ffmpeg/libavcodec/x86
Nuo Mi 15eb10c6de x86/vvcdec: inter, add optical flow avx2 code
BDoF used about 10%–25% of the CPU for some clips.
Here are the FPS for one run; please ignore the negative values, as they may be due to round-to-round variation

clips                                       | before | after | delta
--------------------------------------------|--------|-------|------
RitualDance_1920x1080_60_10_420_37_RA.266   | 310.0  | 363.0 | 14.60%
NovosobornayaSquare_1920x1080.bin           | 322.3  | 339.7 |  5.12%
Tango2_3840x2160_60_10_420_27_LD.266        |  71.0  | 68.7  | -3.35%
RitualDance_1920x1080_60_10_420_32_LD.266   | 250.0  | 245.3 | -1.92%
Chimera_8bit_1080P_1000_frames.vvc          | 359.3  | 422.7 | 15.00%
BQTerrace_1920x1080_60_10_420_22_RA.vvc     | 142.3  | 147.7 |  3.66%

Reviewed-by: James Almer <jamrial@gmail.com>
2024-08-31 14:06:19 +08:00
..
h26x
vvc x86/vvcdec: inter, add optical flow avx2 code 2024-08-31 14:06:19 +08:00
Makefile
aacencdsp.asm
aacencdsp_init.c
aacpsdsp.asm
aacpsdsp_init.c
ac3dsp.asm
ac3dsp_downmix.asm
ac3dsp_init.c
alacdsp.asm
alacdsp_init.c
audiodsp.asm
audiodsp_init.c
blockdsp.asm
blockdsp_init.c
bswapdsp.asm
bswapdsp_init.c
cabac.h
cavsdsp.c
cavsidct.asm
celt_pvq_init.c
celt_pvq_search.asm
cfhddsp.asm
cfhddsp_init.c
cfhdencdsp.asm
cfhdencdsp_init.c
constants.c
constants.h
dcadsp.asm
dcadsp_init.c
dct32.asm
dirac_dwt.asm
dirac_dwt_init.c
diracdsp.asm
diracdsp_init.c
dnxhdenc.asm
dnxhdenc_init.c
exrdsp.asm
exrdsp_init.c
fdct.c
fdct.h
fdctdsp_init.c
flac_dsp_gpl.asm
flacdsp.asm
flacdsp_init.c
flacencdsp_init.c
fmtconvert.asm
fmtconvert_init.c
fpel.asm
fpel.h
g722dsp.asm
g722dsp_init.c
h263_loopfilter.asm
h263dsp_init.c
h264_cabac.c
h264_chromamc.asm
h264_chromamc_10bit.asm
h264_deblock.asm
h264_deblock_10bit.asm
h264_idct.asm
h264_idct_10bit.asm
h264_intrapred.asm
h264_intrapred_10bit.asm
h264_intrapred_init.c
h264_qpel.c
h264_qpel_8bit.asm
h264_qpel_10bit.asm
h264_weight.asm
h264_weight_10bit.asm
h264chroma_init.c
h264dsp_init.c
hevc_add_res.asm
hevc_deblock.asm
hevc_idct.asm
hevc_mc.asm
hevc_sao.asm
hevc_sao_10bit.asm
hevcdsp.h
hevcdsp_init.c
hpeldsp.asm
hpeldsp.h
hpeldsp_init.c
hpeldsp_rnd_template.c
huffyuvdsp.asm
huffyuvdsp_init.c
huffyuvdsp_template.asm
huffyuvencdsp.asm
huffyuvencdsp_init.c
idctdsp.asm
idctdsp.h
idctdsp_init.c
imdct36.asm
inline_asm.h
jpeg2000dsp.asm
jpeg2000dsp_init.c
lossless_audiodsp.asm
lossless_audiodsp_init.c
lossless_videodsp.asm
lossless_videodsp_init.c
lossless_videoencdsp.asm
lossless_videoencdsp_init.c
lpc.asm
lpc_init.c
mathops.h
me_cmp.asm
me_cmp_init.c
mlpdsp.asm
mlpdsp_init.c
mpeg4videodsp.c
mpegaudiodsp.c
mpegvideo.c
mpegvideoenc.c
mpegvideoenc_qns_template.c
mpegvideoenc_template.c
mpegvideoencdsp.asm
mpegvideoencdsp_init.c avcodec/x86/mpegvideoencdsp: speed up draw_edges_mmx by using memcpy() 2024-08-26 12:50:21 +02:00
opusdsp.asm
opusdsp_init.c
pixblockdsp.asm
pixblockdsp_init.c
pngdsp.asm
pngdsp_init.c
proresdsp.asm
proresdsp_init.c
qpel.asm
qpeldsp.asm
qpeldsp_init.c
rnd_template.c
rv34dsp.asm
rv34dsp_init.c
rv40dsp.asm
rv40dsp_init.c
sbcdsp.asm
sbcdsp_init.c
sbrdsp.asm
sbrdsp_init.c
simple_idct.asm
simple_idct.h
simple_idct10.asm
simple_idct10_template.asm
snowdsp.c
svq1enc.asm
svq1enc_init.c
synth_filter.asm
synth_filter_init.c
takdsp.asm
takdsp_init.c
ttadsp.asm
ttadsp_init.c
ttaencdsp.asm
ttaencdsp_init.c
utvideodsp.asm
utvideodsp_init.c
v210-init.c
v210.asm
v210enc.asm
v210enc_init.c
vc1dsp.h
vc1dsp_init.c
vc1dsp_loopfilter.asm
vc1dsp_mc.asm
vc1dsp_mmx.c
videodsp.asm
videodsp_init.c
vorbisdsp.asm
vorbisdsp_init.c
vp3dsp.asm
vp3dsp_init.c
vp6dsp.asm
vp6dsp_init.c
vp8dsp.asm
vp8dsp_init.c
vp8dsp_loopfilter.asm
vp9dsp_init.c
vp9dsp_init.h
vp9dsp_init_10bpp.c
vp9dsp_init_12bpp.c
vp9dsp_init_16bpp.c
vp9dsp_init_16bpp_template.c
vp9intrapred.asm
vp9intrapred_16bpp.asm
vp9itxfm.asm
vp9itxfm_16bpp.asm
vp9itxfm_template.asm
vp9lpf.asm
vp9lpf_16bpp.asm
vp9mc.asm
vp9mc_16bpp.asm
vpx_arith.h
w64xmmtest.c
xvididct.asm
xvididct.h
xvididct_init.c