8915320db9
vp9lpf/x86: make filter_48/84/88_h work on 32-bit.
...
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
725a216481
vp9lpf/x86: make filter_44_h work on 32-bit.
...
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
5bfa96c4b3
vp9lpf/x86: make filter_16_v work on 32-bit.
...
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
b905e8d2fe
vp9lpf/x86: make filter_48/84_v work on 32-bit.
...
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
37637e6590
vp9lpf/x86: make filter_88_v work on 32-bit.
...
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
be10834bd9
vp9lpf/x86: make filter_44_v work on 32-bit.
...
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
7c62891efe
vp9lpf/x86: save one register in SIGN_ADD/SUB.
...
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
c6375a83d1
vp9lpf/x86: store unpacked intermediates for filter6/14 on stack.
...
filter16 goes from 508 to 482 (h) or 346 to 314 (v) cycles; filter88
goes from 240 to 238 (h) or 174 to 165 (v) cycles, measured on TOS.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
4ce8ba72f9
vp9lpf/x86: move variable assigned inside macro branch.
...
The value is not used outside the branch.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
e4961035b2
vp9lpf/x86: simplify ABSSUM_CMP by inverting the comparison meaning.
...
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
683da2788e
vp9lpf/x86: remove unused register from ABSSUB_CMP macro.
...
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
6e74e9636b
vp9lpf/x86: slightly simplify 44/48/84/88 h stores.
...
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
6411c328a2
vp9lpf/x86: make cglobal statement more conservative in register allocation.
...
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
a6e288d624
vp9lpf/x86: save one register in loopfilter surface coverage.
...
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
0ed21bdc9e
vp9lpf/x86: add ff_vp9_loop_filter_[vh]_44_16_{sse2,ssse3,avx}.
...
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
f2e3d706a1
vp9lpf/x86: add ff_vp9_loop_filter_h_{48,84}_16_{sse2,ssse3,avx}().
...
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
92d47550ea
vp9lpf/x86: add an SSE2 version of vp9_loop_filter_[vh]_88_16
...
Similar gains as the ssse3 version once again
Additional improvements by Clément Bœsch <u@pkh.me>.
Signed-off-by: James Almer <jamrial@gmail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
6bea478158
vp9lpf/x86: add ff_vp9_loop_filter_[vh]_88_16_{ssse3,avx}.
...
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
1f451eed60
vp9lpf/x86: add ff_vp9_loop_filter_[vh]_16_16_sse2().
...
Similar gains in performance as the SSSE3 version
Signed-off-by: James Almer <jamrial@gmail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
a692724c58
vp9lpf/x86: add x86 SSSE3/AVX SIMD for vp9_loop_filter_[vh]_16_16.
...
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
42111e8543
avcodec: fix arguments on xmm/neon clobber test wrappers
...
Signed-off-by: James Almer <jamrial@gmail.com>
9 years ago
449f263f9f
avcodec: add missing xmm/neon clobber test wrappers for the new encode API
...
Reviewed-by: Ronald S. Bultje <rsbultje@gmail.com>
Signed-off-by: James Almer <jamrial@gmail.com>
9 years ago
b57e38f52c
ac3dsp: x86: Replace inline asm for in-decoder downmixing with standalone asm
...
Adds a wrapper function for downmixing which detects channel count changes
and updates the selected downmix function accordingly.
Simplification and porting to current x86inc infrastructure by Diego Biurrun.
Signed-off-by: Diego Biurrun <diego@biurrun.de>
9 years ago
43717469f9
ac3dsp: Reverse matrix in/out order in downmix()
...
Also use (float **) instead of (float (*)[2]). This matches the matrix
layout in libavresample so we can reuse assembly code between the two.
Signed-off-by: Diego Biurrun <diego@biurrun.de>
9 years ago
8d1267932c
x86/h264_weight: use appropriate register size for weight parameters
...
This fixes decoding corruption on 64 bit windows.
Signed-off-by: Martin Storsjö <martin@martin.st>
9 years ago
2caa93b813
mpegaudiodsp: Change type of array stride parameters to ptrdiff_t
...
This avoids SIMD-optimized functions having to sign-extend their
stride argument manually to be able to do pointer arithmetic.
9 years ago
e4a94d8b36
h264chroma: Change type of stride parameters to ptrdiff_t
...
This avoids SIMD-optimized functions having to sign-extend their
stride argument manually to be able to do pointer arithmetic.
9 years ago
2ec9fa5ec6
idct: Change type of array stride parameters to ptrdiff_t
...
ptrdiff_t is the correct type for array strides and similar.
