[x265] [arm64] add missing p2s functions

[Pop, Sebastian]

Hi,

Part of using the x86 macros to register the functions makes it possible to automatically spot those functions that were not yet implemented.
The attached patch adds the following missing kernels:

chroma_p2s[  2x4]               7.23x    7.77            56.14
chroma_p2s_aligned[  2x4]               7.36x    7.82            57.53
chroma_p2s[  2x8]               9.33x    11.42           106.57
chroma_p2s_aligned[  2x8]               11.46x   11.40           130.65
chroma_p2s[ 2x16]               12.81x   20.87           267.46
chroma_p2s_aligned[ 2x16]               13.18x   20.45           269.52
chroma_p2s[  4x2]               9.27x    5.94            55.11
chroma_p2s_aligned[  4x2]               9.26x    6.05            56.01
chroma_p2s[ 4x32]               28.73x   30.52           876.99
chroma_p2s_aligned[ 4x32]               33.27x   30.85           1026.31
chroma_p2s[  6x8]               15.33x   19.92           305.40
chroma_p2s_aligned[  6x8]               15.41x   19.92           306.91
chroma_p2s[ 6x16]               16.28x   37.47           609.82
chroma_p2s_aligned[ 6x16]               16.53x   37.42           618.63
chroma_p2s[  8x2]               17.47x   5.87            102.58
chroma_p2s_aligned[  8x2]               16.97x   5.95            100.98
chroma_p2s[  8x6]               36.08x   8.35            301.40
chroma_p2s_aligned[  8x6]               35.69x   8.39            299.25
chroma_p2s[ 8x12]               45.62x   13.12           598.47
chroma_p2s_aligned[ 8x12]               46.13x   13.09           603.87
chroma_p2s[ 8x64]               51.21x   62.76           3214.13
chroma_p2s_aligned[ 8x64]               52.85x   62.51           3303.60
chroma_p2s[12x32]               45.45x   52.36           2379.91
chroma_p2s_aligned[12x32]               45.66x   51.82           2365.99
chroma_p2s[16x24]               59.90x   39.38           2358.73
chroma_p2s_aligned[16x24]               59.66x   39.39           2350.24
chroma_p2s[24x64]               59.29x   159.27          9444.02
chroma_p2s_aligned[24x64]               59.44x   159.39          9473.78
chroma_p2s[32x48]               59.84x   155.86          9326.65
chroma_p2s_aligned[32x48]               59.81x   155.81          9318.74

The patch also makes use of the macros to declare the functions.
That makes it much easier to see through the include file.

Ok to commit?

Thanks,
Sebastian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.videolan.org/pipermail/x265-devel/attachments/20210811/5f7ed32c/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-arm64-add-missing-p2s-functions.patch
Type: application/octet-stream
Size: 59458 bytes
Desc: 0001-arm64-add-missing-p2s-functions.patch
URL: <http://mailman.videolan.org/pipermail/x265-devel/attachments/20210811/5f7ed32c/attachment-0001.obj>