[x265] [arm64] port chroma_hps

[Pop, Sebastian]

Patch attached.

From: "Pop, Sebastian" <spop at amazon.com>
Date: Saturday, July 10, 2021 at 12:07 PM
To: "x265-devel at videolan.org" <x265-devel at videolan.org>
Subject: [arm64] port chroma_hps

Hi,
the attached patch ports to arm64 the following kernels:

chroma_hps[  4x4]           6.31x      66.07                    416.85
chroma_hps[  8x8]           10.72x    116.80                  1252.09
chroma_hps[16x16]        11.26x    379.01                  4267.84
chroma_hps[32x32]        11.45x    1363.69               15614.88
chroma_hps[  4x2]           6.03x      50.01                    301.36
chroma_hps[  8x4]           10.64x    76.22                    810.64
chroma_hps[  4x8]           6.30x      103.44                  651.70
chroma_hps[ 16x8]          11.16x    221.79                  2474.83
chroma_hps[ 8x16]          10.85x    200.29                  2173.45
chroma_hps[32x16]        11.37x    744.57                  8466.88
chroma_hps[16x32]        11.32x    688.54                  7793.67
chroma_hps[  8x6]           10.50x    97.37                    1022.70
chroma_hps[  8x2]           10.31x    55.67                    573.87
chroma_hps[16x12]        11.13x    303.26                  3375.54
chroma_hps[12x16]        9.11x      355.26                  3236.57
chroma_hps[ 16x4]          10.96x    143.24                  1570.09
chroma_hps[ 4x16]          6.48x      175.15                  1134.28
chroma_hps[32x24]        11.53x    1042.16               12012.75
chroma_hps[24x32]        11.40x    1027.45               11717.00
chroma_hps[ 32x8]          11.44x    430.85                  4928.72
chroma_hps[ 8x32]          11.03x    364.04                  4014.17
chroma_hps[  4x8]           6.27x      103.33                  647.55
chroma_hps[ 8x16]          10.88x    199.97                  2176.06
chroma_hps[16x32]        11.45x    688.99                  7888.40
chroma_hps[32x64]        11.40x    2619.84               29865.95
chroma_hps[  4x4]           6.33x      66.06                    418.25
chroma_hps[  8x8]           10.65x    116.94                  1245.63
chroma_hps[ 4x16]          6.45x      175.13                  1130.01
chroma_hps[16x16]        11.26x    378.56                  4261.02
chroma_hps[ 8x32]          10.91x    364.85                  3979.70
chroma_hps[32x32]        11.43x    1363.20               15584.13
chroma_hps[16x64]        11.30x    1322.92               14950.73
chroma_hps[ 8x12]          10.85x    158.73                  1721.85
chroma_hps[  8x4]           10.38x    76.24                    791.71
chroma_hps[16x24]        11.23x    538.16                  6042.79
chroma_hps[12x32]        9.08x      648.79                  5890.13
chroma_hps[ 16x8]          11.09x    222.01                  2461.67
chroma_hps[ 4x32]          6.52x      314.81                  2052.62
chroma_hps[32x48]        11.47x    1997.75               22915.55
chroma_hps[24x64]        11.56x    1955.02               22606.40
chroma_hps[32x16]        11.44x    741.23                  8479.07
chroma_hps[ 8x64]          11.17x    694.65                  7762.46
chroma_hps[  4x4]           6.35x      66.07                    419.64
chroma_hps[  8x8]           10.71x    116.78                  1251.18
chroma_hps[16x16]        11.23x    379.59                  4263.16
chroma_hps[32x32]        11.42x    1363.94               15576.00
chroma_hps[64x64]        11.48x    5146.96               59068.42
chroma_hps[  8x4]           10.55x    76.22                    803.90
chroma_hps[  4x8]           6.32x      103.44                  653.89
chroma_hps[ 16x8]          11.14x    221.50                  2466.70
chroma_hps[ 8x16]          10.86x    200.24                  2175.38
chroma_hps[32x16]        11.45x    741.51                  8488.24
chroma_hps[16x32]        11.39x    688.18                  7839.09
chroma_hps[64x32]        11.43x    2681.22               30655.33
chroma_hps[32x64]        11.48x    2619.91               30067.04
chroma_hps[16x12]        11.11x    303.26                  3369.39
chroma_hps[12x16]        9.09x      355.63                  3231.97
chroma_hps[ 16x4]          10.89x    143.23                  1559.78
chroma_hps[ 4x16]          6.49x      174.79                  1134.09
chroma_hps[32x24]        11.54x    1041.71               12024.53
chroma_hps[24x32]        11.51x    1027.19               11823.64
chroma_hps[ 32x8]          11.44x    430.82                  4929.02
chroma_hps[ 8x32]          10.99x    364.83                  4011.04
chroma_hps[64x48]        11.60x    3879.34               45011.70
chroma_hps[48x64]        11.61x    3828.37               44464.40
chroma_hps[64x16]        11.50x    1461.87               16816.20
chroma_hps[16x64]        11.21x    1332.53               14938.27

Ok to commit?

Thanks,
Sebastian

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.videolan.org/pipermail/x265-devel/attachments/20210710/fde38310/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-arm64-port-chroma_hps.patch
Type: application/octet-stream
Size: 25122 bytes
Desc: 0001-arm64-port-chroma_hps.patch
URL: <http://mailman.videolan.org/pipermail/x265-devel/attachments/20210710/fde38310/attachment-0001.obj>