ccca62ef99 added new VP9 VDPAU profiles
and as a consequence AV_PIX_FMT_VDPAU can now be twice in the list of
pixel formats used for format negotiation by ff_thread_get_format(); yet
there is only one entry in said list reserved for VDPAU, leading to a
stack-buffer overflow. This commit fixes this by making sure that
AV_PIX_FMT_VDPAU will not occur twice in said list.
Fixes Coverity ticket 1468046.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>
Added VDPAU to list of supported formats for VP9 420 10 and 12 bit
formats. Add VP9 10/12 Bit support for VDPAU
Signed-off-by: Philip Langdale <philipl@overt.org>
With the description in frame size with refs semantics (SPEC 7.2.5),
it is a requirement of bitstream conformance that for at least one
reference frame has the valid dimensions.
Modify the check to make sure the decoder works well in SINGLE_REFERENCE
mode that not all reference frames have valid dimensions.
Check and error out if invalid reference frame is used in inter_recon.
One of the failure case is a 480x272 inter frame (SINGLE_REFERENCE mode)
with following reference pool:
0. 960x544 LAST valid
1. 1920x1088 GOLDEN invalid, but not used in single reference mode
2. 1920x1088 ALTREF invalid, but not used in single reference mode
3~7 ... Unused
Identical logic in libvpx:
<https://github.com/webmproject/libvpx/blob/master/vp9/decoder/vp9_decodeframe.c#L736>
Signed-off-by: Linjie Fu <linjie.fu@intel.com>
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
The current design, where
- proper init is called for the first per-thread context
- first thread's private data is copied into private data for all the
other threads
- a "fixup" function is called for all the other threads to e.g.
allocate dynamically allocated data
is very fragile and hard to follow, so it is abandoned. Instead, the
same init function is used to init each per-thread context. Where
necessary, AVCodecInternal.is_copy can be used to differentiate between
the first thread and the other ones (e.g. for decoding the extradata
just once).
Support for VDPAU accelerated VP9 decoding was added with libvdpau-1.3.
Support for the same in ffmpeg is added with this patch. Profiles
related to VDPAU VP9 can be found in latest vdpau.h present in
libvdpau-1.3. DRC clips are not supported yet due to
http://trac.ffmpeg.org/ticket/8068
Add VP9 VDPAU to list of hwaccels and supported formats
Added file vdpau_vp9.c and Modified configure to add VDPAU VP9 support.
Mapped VP9 profiles to VDPAU VP9 profiles. Populated the codec specific
params that need to be passed to VDPAU.
Signed-off-by: Philip Langdale <philipl@overt.org>
This commit uses smaller types for some static const arrays to reduce
their size in case the entries can be represented in the smaller type.
The biggest savings came from inv_map_table in vp9.c.
Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>
Signed-off-by: James Almer <jamrial@gmail.com>
webm usually has invisible superframes merged with normal frames.
(vpxenc muxes them in this form, which is evidence enough that this is
the standard webm packet format. It's rather unclear whether ffmpeg is
even allowed to remux them with split packets.)
The vp9 decoder needs them to be in separate packets for multithreading
to work. Add the BSF to the decoder, so the conversion happens
automatically.
This contains the important part of fa1749dd34, which
was apparently skipped in commit d417e95af7. This restores Libav API
compatibility.
This includes a pointer to the associated hwaccel for decoders using
hwaccels - these will be used later to implement the hwaccel setup
without needing a global list.
Also added is a new file listing all hwaccels as external declarations -
this will be used later to generate the hwaccel list at configure time.
this is undocumented in the vp9 bitstream and decoding specification
doc, but matches libvpx
Reviewed-by: "Ronald S. Bultje" <rsbultje@gmail.com>
Signed-off-by: James Zern <jzern@google.com>
This also adds support to avconv (which is trivial due to the new
hwaccel API being generic enough).
The new decoder setup code in dxva2.c is significantly based on work by
Steve Lhomme <robux4@gmail.com>, but with heavy changes/rewrites.
Merges Libav commit f9e7a2f95a.
Also adds untested VP9 support.
The check for DXVA2 COBJs is removed. Just update your MinGW to
something newer than a 5 year old release.
