[x265] [PATCH] asm: Add sse_ss for [16x16], [32x32] & [64x64] for 8bpp avx2

[chen]

At 2015-09-30 13:53:18,ramya at multicorewareinc.com wrote:
># HG changeset patch
># User Ramya Sriraman ramya at multicorewareinc.com>
># Date 1443592336 -19800
>#      Wed Sep 30 11:22:16 2015 +0530
># Node ID 29b61906162c657da241aecee9012e3f2da34b6d
># Parent  5f1451e5842252b31442e8b6519138d8033bbb2b
>asm: Add sse_ss for [16x16],[32x32] & [64x64] for 8bpp avx2
>
diff -r 5f1451e58422 -r 29b61906162c source/common/x86/ssd-a.asm
>--- a/source/common/x86/ssd-a.asm	Mon Sep 28 16:43:47 2015 +0530
>+++ b/source/common/x86/ssd-a.asm	Wed Sep 30 11:22:16 2015 +0530
>@@ -1100,8 +1100,195 @@
> SSD_SS_32xN
> SSD_SS_48
> SSD_SS_64xN
>+
>+INIT_YMM avx2
>+cglobal pixel_ssd_ss_16x16, 4,4,5
>+    add     r1d, r1d
>+    add     r3d, r3d
>+    pxor    m4, m4
>+
>+    movu        m0, [r0]
>+    movu        m1, [r0+r1]
>+    movu        m2, [r2]
>+    movu        m3, [r2+r3]
>+    psubw       m0, m2
in avx2, vpsubw can work on unaligned address

>+    psubw       m1, m3
>+    lea         r0, [r0+2*r1]
>+    lea         r2, [r2+2*r3]
>+    pmaddwd     m0, m0
>+    pmaddwd     m1, m1
>+    paddd       m0 , m1
>+    paddd       m4, m0


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.videolan.org/pipermail/x265-devel/attachments/20150930/e79ea8f2/attachment-0001.html>