|
|
|
|
ARCH_HEADERS = mathops.h
|
|
|
|
|
|
|
|
|
|
# subsystems
|
|
|
|
|
OBJS-$(CONFIG_AC3DSP) += arm/ac3dsp_init_arm.o \
|
|
|
|
|
arm/ac3dsp_arm.o
|
|
|
|
|
OBJS-$(CONFIG_AUDIODSP) += arm/audiodsp_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_BLOCKDSP) += arm/blockdsp_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_FFT) += arm/fft_init_arm.o \
|
|
|
|
|
arm/fft_fixed_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_FLACDSP) += arm/flacdsp_init_arm.o \
|
|
|
|
|
arm/flacdsp_arm.o
|
|
|
|
|
OBJS-$(CONFIG_FMTCONVERT) += arm/fmtconvert_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_G722DSP) += arm/g722dsp_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_H264CHROMA) += arm/h264chroma_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_H264DSP) += arm/h264dsp_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_H264PRED) += arm/h264pred_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_H264QPEL) += arm/h264qpel_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_HPELDSP) += arm/hpeldsp_init_arm.o \
|
|
|
|
|
arm/hpeldsp_arm.o
|
|
|
|
|
OBJS-$(CONFIG_IDCTDSP) += arm/idctdsp_init_arm.o \
|
|
|
|
|
arm/idctdsp_arm.o \
|
|
|
|
|
arm/jrevdct_arm.o \
|
|
|
|
|
arm/simple_idct_arm.o
|
|
|
|
|
OBJS-$(CONFIG_LLAUDDSP) += arm/lossless_audiodsp_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_ME_CMP) += arm/me_cmp_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_MPEGAUDIODSP) += arm/mpegaudiodsp_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_MPEGVIDEO) += arm/mpegvideo_arm.o
|
|
|
|
|
OBJS-$(CONFIG_MPEGVIDEOENC) += arm/mpegvideoencdsp_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_NEON_CLOBBER_TEST) += arm/neontest.o
|
|
|
|
|
OBJS-$(CONFIG_PIXBLOCKDSP) += arm/pixblockdsp_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_RDFT) += arm/rdft_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_RV34DSP) += arm/rv34dsp_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_VC1DSP) += arm/vc1dsp_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_VIDEODSP) += arm/videodsp_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_VP3DSP) += arm/vp3dsp_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_VP8DSP) += arm/vp8dsp_init_arm.o
|
|
|
|
|
|
|
|
|
|
# decoders/encoders
|
|
|
|
|
OBJS-$(CONFIG_AAC_DECODER) += arm/aacpsdsp_init_arm.o \
|
|
|
|
|
arm/sbrdsp_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_DCA_DECODER) += arm/synth_filter_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_HEVC_DECODER) += arm/hevcdsp_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_MLP_DECODER) += arm/mlpdsp_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_RV40_DECODER) += arm/rv40dsp_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_VORBIS_DECODER) += arm/vorbisdsp_init_arm.o
|
|
|
|
|
OBJS-$(CONFIG_VP6_DECODER) += arm/vp6dsp_init_arm.o
|
arm: vp9: Add NEON optimizations of VP9 MC functions
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
ffbd1d2b0002576ef0d976a41ff959c635373fdc,
392caa65df3efa8b2d48a80f08a6af4892c61c08,
557c1675cf0e803b2fee43b4c8b58433842c84d0 and
11623217e3c9b859daee544e31acdd0821b61039.