Signed-off-by: Diego Biurrun <diego@biurrun.de>
Fixes: timeout in 730/clusterfuzz-testcase-5265113739165696 (part 1 of 2)
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/targets/ffmpeg
Reviewed-by: BBB
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
This work is sponsored by, and copyright, Google.
These are ported from the ARM version; it is essentially a 1:1
port with no extra added features, but with some hand tuning
(especially for the plain copy/avg functions). The ARM version
isn't very register starved to begin with, so there's not much
to be gained from having more spare registers here - we only
avoid having to clobber callee-saved registers.
Examples of runtimes vs the 32 bit version, on a Cortex A53:
ARM AArch64
vp9_avg4_neon: 27.2 23.7
vp9_avg8_neon: 56.5 54.7
vp9_avg16_neon: 169.9 167.4
vp9_avg32_neon: 585.8 585.2
vp9_avg64_neon: 2460.3 2294.7
vp9_avg_8tap_smooth_4h_neon: 132.7 125.2
vp9_avg_8tap_smooth_4hv_neon: 478.8 442.0
vp9_avg_8tap_smooth_4v_neon: 126.0 93.7
vp9_avg_8tap_smooth_8h_neon: 241.7 234.2
vp9_avg_8tap_smooth_8hv_neon: 690.9 646.5
vp9_avg_8tap_smooth_8v_neon: 245.0 205.5
vp9_avg_8tap_smooth_64h_neon: 11273.2 11280.1
vp9_avg_8tap_smooth_64hv_neon: 22980.6 22184.1
vp9_avg_8tap_smooth_64v_neon: 11549.7 10781.1
vp9_put4_neon: 18.0 17.2
vp9_put8_neon: 40.2 37.7
vp9_put16_neon: 97.4 99.5
vp9_put32_neon/armv8: 346.0 307.4
vp9_put64_neon/armv8: 1319.0 1107.5
vp9_put_8tap_smooth_4h_neon: 126.7 118.2
vp9_put_8tap_smooth_4hv_neon: 465.7 434.0
vp9_put_8tap_smooth_4v_neon: 113.0 86.5
vp9_put_8tap_smooth_8h_neon: 229.7 221.6
vp9_put_8tap_smooth_8hv_neon: 658.9 621.3
vp9_put_8tap_smooth_8v_neon: 215.0 187.5
vp9_put_8tap_smooth_64h_neon: 10636.7 10627.8
vp9_put_8tap_smooth_64hv_neon: 21076.8 21026.9
vp9_put_8tap_smooth_64v_neon: 9635.0 9632.4
These are generally about as fast as the corresponding ARM
routines on the same CPU (at least on the A53), in most cases
marginally faster.
The speedup vs C code is pretty much the same as for the 32 bit
case; on the A53 it's around 6-13x for ther larger 8tap filters.
The exact speedup varies a little, since the C versions generally
don't end up exactly as slow/fast as on 32 bit.
This is an adapted cherry-pick from libav commit
383d96aa22.
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
This work is sponsored by, and copyright, Google.
The filter coefficients are signed values, where the product of the
multiplication with one individual filter coefficient doesn't
overflow a 16 bit signed value (the largest filter coefficient is
127). But when the products are accumulated, the resulting sum can
overflow the 16 bit signed range. Instead of accumulating in 32 bit,
we accumulate the largest product (either index 3 or 4) last with a
saturated addition.
(The VP8 MC asm does something similar, but slightly simpler, by
accumulating each half of the filter separately. In the VP9 MC
filters, each half of the filter can also overflow though, so the
largest component has to be handled individually.)
