mirror of
https://github.com/FFmpeg/FFmpeg.git
synced 2024-12-12 19:18:44 +02:00
dd299a2d6d
This work is sponsored by, and copyright, Google.
The implementation tries to have smart handling of cases
where no pixels need the full filtering for the 8/16 width
filters, skipping both calculation and writeback of the
unmodified pixels in those cases. The actual effect of this
is hard to test with checkasm though, since it tests the
full filtering, and the benefit depends on how many filtered
blocks use the shortcut.
Examples of relative speedup compared to the C version, from checkasm:
Cortex A7 A8 A9 A53
vp9_loop_filter_h_4_8_neon: 2.72 2.68 1.78 3.15
vp9_loop_filter_h_8_8_neon: 2.36 2.38 1.70 2.91
vp9_loop_filter_h_16_8_neon: 1.80 1.89 1.45 2.01
vp9_loop_filter_h_16_16_neon: 2.81 2.78 2.18 3.16
vp9_loop_filter_mix2_h_44_16_neon: 2.65 2.67 1.93 3.05
vp9_loop_filter_mix2_h_48_16_neon: 2.46 2.38 1.81 2.85
vp9_loop_filter_mix2_h_84_16_neon: 2.50 2.41 1.73 2.85
vp9_loop_filter_mix2_h_88_16_neon: 2.77 2.66 1.96 3.23
vp9_loop_filter_mix2_v_44_16_neon: 4.28 4.46 3.22 5.70
vp9_loop_filter_mix2_v_48_16_neon: 3.92 4.00 3.03 5.19
vp9_loop_filter_mix2_v_84_16_neon: 3.97 4.31 2.98 5.33
vp9_loop_filter_mix2_v_88_16_neon: 3.91 4.19 3.06 5.18
vp9_loop_filter_v_4_8_neon: 4.53 4.47 3.31 6.05
vp9_loop_filter_v_8_8_neon: 3.58 3.99 2.92 5.17
vp9_loop_filter_v_16_8_neon: 3.40 3.50 2.81 4.68
vp9_loop_filter_v_16_16_neon: 4.66 4.41 3.74 6.02
The speedup vs C code is around 2-6x. The numbers are quite
inconclusive though, since the checkasm test runs multiple filterings
on top of each other, so later rounds might end up with different
codepaths (different decisions on which filter to apply, based
on input pixel differences). Disabling the early-exit in the asm
doesn't give a fair comparison either though, since the C code
only does the necessary calcuations for each row.
Based on START_TIMER/STOP_TIMER wrapping around a few individual
functions, the speedup vs C code is around 4-9x.
This is pretty similar in runtime to the corresponding routines
in libvpx. (This is comparing vpx_lpf_vertical_16_neon,
vpx_lpf_horizontal_edge_8_neon and vpx_lpf_horizontal_edge_16_neon
to vp9_loop_filter_h_16_8_neon, vp9_loop_filter_v_16_8_neon
and vp9_loop_filter_v_16_16_neon - note that the naming of horizonal
and vertical is flipped between the libraries.)
In order to have stable, comparable numbers, the early exits in both
asm versions were disabled, forcing the full filtering codepath.
Cortex A7 A8 A9 A53
vp9_loop_filter_h_16_8_neon: 597.2 472.0 482.4 415.0
libvpx vpx_lpf_vertical_16_neon: 626.0 464.5 470.7 445.0
vp9_loop_filter_v_16_8_neon: 500.2 422.5 429.7 295.0
libvpx vpx_lpf_horizontal_edge_8_neon: 586.5 414.5 415.6 383.2
vp9_loop_filter_v_16_16_neon: 905.0 784.7 791.5 546.0
libvpx vpx_lpf_horizontal_edge_16_neon: 1060.2 751.7 743.5 685.2
Our version is consistently faster on on A7 and A53, marginally slower on
A8, and sometimes faster, sometimes slower on A9 (marginally slower in all
three tests in this particular test run).
Signed-off-by: Martin Storsjö <martin@martin.st>
145 lines
7.6 KiB
Makefile
145 lines
7.6 KiB
Makefile
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_MDCT) += arm/mdct_init_arm.o \
|
|
arm/mdct_fixed_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_APE_DECODER) += arm/apedsp_init_arm.o
|
|
OBJS-$(CONFIG_DCA_DECODER) += arm/dcadsp_init_arm.o
|
|
OBJS-$(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-$(CONFIG_VP6_DECODER) += arm/vp6dsp_init_arm.o
|
|
OBJS-$(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/dcadsp_vfp.o \
|
|
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_APE_DECODER) += arm/apedsp_neon.o
|
|
NEON-OBJS-$(CONFIG_DCA_DECODER) += arm/dcadsp_neon.o \
|
|
arm/synth_filter_neon.o
|
|
NEON-OBJS-$(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-$(CONFIG_VP6_DECODER) += arm/vp6dsp_neon.o
|
|
NEON-OBJS-$(CONFIG_VP9_DECODER) += arm/vp9itxfm_neon.o \
|
|
arm/vp9lpf_neon.o \
|
|
arm/vp9mc_neon.o
|