[x265] [PATCH] Luma_hpp[16x16] avx2 asm code : improved 2307c->1610c

[chen]

At 2014-11-13 13:39:58,aasaipriya at multicorewareinc.com wrote:
># HG changeset patch
># User Aasaipriya
># Date 1415856424 -19800
>#      Thu Nov 13 10:57:04 2014 +0530
># Node ID c6994599bf61ba8c5a0ede52e6062196b340dce4
># Parent  f0a17f2b4c22ff8aa05afb40e3360e5ac03590a6
>Luma_hpp[16x16] avx2 asm code : improved 2307c->1610c
>
>diff -r f0a17f2b4c22 -r c6994599bf61 source/common/x86/ipfilter8.asm
>--- a/source/common/x86/ipfilter8.asm	Wed Nov 12 15:57:27 2014 +0530
>+++ b/source/common/x86/ipfilter8.asm	Thu Nov 13 10:57:04 2014 +0530
>@@ -897,8 +897,7 @@
>     packuswb        m4, m4
>     vpermq          m4, m4, 11011000b
>     pshufd          xm4, xm4, 11011000b
>-    movq            [r2], xm4
>-    movhps          [r2 + 8], xm4
>+    movdqu          [r2],xm4
>     lea             r0, [r0 + r1]
>     lea             r2, [r2 + r3]
>     dec             r4d
movu is better, and please send full patch
 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.videolan.org/pipermail/x265-devel/attachments/20141114/b115c9b3/attachment.html>