Examples of relative speedup compared to the C version, from checkasm:
Cortex A7 A8 A9 A53
vp9_avg4_neon: 1.71 1.15 1.42 1.49
vp9_avg8_neon: 2.51 3.63 3.14 2.58
vp9_avg16_neon: 2.95 6.76 3.01 2.84
vp9_avg32_neon: 3.29 6.64 2.85 3.00
vp9_avg64_neon: 3.47 6.67 3.14 2.80
vp9_avg_8tap_smooth_4h_neon: 3.22 4.73 2.76 4.67
vp9_avg_8tap_smooth_4hv_neon: 3.67 4.76 3.28 4.71
vp9_avg_8tap_smooth_4v_neon: 5.52 7.60 4.60 6.31
vp9_avg_8tap_smooth_8h_neon: 6.22 9.04 5.12 9.32
vp9_avg_8tap_smooth_8hv_neon: 6.38 8.21 5.72 8.17
vp9_avg_8tap_smooth_8v_neon: 9.22 12.66 8.15 11.10
vp9_avg_8tap_smooth_64h_neon: 7.02 10.23 5.54 11.58
vp9_avg_8tap_smooth_64hv_neon: 6.76 9.46 5.93 9.40
vp9_avg_8tap_smooth_64v_neon: 10.76 14.13 9.46 13.37
vp9_put4_neon: 1.11 1.47 1.00 1.21
vp9_put8_neon: 1.23 2.17 1.94 1.48
vp9_put16_neon: 1.63 4.02 1.73 1.97
vp9_put32_neon: 1.56 4.92 2.00 1.96
vp9_put64_neon: 2.10 5.28 2.03 2.35
vp9_put_8tap_smooth_4h_neon: 3.11 4.35 2.63 4.35
vp9_put_8tap_smooth_4hv_neon: 3.67 4.69 3.25 4.71
vp9_put_8tap_smooth_4v_neon: 5.45 7.27 4.49 6.52
vp9_put_8tap_smooth_8h_neon: 5.97 8.18 4.81 8.56
vp9_put_8tap_smooth_8hv_neon: 6.39 7.90 5.64 8.15
vp9_put_8tap_smooth_8v_neon: 9.03 11.84 8.07 11.51
vp9_put_8tap_smooth_64h_neon: 6.78 9.48 4.88 10.89
vp9_put_8tap_smooth_64hv_neon: 6.99 8.87 5.94 9.56
vp9_put_8tap_smooth_64v_neon: 10.69 13.30 9.43 14.34
For the larger 8tap filters, the speedup vs C code is around 5-14x.
This is significantly faster than libvpx's implementation of the same
functions, at least when comparing the put_8tap_smooth_64 functions
(compared to vpx_convolve8_horiz_neon and vpx_convolve8_vert_neon from
libvpx).
Absolute runtimes from checkasm:
Cortex A7 A8 A9 A53
vp9_put_8tap_smooth_64h_neon: 20150.3 14489.4 19733.6 10863.7
libvpx vpx_convolve8_horiz_neon: 52623.3 19736.4 21907.7 25027.7
vp9_put_8tap_smooth_64v_neon: 14455.0 12303.9 13746.4 9628.9
libvpx vpx_convolve8_vert_neon: 42090.0 17706.2 17659.9 16941.2
Thus, on the A9, the horizontal filter is only marginally faster than
libvpx, while our version is significantly faster on the other cores,
and the vertical filter is significantly faster on all cores. The
difference is especially large on the A7.
The libvpx implementation does the accumulation in 32 bit, which
probably explains most of the differences.
This is an adapted cherry-pick from libav commits
ffbd1d2b00,
392caa65df,
557c1675cf and
11623217e3.
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
This is intended to workaround bug "665 Integer Divide Instruction May
Cause Unpredictable Behavior" on some early AMD CPUs, which causes a
div-by-zero in this codepath, such as reported in Mozilla bug #1293996.
Note that this isn't guaranteed to fix the bug, since a compiler is free
to reorder instructions that don't depend on each other. However, it
appears to fix the bug in Firefox, and a similar patch was applied to
libvpx also (see Chrome bug #599899).
We need more information from last/cur_frame than from reference
buffers, so we can use a simplified structure for reference buffers,
and then store mvs and segmentation map information in last/cur.
This prepares the decoder for frame threading support.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
According to avcodec.h, avcodec_decode_video2 should return the number of
bytes used if a frame was decoded.
The current implementation returns size - used size of all the subframes.
This fixes the VLC's bug https://trac.videolan.org/vlc/ticket/16836.
The superframe is always fully consumed.
Signed-off-by: Diego Biurrun <diego@biurrun.de>
Andreas Rheinhardt
ManojGuptaBonda
Anton Khirnov
Linjie Fu
Mark Thompson
Peter Ross
James Almer
Michael Niedermayer
Hendrik Leppkes
wm4
Timo Rothenpieler
Ronald S. Bultje
Ilia Valiakhmetov
James Zern
Clément Bœsch
Mathieu Velten
Martin Storsjö
Denis Charmet