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
8 years ago
|
|
|
OBJS-$(CONFIG_VP9_DECODER) += arm/vp9dsp_init_arm.o
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
# ARMv5 optimizations
|
|
|
|
|
# subsystems
|
|
|
|
|
ARMV5TE-OBJS-$(CONFIG_IDCTDSP) += arm/idctdsp_init_armv5te.o \
|
|
|
|
|
arm/simple_idct_armv5te.o
|
|
|
|
|
ARMV5TE-OBJS-$(CONFIG_MPEGVIDEO) += arm/mpegvideo_armv5te.o \
|
|
|
|
|
arm/mpegvideo_armv5te_s.o
|
|
|
|
|
ARMV5TE-OBJS-$(CONFIG_VIDEODSP) += arm/videodsp_init_armv5te.o \
|
|
|
|
|
arm/videodsp_armv5te.o
|
|
|
|
|
|
|
|
|
|
# decoders/encoders
|
|
|
|
|
ARMV5TE-OBJS-$(CONFIG_MLP_DECODER) += arm/mlpdsp_armv5te.o
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
# ARMv6 optimizations
|
|
|
|
|
# subsystems
|
|
|
|
|
ARMV6-OBJS-$(CONFIG_AC3DSP) += arm/ac3dsp_armv6.o
|
|
|
|
|
ARMV6-OBJS-$(CONFIG_HPELDSP) += arm/hpeldsp_init_armv6.o \
|
|
|
|
|
arm/hpeldsp_armv6.o
|
|
|
|
|
ARMV6-OBJS-$(CONFIG_IDCTDSP) += arm/idctdsp_init_armv6.o \
|
|
|
|
|
arm/idctdsp_armv6.o \
|
|
|
|
|
arm/simple_idct_armv6.o
|
|
|
|
|
ARMV6-OBJS-$(CONFIG_ME_CMP) += arm/me_cmp_armv6.o
|
|
|
|
|
ARMV6-OBJS-$(CONFIG_MPEGAUDIODSP) += arm/mpegaudiodsp_fixed_armv6.o
|
|
|
|
|
ARMV6-OBJS-$(CONFIG_MPEGVIDEOENC) += arm/mpegvideoencdsp_armv6.o
|
|
|
|
|
ARMV6-OBJS-$(CONFIG_PIXBLOCKDSP) += arm/pixblockdsp_armv6.o
|
|
|
|
|
ARMV6-OBJS-$(CONFIG_STARTCODE) += arm/startcode_armv6.o
|
|
|
|
|
ARMV6-OBJS-$(CONFIG_VP8DSP) += arm/vp8_armv6.o \
|
|
|
|
|
arm/vp8dsp_init_armv6.o \
|
|
|
|
|
arm/vp8dsp_armv6.o
|
|
|
|
|
|
|
|
|
|
# decoders/encoders
|
|
|
|
|
ARMV6-OBJS-$(CONFIG_MLP_DECODER) += arm/mlpdsp_armv6.o
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
# VFP optimizations
|
|
|
|
|
|
|
|
|
|
# subsystems
|
|
|
|
|
VFP-OBJS-$(CONFIG_FFT) += arm/fft_vfp.o
|
|
|
|
|
VFP-OBJS-$(CONFIG_FMTCONVERT) += arm/fmtconvert_vfp.o
|
|
|
|
|
VFP-OBJS-$(CONFIG_MDCT) += arm/mdct_vfp.o
|
|
|
|
|
|
|
|
|
|
# decoders/encoders
|
|
|
|
|
VFP-OBJS-$(CONFIG_DCA_DECODER) += arm/synth_filter_vfp.o
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
# NEON optimizations
|
|
|
|
|
|
|
|
|
|
# subsystems
|
|
|
|
|
NEON-OBJS-$(CONFIG_AC3DSP) += arm/ac3dsp_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_AUDIODSP) += arm/audiodsp_init_neon.o \
|
|
|
|
|
arm/audiodsp_neon.o \
|
|
|
|
|
arm/int_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_BLOCKDSP) += arm/blockdsp_init_neon.o \
|
|
|
|
|
arm/blockdsp_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_FFT) += arm/fft_neon.o \
|
|
|
|
|
arm/fft_fixed_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_FMTCONVERT) += arm/fmtconvert_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_G722DSP) += arm/g722dsp_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_H264CHROMA) += arm/h264cmc_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_H264DSP) += arm/h264dsp_neon.o \
|
|
|
|
|
arm/h264idct_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_H264PRED) += arm/h264pred_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_H264QPEL) += arm/h264qpel_neon.o \
|
|
|
|
|
arm/hpeldsp_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_HPELDSP) += arm/hpeldsp_init_neon.o \
|
|
|
|
|
arm/hpeldsp_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_IDCTDSP) += arm/idctdsp_init_neon.o \
|
|
|
|
|
arm/idctdsp_neon.o \
|
|
|
|
|
