and potentially dangerous.
With this commit, the frame pool is made refcounted, with the reference
being propagated across threads along with other context variables. The
frame pool is now also immutable - when the stream parameters change we
drop the old reference and create a new one.
This allows the fuzzer to target meaningfull codec tags instead
of hunting the 4gb space, which it seems to have problems with.
Suggested-by: James
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Some decoders may not need a writable buffer in some specific cases, but only
a reference to the existing buffer with updated frame properties instead, for
the purpose of returning duplicate frames. For this, the
FF_REGET_BUFFER_FLAG_READONLY flag is added, which will prevent potential
allocations and buffer copies when they are not needed.
Signed-off-by: James Almer <jamrial@gmail.com>
This allows decoding more als reference samples
Suggested-by: Thilo Borgmann <thilo.borgmann@mail.de>
Reviewed-by: Thilo Borgmann <thilo.borgmann@mail.de>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
This was reduced from 128 in libav commit
192f1984b1, but since we support unknown channel
layouts, we can increase this limit.
Fixes ticket #6332.
Signed-off-by: Marton Balint <cus@passwd.hu>
And remove the function altogether while at it. It's a duplicate of
another.
Reviewed-by: wm4 <nfxjfg@googlemail.com>
Signed-off-by: James Almer <jamrial@gmail.com>
This is just a lot of complicated and confusing code that had no purpose
anymore.
Also, the functions return values were checked only sometimes. Locking
shouldn't fail anyway, so remove the return values. Barely any other
pthread lock calls check the return value (including more important code
that is more likely to fail horribly if locking fails).
It could be argued that it might be helpful in some debugging
situations, or when the user built FFmpeg without thread support against
all good advice.
But there are dummy atomics too, so the atomic check won't help with
ensuring correctness absolutely. You gain very little.
Also, for debugging, you can just raise the ASSERT_LEVEL, and then
libavutil/thread.h will redefine the locking functions to explicitly
check the return values.
It's completely absurd that libavcodec would care about libavformat
locking, but it was there because the lock manager was in libavcodec.
This is more stright forward. Changes ABI, but we don't require ABI
compatibility currently.
This reverts commit 590136e78d.
Atomics are not required for this variable, because it is protected
through the lock manager, and the use of atomics here is not compatible
with the c11 emulation wrappersi.
Fixes FATE on MSVC, among other setups which use the compat wrappers.
This removes the dependency that hardware pixel formats previously had on
AVHWAccel instances, meaning only those which actually do something need
exist after this patch.
Also updates avcodec_default_get_format() to be able to choose hardware
formats if either a matching device has been supplied or no additional
external configuration is required, and avcodec_get_hw_frames_parameters()
to use the hardware config rather than searching the old hwaccel list.
The FF_CODEC_CAP_HWACCEL_REQUIRE_CLASS mechanism is deleted because it
no longer does anything (the codec already contains the pointers to the
matching hwaccels).
* commit '3152058bf1dca318898550efacf0286f4836cae6':
libavcodec: Don't use dllexport, only dllimport when building DLLs
Merged-by: James Almer <jamrial@gmail.com>
The only purpose of dllexport (which is set while building the library
that exports the symbols) is to have the linker automatically
export such symbols into a DLL without using a def file - it doesn't
affect the generated code.
For both MSVC and mingw builds, this isn't essential since we override
what symbols to export via an autogenerated def file instead.
Update a comment in configure to refer to the right concept.
With lld, this avoids warnings about duplicate export directives,
when some symbols are requested to be exported both via dllexport
attributes and via the autogenerated def file.
This also reduces the number of lines of code marginally.
Signed-off-by: Martin Storsjö <martin@martin.st>
* commit 'accb06120c13a4ead442464d96f2fa318fa07a4e':
configure: Use dllexport/dllimport for data symbols across DLLs with mingw
Merged-by: James Almer <jamrial@gmail.com>
* commit 'abf1c058d1bd0ed1b820ea5e501a4484756f00b0':
msvc: Properly specify dllexport for data symbols shared across dll boundaries
Merged-by: James Almer <jamrial@gmail.com>
Currently, AVHWAccels are looked up using a (codec_id, pixfmt) tuple.
This means it's impossible to have 2 decoders for the same codec and
using the same opaque hardware pixel format.
This breaks merging Libav's CUVID hwaccel. FFmpeg has its own CUVID
support, but it's a full stream decoder, using NVIDIA's codec parser.
The Libav one is a true hwaccel, which is based on the builtin software
decoders.
Fix this by introducing another field to disambiguate AVHWAccels, and
use it for our CUVID decoders. FF_CODEC_CAP_HWACCEL_REQUIRE_CLASS makes
this mechanism backwards compatible and optional.
* commit '94eed68ace9f2416af8457fcbf142b175928c06b':
lavc: Drop deprecated options moved to private contexts
Merged-by: James Almer <jamrial@gmail.com>
Their use in the public header is deprecated and will be removed, but
they are still needed by some codecs at least as long as qscale related
deprecated fields in the AVFrame struct remain in the tree.
