[x265] [PATCH] asm: avx2 code for sse_ss[16x16] for 10 bpp (1935 -> 1030)

[Sumalatha Polureddy]

please ignore this patch, will send new patch with changed commit message

Regards
Sumalatha

On Thu, May 21, 2015 at 11:58 AM, <sumalatha at multicorewareinc.com> wrote:

> # HG changeset patch
> # User Sumalatha Polureddy
> # Date 1432189730 -19800
> #      Thu May 21 11:58:50 2015 +0530
> # Node ID f33e89178120b981ae3ec0e717a74bc608741722
> # Parent  dc4fcfc574ade14ecc841797ad08be9753fad58e
> asm: avx2 code for sse_ss[16x16] for 10 bpp (1935 -> 1030)
>
> sse2:
> sse_ss[32x32]  4.71x    1935.46         9113.11
>
> avx2:
> sse_ss[32x32]  8.85x    1030.08         9112.05
>
> diff -r dc4fcfc574ad -r f33e89178120 source/common/x86/asm-primitives.cpp
> --- a/source/common/x86/asm-primitives.cpp      Wed May 20 12:17:44 2015
> -0500
> +++ b/source/common/x86/asm-primitives.cpp      Thu May 21 11:58:50 2015
> +0530
> @@ -1259,6 +1259,7 @@
>
>          p.cu[BLOCK_32x32].ssd_s = x265_pixel_ssd_s_32_avx2;
>          p.cu[BLOCK_16x16].sse_ss = x265_pixel_ssd_ss_16x16_avx2;
> +        p.cu[BLOCK_32x32].sse_ss = x265_pixel_ssd_ss_32x32_avx2;
>
>          p.quant = x265_quant_avx2;
>          p.nquant = x265_nquant_avx2;
> diff -r dc4fcfc574ad -r f33e89178120 source/common/x86/ssd-a.asm
> --- a/source/common/x86/ssd-a.asm       Wed May 20 12:17:44 2015 -0500
> +++ b/source/common/x86/ssd-a.asm       Thu May 21 11:58:50 2015 +0530
> @@ -347,6 +347,7 @@
>  INIT_YMM avx2
>  SSD_ONE    16,  8
>  SSD_ONE    16, 16
> +SSD_ONE    32, 32
>  %endif ; HIGH_BIT_DEPTH
>
>
>  ;-----------------------------------------------------------------------------
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.videolan.org/pipermail/x265-devel/attachments/20150521/ebee7c4f/attachment.html>