9 years ago
009adfd4fb
x86: fpel: Remove unnecessary sign extend
9 years ago
de2ae3c1fa
lavc: add clobber tests for the new encoding/decoding API
9 years ago
5ae0ad001a
x86/h264_weight: use appropriate register size for weight parameters
...
Fixes trac 5579
Reviewed-by: Ronald S. Bultje <rsbultje@gmail.com>
Acked-by: Michael Niedermayer <michael@niedermayer.cc>
9 years ago
bc26fe8927
avcodec/h264: Use ptrdiff_t for (bi)weight functions
...
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
9 years ago
12004a9a7f
audiodsp/x86: yasmify vector_clipf_sse
9 years ago
eea9857bfd
blockdsp: drop the high_bit_depth parameter
...
It has no effect, since the code is supposed to operate the same way for
any bit depth.
9 years ago
683da86aab
audiodsp: reorder arguments for vector_clipf
...
This will make the x86 asm simpler.
ARM conversion by Martin Storsjö <martin@martin.st> and Janne Grunau
<janne-libav@jannau.net>
9 years ago
75d98e30af
audiodsp/x86: clear the high bits of the order parameter on 64bit
...
Also change shl to add, since it can be faster on some CPUs.
CC: libav-stable@libav.org
9 years ago
1d6c76e11f
audiodsp/x86: fix ff_vector_clip_int32_sse2
...
This version, which is the only one doing two processing cycles per loop
iteration, computes the load/store indices incorrectly for the second
cycle.
CC: libav-stable@libav.org
9 years ago
de452e5037
pixblockdsp: Change type of stride parameters to ptrdiff_t
...
This avoids SIMD-optimized functions having to sign-extend their
line size argument manually to be able to do pointer arithmetic.
Also adjust parameter names to be "stride" everywhere.
9 years ago
721d57e608
vp56: Separate VP5 and VP6 dsp initialization
...
VP5 has no arch-specific optimizations (nor will it get some in the
future), so it makes no sense to try to share dsp init code with VP6.
9 years ago
3fd22538bc
prores: Change type of stride parameters to ptrdiff_t
...
This avoids SIMD-optimized functions having to sign-extend their
line size argument manually to be able to do pointer arithmetic.
Also adjust parameter names to be "linesize" everywhere.
9 years ago
f81be06cf6
cavs: Change type of stride parameters to ptrdiff_t
...
ptrdiff_t is the correct type for array strides and similar.
9 years ago
802727b538
vp8: Update some assembly comments left unchanged in bd66f073fe
9 years ago
d9d26a3674
vp56: Change type of stride parameters to ptrdiff_t
...
This avoids SIMD-optimized functions having to sign-extend their
line size argument manually to be able to do pointer arithmetic.
9 years ago
6892df9294
vp3: Change type of stride parameters to ptrdiff_t
...
This avoids SIMD-optimized functions having to sign-extend their
stride argument manually to be able to do pointer arithmetic.
Also adjust parameter names to be "stride" everywhere.
9 years ago
e2b9993558
simple_idct: x86: Drop disabled IDCT implementation
...
This gem has been disabled since 2001.
9 years ago
d950279cbf
avcodec/ttadsp: cosmetics
...
Clean some header includes and use the same naming scheme as
in ttaencdsp
Signed-off-by: James Almer <jamrial@gmail.com>
9 years ago
9790b44a89
vp9mc/x86: sse2 MC assembly.
...
Also a slight change to the ssse3 code, which prevents a theoretical
overflow in the sharp filter.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
67922b4ee4
vp9mc/x86: add AVX and AVX2 MC
...
Roughly 25% faster MC than ssse3 for blocksizes 32 and 64.
Reviewed-by: Ronald S. Bultje <rsbultje@gmail.com>
Signed-off-by: James Almer <jamrial@gmail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
3cda179f18
vp9mc/x86: rename ff_* to ff_vp9_*
...
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
8be8444d01
vp9mc/x86: rename ff_avg[48]_sse to ff_avg[48]_mmxext
...
pavgb is an sse integer instruction, so the mmxext flag is enough
Signed-off-by: James Almer <jamrial@gmail.com>
Reviewed-by: "Ronald S. Bultje" <rsbultje@gmail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
9 years ago
Ronald S. Bultje
Clément Bœsch
James Almer
Justin Ruggles
Hendrik Leppkes
Diego Biurrun
Anton Khirnov
Michael Niedermayer