HWAccels with frame threads are fundamentally flawed in avcodecs current
design, and there are several known problems ranging from image corruption
to driver crashes.
These problems come down to two design problems in the interaction of
threads and HWAccel decoding:
(1)
While avcodec prevents parallel decoding and as such simultaneous access
to the hardware accelerator from the decoding threads, it cannot account
for the user code and its access to the hardware surfaces and the hardware
itself.
This can result in image corruption or even driver crashes if the
user code locks image surfaces while they are being used by the decoder
threads as reference frames.
The current HWAccel API does not offer any way to ensure exclusive access
to the hardware or the surfaces if frame threading is used.
(2)
Initialization of the HWAccel with frame threads is non-trivial, and many
decoders had and still have issues that cause excess calls to the
get_format callback.
This will potentially cause duplicate HWAccel initialization, which in
extreme cases can even lead to driver crashes if the HWAccel is
re-initialized while the user code is actively accessing the hardware
surfaces associated with it, or lead to image corruption due to lost
reference frames.
While both of these issues are solvable, fixing (1) would at least require
a huge API redesign which would move a lot of complexity into the user
code.
The only reason the combination of frame threads and HWAccel was
considered useful is to allow a seamless fallback to multi-threaded
software decoding if the HWAccel is not available, however the issues
outlined above far outweigh this.
The proper solution for a fallback is to re-open the AVCodecContext with
threading enabled if the HWAccel failed, which is a practice commonly used
by various user applications using avcodec today already.
Reviewed-by: Gwenole Beauchesne <gb.devel@gmail.com>
Reviewed-by: wm4 <nfxjfg@googlemail.com>
Signed-off-by: Hendrik Leppkes <h.leppkes@gmail.com>
It is only used inside libavcodec.
Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
Use the new fields directly instead of the ones from AVPicture.
This removes a layer of indirection which serves no pratical purpose
whatsoever, and will help in removing AVPicture structure completely
later.
Every subtitle encoder/decoder seamlessly points to the new arrays,
so it is possible to deprecate AVSubtitleRect.pict.
Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>
It was merged with the iff_ilbm decoder in commit
929a24efff.
Define AV_CODEC_ID_IFF_BYTERUN1 as AV_CODEC_ID_IFF_ILBM for API
compatibility.
Reviewed-by: Ronald S. Bultje <rsbultje@gmail.com>
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
The generic code in utils.c sets the AVFrame.pkt_dts field from the
packet it was supposedly decoded. This does not have to be true for a
fully asynchronous decoder like mmaldec. It could be overwritten with an
incorrect value. Even if the decoder doesn't determine the DTS (but sets
it to AV_NOPTS_VALUE), it's impossible to determine a correct value in
utils.c.
Decoders can now be marked with FF_CODEC_CAP_SETS_PKT_DTS, in which case
utils.c won't overwrite the field. The decoders are expected to set this
field (even if they only set it to AV_NOPTS_VALUE).
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Assert on `avctx->codec->encode2` to avoid a SEGFAULT on the subsequent
function call.
avcodec_encode_video2() uses a similar assertion.
Calling the wrong function on a stream is a serious inconsistency
which could at other places be potentially dangerous and exploitable,
it is thus safer to stop execution and not continue with such
inconsistency after returning an error.
Commit-message-extended-by commiter
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
All DSD codes have 1 bit per sample.
https://en.wikipedia.org/wiki/Direct_Stream_Digital
Signed-off-by: Ihar A. Tumashyk <itumashyk@gmail.com>
Reviewed-by: Ganesh Ajjanagadde <gajjanag@mit.edu>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
This makes it possible to remove more coded_frame usage without breaking the
publically visible coded_frame
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
This parameter can be used to inform the allocation code about how much
downsizing might occur, and can be used to optimize how to allocate the
packet
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
And prevent eventual compilation failures once the relevant functions
and fields are removed.
Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: James Almer <jamrial@gmail.com>
With side data now being added to every MPEG*/H26* video encoder
in the form of a quality scalar, software which simply stores
the pkt->data to get an ES stream would break.
This also as a side-effect makes the code faster
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
The rationale is that coded_frame was only used to communicate key_frame,
pict_type and quality to the caller, as well as a few other random fields,
in a non predictable, let alone consistent way.
There was agreement that there was no use case for coded_frame, as it is
a full-sized AVFrame container used for just 2-3 int-sized properties,
which shouldn't even belong into the AVCodecContext in the first place.
The appropriate AVPacket flag can be used instead of key_frame, while
quality is exported with the new AVPacketSideData quality factor.
There is no replacement for the other fields as they were unreliable,
mishandled or just not used at all.
Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>
Allocating coded_frame is what most encoders do anyway, so it makes
sense to always allocate and free it in a single place. Moreover a lot
of encoders freed the frame with av_freep() instead of the correct API
av_frame_free().
This bring uniformity to encoder behaviour and prevents applications
from erroneusly accessing this field when not allocated. Additionally
this helps isolating encoders that export information with coded_frame,
and heavily simplifies its deprecation.
Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>
internal.h is difficult to use due to circular dependancies
mem.h is a public header ff_* is not public
Alternative solutions probably are possible too
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Michael Niedermayer
Paul B Mahol
Hendrik Leppkes
Luca Barbato
Vittorio Giovara
Andreas Cadhalpun
Derek Buitenhuis
wm4
Carl Eugen Hoyos
James Almer
lummax
Ronald S. Bultje
Ihar A. Tumashyk