arm/simple_idct_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_MDCT) += arm/mdct_neon.o \
|
|
|
|
|
arm/mdct_fixed_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_MPEGVIDEO) += arm/mpegvideo_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_RDFT) += arm/rdft_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_VC1DSP) += arm/vc1dsp_init_neon.o \
|
|
|
|
|
arm/vc1dsp_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_VP3DSP) += arm/vp3dsp_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_VP8DSP) += arm/vp8dsp_init_neon.o \
|
|
|
|
|
arm/vp8dsp_neon.o
|
|
|
|
|
|
|
|
|
|
# decoders/encoders
|
|
|
|
|
NEON-OBJS-$(CONFIG_AAC_DECODER) += arm/aacpsdsp_neon.o \
|
|
|
|
|
arm/sbrdsp_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_LLAUDDSP) += arm/lossless_audiodsp_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_DCA_DECODER) += arm/synth_filter_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_HEVC_DECODER) += arm/hevcdsp_init_neon.o \
|
|
|
|
|
arm/hevcdsp_deblock_neon.o \
|
|
|
|
|
arm/hevcdsp_idct_neon.o \
|
|
|
|
|
arm/hevcdsp_qpel_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_RV30_DECODER) += arm/rv34dsp_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_RV40_DECODER) += arm/rv34dsp_neon.o \
|
|
|
|
|
arm/rv40dsp_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_VORBIS_DECODER) += arm/vorbisdsp_neon.o
|
|
|
|
|
NEON-OBJS-$(CONFIG_VP6_DECODER) += arm/vp6dsp_neon.o
|
arm: vp9: Add NEON itxfm routines
This work is sponsored by, and copyright, Google.
For the transforms up to 8x8, we can fit all the data (including
temporaries) in registers and just do a straightforward transform
of all the data. For 16x16, we do a transform of 4x16 pixels in
4 slices, using a temporary buffer. For 32x32, we transform 4x32
pixels at a time, in two steps of 4x16 pixels each.
Examples of relative speedup compared to the C version, from checkasm:
Cortex A7 A8 A9 A53
vp9_inv_adst_adst_4x4_add_neon: 3.39 5.83 4.17 4.01
vp9_inv_adst_adst_8x8_add_neon: 3.79 4.86 4.23 3.98
vp9_inv_adst_adst_16x16_add_neon: 3.33 4.36 4.11 4.16
vp9_inv_dct_dct_4x4_add_neon: 4.06 6.16 4.59 4.46
vp9_inv_dct_dct_8x8_add_neon: 4.61 6.01 4.98 4.86
vp9_inv_dct_dct_16x16_add_neon: 3.35 3.44 3.36 3.79
vp9_inv_dct_dct_32x32_add_neon: 3.89 3.50 3.79 4.42
vp9_inv_wht_wht_4x4_add_neon: 3.22 5.13 3.53 3.77
Thus, the speedup vs C code is around 3-6x.
This is mostly marginally faster than the corresponding routines
in libvpx on most cores, tested with their 32x32 idct (compared to
vpx_idct32x32_1024_add_neon). These numbers are slightly in libvpx's
favour since their version doesn't clear the input buffer like ours
do (although the effect of that on the total runtime probably is
negligible.)
Cortex A7 A8 A9 A53
vp9_inv_dct_dct_32x32_add_neon: 18436.8 16874.1 14235.1 11988.9
libvpx vpx_idct32x32_1024_add_neon 20789.0 13344.3 15049.9 13030.5
Only on the Cortex A8, the libvpx function is faster. On the other cores,
ours is slightly faster even though ours has got source block clearing
integrated.
This is an adapted cherry-pick from libav commits
a67ae67083151f2f9595a1f2d17b601da19b939e and
52d196fb30fb6628921b5f1b31e7bd11eb7e1d9a.
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
8 years ago
|
|
|
NEON-OBJS-$(CONFIG_VP9_DECODER) += arm/vp9itxfm_neon.o \
|
|
|
|
|
arm/vp9mc_neon.o
|
