[x265] [PATCH] asm code blockcopy_ps_16x4, hiden register copy

[praveen at multicorewareinc.com]

# HG changeset patch
# User Praveen Tiwari
# Date 1384168071 -19800
# Node ID a4198e18262ff5ce0655fa322e1ccdb69a44f933
# Parent  859aa956515ccc0b1717edaed67c7e8f59f915fc
asm code blockcopy_ps_16x4, hiden register copy

diff -r 859aa956515c -r a4198e18262f source/common/x86/blockcopy8.asm
--- a/source/common/x86/blockcopy8.asm	Mon Nov 11 16:29:44 2013 +0530
+++ b/source/common/x86/blockcopy8.asm	Mon Nov 11 16:37:51 2013 +0530
@@ -1797,19 +1797,19 @@
 pxor       m0,      m0
 
 movu       m1,                 [r2]
-punpcklbw  m2,                 m1,            m0
+pmovzxbw   m2,                 m1
 movu       [r0],               m2
 punpckhbw  m1,                 m0
 movu       [r0 + 16],          m1
 
 movu       m1,                 [r2 + r3]
-punpcklbw  m2,                 m1,            m0
+pmovzxbw   m2,                 m1
 movu       [r0 + r1],          m2
 punpckhbw  m1,                 m0
 movu       [r0 + r1 + 16],     m1
 
 movu       m1,                 [r2 + 2 * r3]
-punpcklbw  m2,                 m1,            m0
+pmovzxbw   m2,                 m1
 movu       [r0 + 2 * r1],      m2
 punpckhbw  m1,                 m0
 movu       [r0 + 2 * r1 + 16], m1
@@ -1818,7 +1818,7 @@
 lea        r2,                 [r2 + 2 * r3]
 
 movu       m1,                 [r2 + r3]
-punpcklbw  m2,                 m1,            m0
+pmovzxbw   m2,                 m1
 movu       [r0 + r1],          m2
 punpckhbw  m1,                 m0
 movu       [r0 + r1 + 16],     m1