mirror of
https://github.com/FFmpeg/FFmpeg.git
synced 2024-12-18 03:19:31 +02:00
9f10cff610
This work is sponsored by, and copyright, Google.
This is similar to the arm version, but due to the larger registers
on aarch64, we can do 8 pixels at a time for all filter sizes.
Examples of runtimes vs the 32 bit version, on a Cortex A53:
ARM AArch64
vp9_loop_filter_h_4_8_10bpp_neon: 213.2 172.6
vp9_loop_filter_h_8_8_10bpp_neon: 281.2 244.2
vp9_loop_filter_h_16_8_10bpp_neon: 657.0 444.5
vp9_loop_filter_h_16_16_10bpp_neon: 1280.4 877.7
vp9_loop_filter_mix2_h_44_16_10bpp_neon: 397.7 358.0
vp9_loop_filter_mix2_h_48_16_10bpp_neon: 465.7 429.0
vp9_loop_filter_mix2_h_84_16_10bpp_neon: 465.7 428.0
vp9_loop_filter_mix2_h_88_16_10bpp_neon: 533.7 499.0
vp9_loop_filter_mix2_v_44_16_10bpp_neon: 271.5 244.0
vp9_loop_filter_mix2_v_48_16_10bpp_neon: 330.0 305.0
vp9_loop_filter_mix2_v_84_16_10bpp_neon: 329.0 306.0
vp9_loop_filter_mix2_v_88_16_10bpp_neon: 386.0 365.0
vp9_loop_filter_v_4_8_10bpp_neon: 150.0 115.2
vp9_loop_filter_v_8_8_10bpp_neon: 209.0 175.5
vp9_loop_filter_v_16_8_10bpp_neon: 492.7 345.2
vp9_loop_filter_v_16_16_10bpp_neon: 951.0 682.7
This is significantly faster than the ARM version in almost
all cases except for the mix2 functions.
Based on START_TIMER/STOP_TIMER wrapping around a few individual
functions, the speedup vs C code is around 2-3x.
Signed-off-by: Martin Storsjö <martin@martin.st>
51 lines
2.7 KiB
Makefile
51 lines
2.7 KiB
Makefile
# subsystems
|
|
OBJS-$(CONFIG_FFT) += aarch64/fft_init_aarch64.o
|
|
OBJS-$(CONFIG_FMTCONVERT) += aarch64/fmtconvert_init.o
|
|
OBJS-$(CONFIG_H264CHROMA) += aarch64/h264chroma_init_aarch64.o
|
|
OBJS-$(CONFIG_H264DSP) += aarch64/h264dsp_init_aarch64.o
|
|
OBJS-$(CONFIG_H264PRED) += aarch64/h264pred_init.o
|
|
OBJS-$(CONFIG_H264QPEL) += aarch64/h264qpel_init_aarch64.o
|
|
OBJS-$(CONFIG_HPELDSP) += aarch64/hpeldsp_init_aarch64.o
|
|
OBJS-$(CONFIG_MPEGAUDIODSP) += aarch64/mpegaudiodsp_init.o
|
|
OBJS-$(CONFIG_NEON_CLOBBER_TEST) += aarch64/neontest.o
|
|
OBJS-$(CONFIG_VIDEODSP) += aarch64/videodsp_init.o
|
|
|
|
# decoders/encoders
|
|
OBJS-$(CONFIG_DCA_DECODER) += aarch64/synth_filter_init.o
|
|
OBJS-$(CONFIG_RV40_DECODER) += aarch64/rv40dsp_init_aarch64.o
|
|
OBJS-$(CONFIG_VC1DSP) += aarch64/vc1dsp_init_aarch64.o
|
|
OBJS-$(CONFIG_VORBIS_DECODER) += aarch64/vorbisdsp_init.o
|
|
OBJS-$(CONFIG_VP9_DECODER) += aarch64/vp9dsp_init_10bpp_aarch64.o \
|
|
aarch64/vp9dsp_init_12bpp_aarch64.o \
|
|
aarch64/vp9dsp_init_aarch64.o
|
|
|
|
# ARMv8 optimizations
|
|
|
|
# subsystems
|
|
ARMV8-OBJS-$(CONFIG_VIDEODSP) += aarch64/videodsp.o
|
|
|
|
# NEON optimizations
|
|
|
|
# subsystems
|
|
NEON-OBJS-$(CONFIG_FFT) += aarch64/fft_neon.o
|
|
NEON-OBJS-$(CONFIG_FMTCONVERT) += aarch64/fmtconvert_neon.o
|
|
NEON-OBJS-$(CONFIG_H264CHROMA) += aarch64/h264cmc_neon.o
|
|
NEON-OBJS-$(CONFIG_H264DSP) += aarch64/h264dsp_neon.o \
|
|
aarch64/h264idct_neon.o
|
|
NEON-OBJS-$(CONFIG_H264PRED) += aarch64/h264pred_neon.o
|
|
NEON-OBJS-$(CONFIG_H264QPEL) += aarch64/h264qpel_neon.o \
|
|
aarch64/hpeldsp_neon.o
|
|
NEON-OBJS-$(CONFIG_HPELDSP) += aarch64/hpeldsp_neon.o
|
|
NEON-OBJS-$(CONFIG_MDCT) += aarch64/mdct_neon.o
|
|
NEON-OBJS-$(CONFIG_MPEGAUDIODSP) += aarch64/mpegaudiodsp_neon.o
|
|
|
|
# decoders/encoders
|
|
NEON-OBJS-$(CONFIG_DCA_DECODER) += aarch64/synth_filter_neon.o
|
|
NEON-OBJS-$(CONFIG_VORBIS_DECODER) += aarch64/vorbisdsp_neon.o
|
|
NEON-OBJS-$(CONFIG_VP9_DECODER) += aarch64/vp9itxfm_16bpp_neon.o \
|
|
aarch64/vp9itxfm_neon.o \
|
|
aarch64/vp9lpf_16bpp_neon.o \
|
|
aarch64/vp9lpf_neon.o \
|
|
aarch64/vp9mc_16bpp_neon.o \
|
|
aarch64/vp9mc_neon.o
|