[x265] [arm64] Status and combined patch

Fri Aug 6 15:27:36 UTC 2021

Hi,
I combined all the patches into one diff file.
This will make it easier for people to review and to try the current arm64 port.

I am currently finishing porting all the functions from the arm32/ port.
After that I will be working on porting functions from x86/.
Please see below the status of the port to arm64.

arm/blockcopy8.S
[x] blockcopy_{sp,ps,ss}_{4x4,8x8,16x16,32x32,64x64} - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013327.html
[x] chroma blockcopy_{sp,ps,ss}_{4x8,8x16,16x32,32x64} - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013327.html
[x] blockfill_s_{4x4,8x8,16x16,32x32} - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013327.html
[x] copy_cnt_{4,8,16,32} - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013327.html
[x] count_nonzero_{4,8,16,32} - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013327.html

arm/dct-a.S
[ ] dct_{4x4,8x8,16x16}

arm/ipfilter8.S
[x] filterPixelToShort_{4x{4,8,16},8x{4,8,16,32},12x16,16x{4,8,12,16,32,64},24x32,32x{8,16,24,32,64},64x{16,32,48,64},48x64} - https://mailman.videolan.org/pipermail/x265-devel/2021-June/013282.html
[x] LUMA_{VPP_4xN,VPP} - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013301.html
[x] LUMA_{VSP_4xN,VSP} - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013307.html
[x] LUMA_{VPS_4xN,VPS} - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013308.html
[x] CHROMA_VPP - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013309.html
[x] CHROMA_VPS - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013310.html
[x] CHROMA_VSP - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013311.html
[x] LUMA_HPP - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013312.html
[x] LUMA_HPS - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013314.html
[x] CHROMA_HPP - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013315.html
[x] CHROMA_HPS - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013317.html
[x] LUMA_HVPP - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013319.html

x86/ipfilter8.asm
[ ] LUMA_VSS
[ ] CHROMA_VSS

arm/mc-a.S
[x] blockcopy_pp - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013328.html
[x] pixel_avg_pp_{4,8}
[x] pixel_avg_pp_{12,16,24,32,48,64} - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013334.html
[x] cpy2Dto1D_shr - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013341.html
[x] addAvg_{4,6,12,16,24,32,48,64} - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013342.html

arm/pixel-util.S
[x] pixel_var_{8,16,32,64} - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013348.html
[x] getResidual{4,8,16,32} - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013349.html
[x] pixel_sub_ps_{4,8,16,32,64} - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013350.html
[x] chroma sub_ps_{4,8,16,32} - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013350.html
[x] pixel_add_ps_{4,8,16,32,64} - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013351.html
[x] chroma add_ps_{4,8} - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013351.html
[x] scale1D_{128,64} - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013352.html
[x] pixel_planecopy_cp - https://mailman.videolan.org/pipermail/x265-devel/2021-August/013365.html
[x] pixel_satd_{4x4,4x8,4x16,12x16,12x32,8x4,8x8} - https://mailman.videolan.org/pipermail/x265-devel/2021-August/013366.html
[x] pixel_satd_{8x{12,16,32,64},16x{4,8,12,16,24,32,64},24x{32,64},32x{8,16,24,32,48,64},64x{16,32,48,64},48x64} - https://mailman.videolan.org/pipermail/x265-devel/2021-August/013366.html
[x] satd_{8x8,16x4} - https://mailman.videolan.org/pipermail/x265-devel/2021-August/013366.html
[x] satd_8x4v_8x8h - https://mailman.videolan.org/pipermail/x265-devel/2021-August/013366.html
[x] sa8d_satd_8x8 - https://mailman.videolan.org/pipermail/x265-devel/2021-August/013366.html
[x] pixel_sa8d_{8x{8,16},16x{16,32},32x{32,64},64x64} - https://mailman.videolan.org/pipermail/x265-devel/2021-August/013366.html
[x] quant - https://mailman.videolan.org/pipermail/x265-devel/2021-August/013369.html
[x] nquant - https://mailman.videolan.org/pipermail/x265-devel/2021-August/013369.html
[x] dequant_{scaling,normal} - https://mailman.videolan.org/pipermail/x265-devel/2021-August/013369.html
[ ] ssim_4x4x2_core
[x] psyCost_4x4
[ ] psyCost_8x8

arm/sad-a.S
[x] SAD{4,8,16,32,64} - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013321.html
[x] pixel_sad_{24x32,48x64,12x16} - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013321.html
[x] SAD_X{3,4}_{4xN,16xN,16xN_LOOP,32xN,64xN,48x64,24x32,12x16,} - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013325.html
[x] SAD_X_8xN

arm/ssd-a.S
[ ] pixel_sse_pp_{4x4,8x8,16x16,32x32,64x64}
[ ] pixel_sse_ss_{4x4,8x8,16x16,32x32,64x64}
[ ] pixel_ssd_s_{4x4,8x8,16x16,32x32}

x86/blockcopy8.h
[x] cpy2Dto1D_shl - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013341.html
[x] cpy1Dto2D_{shl,shr} - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013341.html
[x] copy_cnt - https://mailman.videolan.org/pipermail/x265-devel/2021-July/013327.html

x86/dct8.asm
[ ] idct{8,16,32}
[ ] dst4
[ ] idst4
[ ] denoise_dct
[ ] nonPsyRdoQuant
[ ] psyRdoQuant
[ ] psyRdoQuant_{1,2}p

x86/intrapred{8,16}.asm
[ ] all_angs_pred
[ ] intra_filter
[ ] intra_pred_{ang,dc,planar}

x86/loopfilter.h
[ ] saoCu{Org,Stats}E{0,1,2,3}
[ ] saoCu{Org,Stats}B0
[ ] calSign
[ ] pelFilter{LumaStrong,Chroma}_{V,H}

x86/mc.h
[ ] frame_init_lowres_core
[ ] mbtree_propagate_cost
[ ] cutree_fix8_{pack,unpack}

x86/pixel-a.asm
[ ] ssimDist

x86/pixel-util8.h
[ ] getResidual{_aligned}
[ ] transpose
[ ] weight_{pp,sp}
[ ] scale1D_{128to64,64to32}
[ ] costCoeffRemain
[ ] costC1C2Flag
[ ] pixel_ssim_4x4x2_core
[ ] pixel_ssim_end4
[ ] scanPosLast
[ ] findPosFirstLast
[ ] costCoeff

x86/pixel.h
[ ] downShift_16
[ ] upShift_{8,16}
[ ] planeClipAndMax
[ ] pixel_{ssd,sa8d,sad_x{3,4},avg,add_ps,sub_ps,satd,sad,ssd,addAvg,ssd_s,var,psyCost_{pp,ss},ssimDist,normFact}

x86/seaintegral.h
[ ] integral{4,8,12,16,24,32}{v,h}
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.videolan.org/pipermail/x265-devel/attachments/20210806/58ead4bb/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: a.diff.gz
Type: application/x-gzip
Size: 36517 bytes
Desc: a.diff.gz
URL: <http://mailman.videolan.org/pipermail/x265-devel/attachments/20210806/58ead4bb/attachment-0001.bin>