This avoids having to use pseudo relocations.
The version script used for exporting functions is skipped as soon
as the set of object files contains symbols marked with dllexport,
therefore we need to use makedef to produce the full list of symbols
to be exported.
Signed-off-by: Martin Storsjö <martin@martin.st>
We currently only have exported data symbols within libavcodec, but
the concept is easy to extend to other libraries if necessary.
The attribute declaration needs to be in a private header though,
since we can't use CONFIG_SHARED in public installed headers.
Signed-off-by: Martin Storsjö <martin@martin.st>
* commit '019ab88a95cb31b698506d90e8ce56695a7f1cc5':
lavc: add an option for exporting cropping information to the caller
Merged-by: James Almer <jamrial@gmail.com>
So, all frames and errors are correctly reported in order.
Also limit the numbers of error during draining to prevent infinite loop.
This fix fate failure with THREADS>=4:
make fate-h264-attachment-631 THREADS=4
This also reverts a755b725ec.
Suggested-by: wm4, Ronald S. Bultje, Marton Balint
Reviewed-by: w4 <nfxjfg@googlemail.com>
Reviewed-by: Ronald S. Bultje <rsbultje@gmail.com>
Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: Muhammad Faiz <mfcc64@gmail.com>
* commit '972c71e9cb63e24f57ee481e413199c7d88a8813':
lavc: add support for filtering packets before decoding
Merged-by: James Almer <jamrial@gmail.com>
* commit '549d0bdca53af7a6e0c612ab4b03baecf3a5878f':
decode: be more explicit about storing the last packet properties
Also copy pkt->size in extract_packet_props(), as it's needed for
AVFrame.pkt_size
Merged-by: James Almer <jamrial@gmail.com>
Allows to get a more realistic total bitrate (and estimated file size)
in avi_write_header. Previously a static default value of 200k was
assumed.
Adds an internal helper function for bitrate guessing.
Signed-off-by: Tobias Rapp <t.rapp@noa-archive.com>
Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Apparently the demuxer outputs the wrong padding for HE-AAC (based on
the raw sample rate, or so). aacdec contains a hack to adjust the muxer
padding accordingly before it's used to trim the decoder output. This
modified the packet side data, which in combination with the old
decoding API would change the packet the user passed to the decoder.
This is clearly not allowed, and it breaks running some gapless fate
tests with "-fflags +keepside" applied (without keepside, the packet
metadata is typically newly allocated, essentially making a copy and not
modifying the user's input packet).
This should probably be fixed in the demuxer (and consequently also the
muxer), but for now only fix the immediate problem.
Regression since 946ed78f5f (2012).
Currently, the new decoding API is pretty much just a wrapper around the
old deprecated one. This is problematic, since it interferes with making
full use of the flexibility added by the new API. The old API should
also be removed at some future point.
Reorganize the code so that the new send_packet/receive_frame functions
call the actual decoding directly and change the old deprecated
avcodec_decode_* functions into wrappers around the new API.
The new internal API for decoders is now changing as well. Before this
commit, it mirrors the public API, so the decoders need to implement
send_packet() and receive_frame() callbacks. This turns out to require
awkward constructs in both the decoders and the generic code. After this
commit, the decoders only implement the receive_frame() callback and
call a new internal function, ff_decode_get_packet() to obtain input
data, in the same manner to how the bitstream filters now work.
avcodec will now always make a reference to the input packet, which means
that non-refcounted input packets will be copied. Keeping the previous
behaviour, where this copy could sometimes be avoided, would make the
code significantly more complex and fragile for only dubious gains,
since packets are typically small and everyone who cares about
performance should use refcounted packets anyway.
The current code stores a pointer to the packet passed to the decoder,
which is then used during get_buffer() for timestamps and side data
passthrough. However, since this is a pointer to user data which we do
not own, storing it is potentially dangerous. It is also ill defined for
the new decoding API with split input/output.
Fix this problem by making an explicit internally owned copy of the
packet properties.
With the new decode API, doing this in ffmpeg.c is impractical. There
was resistance against removing the warning, so put it into libavcodec.
Not bothering with reducing the warning to verbose log level for
subsequent wanrings. The warning should be rare, and only happen when
developing new codecs for the old API.
Includes a change suggested by Michael Niedermayer.
Until now, the decoding API was restricted to outputting 0 or 1 frames
per input packet. It also enforces a somewhat rigid dataflow in general.
This new API seeks to relax these restrictions by decoupling input and
output. Instead of doing a single call on each decode step, which may
consume the packet and may produce output, the new API requires the user
to send input first, and then ask for output.
For now, there are no codecs supporting this API. The API can work with
codecs using the old API, and most code added here is to make them
interoperate. The reverse is not possible, although for audio it might.
From Libav commit 05f66706d1.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
