This avoids reloading them if they haven't been clobbered, if the
first pass also was idct.
This is similar to what was done in the aarch64 version.
This is cherrypicked from libav commit
3c87039a404c5659ae9bf7454a04e186532eb40b.
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Since the same parameter is used for both input and output,
the name inout is more fitting.
This matches the naming used below in the dmbutterfly macro.
This is cherrypicked from libav commit
79566ec8c77969d5f9be533de04b1349834cca62.
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
The clobbering tests in checkasm are only invoked when testing
correctness, so this bug didn't show up when benchmarking the
dc-only version.
This is cherrypicked from libav commit
4d960a11855f4212eb3a4e470ce890db7f01df29.
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
This is one instruction less for thumb, and only have got
1/2 arm/thumb specific instructions.
This is cherrypicked from libav commit
e5b0fc170f85b00f7dd0ac514918fb5c95253d39.
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
The latter is 1 cycle faster on a cortex-53 and since the operands are
bytewise (or larger) bitmask (impossible to overflow to zero) both are
equivalent.
This is cherrypicked from libav commit
e7ae8f7a715843a5089d18e033afb3ee19ab3057.
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Since aarch64 has enough free general purpose registers use them to
branch to the appropiate storage code. 1-2 cycles faster for the
functions using loop_filter 8/16, ... on a cortex-a53. Mixed results
(up to 2 cycles faster/slower) on a cortex-a57.
This is cherrypicked from libav commit
d7595de0b25e7064fd9e06dea5d0425536cef6dc.
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Seemingly ff_clear_block_sse assumed that the block array is aligned,
so make sure it is.
Fixes ticket #6079
Signed-off-by: James Almer <jamrial@gmail.com>
Fixes compilation with hardcoded tables after eaff1aa09e90e2711207c9463db8bf8e8dec8178
and e71b8119e7db675dd2dac3f7fb069b0df2943c38
Reviewed-by: Timo Rothenpieler <timo@rothenpieler.org>
Signed-off-by: James Almer <jamrial@gmail.com>
This decoder can decode all existing SpeedHQ formats (SHQ0–5, 7, and 9),
including correct decoding of the alpha channel.
1080p is decoded in 142 fps on one core of my i7-4600U (2.1 GHz Haswell),
about evenly split between bitstream reader and IDCT. There is currently
no attempt at slice or frame threading, even though the format trivially
supports both.
NewTek very helpfully provided a full set of SHQ samples, as well as
source code for an SHQ2 encoder (not included) and assistance with
understanding some details of the format.
This is what gimp, ImageMagick and FreeImage do and what the
Adobe Photoshop file format specification suggests.
Fixes a sample from ticket #6045.
Reviewed-by: Martin Vignali
cid: 1396268
when av_strdup(str) error, the lst need release
Reviewed-by: James Almer <jamrial@gmail.com>
Signed-off-by: Steven Liu <lq@chinaffmpeg.org>
This should make no difference as the value should not be able to be that large
but its more correct this way
Fixes CID1348138
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Disable B frames when using baseline/constrained baseline profile,
following H.264 spec Annex A.2.1.
Signed-off-by: Jun Zhao <jun.zhao@intel.com>
Signed-off-by: Yi A Wang <yi.a.wang@intel.com>
Signed-off-by: Mark Thompson <sw@jkqxz.net>
looks like there is a bug in commit
1a08758e7c4e14a9ea8d2fef6c33ad411b2d3c40 relating to the handling of
ptr in decode_frame after decode_block is called, before this commit
ptr would have been incremented for each line in the data window, now
after the commit it is left at the start of the first included line
rather than the line after the data window then the code sets the
remaining lines to 0 and thus the whole image is over written.
Fix by adjusting ptr to the correct line after decode_block returns
Signed-off-by: Kevin Wheatley <kevin.j.wheatley@gmail.com>
This commit replaces the current inefficient non-power-of-two FFT with a
much faster FFT based on the Prime Factor Algorithm.
Although it is already much faster than the old algorithm without SIMD,
the new algorithm makes use of the already very throughouly SIMD'd power
of two FFT, which improves performance even more across all platforms
which we have SIMD support for.
Most of the work was done by Peter Barfuss, who passed the code to me to
implement into the iMDCT and the current codebase. The code for a
5-point and 15-point FFT was derived from the previous implementation,
although it was optimized and simplified, which will make its future
SIMD easier. The 15-point FFT is currently using 6% of the current
overall decoder overhead.
The FFT can now easily be used as a forward transform by simply not
multiplying the 5-point FFT's imaginary component by -1 (which comes
from the fact that changing the complex exponential's angle by -1 also
changes the output by that) and by multiplying the "theta" angle of the
main exptab by -1. Hence the deliberately left multiplication by -1 at
the end.
FATE passes, and performance reports on other platforms/CPUs are
welcome.
Performance comparisons:
iMDCT, PFA:
101127 decicycles in speed, 32765 runs, 3 skips
iMDCT, Old:
211022 decicycles in speed, 32768 runs, 0 skips
Standalone FFT, 300000 transforms of size 960:
PFA Old FFT kiss_fft libfftw3f
3.659695s, 15.726912s, 13.300789s, 1.182222s
Being only 3x slower than libfftw3f is a big achievement by itself.
There appears to be something capping the performance in the iMDCT side
of things, possibly during the pre-stage reindexing. However, it is
certainly fast enough for now.
Signed-off-by: Rostislav Pehlivanov <atomnuker@gmail.com>
Prep work for the next commit, which will add a new FFT algorithm
which makes the iMDCT over 3x faster than it is currently (standalone,
the FFT is with some framesizes over 10x faster).
The new FFT algorithm uses the already thouroughly SIMD'd power of two
FFT which already has SIMD for AArch64, so users of that platform will
still see an improvement.
The previous FFT+SIMD was barely 2.5x faster than the C versions on these
platforms.
Signed-off-by: Rostislav Pehlivanov <atomnuker@gmail.com>
This prevents a division by zero crash in wmavoice_decode_packet.
The problem was introduced by commit
3deb4b54a24f8cddce463d9f5751b01efeb976af.
Reviewed-by: Ronald S. Bultje <rsbultje@gmail.com>
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
This fixes triggering the av_assert0(ret <= tmp.size).
The problem was reintroduced by commit
7b27dd5c16de785297ce4de4b88afa0b6685f61d and originally fixed by
2a4700a4f03280fa8ba4fc0f8a9987bb550f0d1e.
Reviewed-by: Ronald S. Bultje <rsbultje@gmail.com>
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
