[x265] fix copy16to16_shl

[Satoshi Nakagawa]

# HG changeset patch
# User Satoshi Nakagawa <nakagawa424 at oki.com>
# Date 1416530954 -32400
#      Fri Nov 21 09:49:14 2014 +0900
# Node ID d330d286791b70b00436cee62a1956634635a638
# Parent  2abf89f5c4f2b797705f3b6e8d6670962daa38b9
fix copy16to16_shl

diff -r 2abf89f5c4f2 -r d330d286791b source/common/pixel.cpp
--- a/source/common/pixel.cpp	Thu Nov 20 14:31:04 2014 -0800
+++ b/source/common/pixel.cpp	Fri Nov 21 09:49:14 2014 +0900
@@ -493,11 +493,12 @@
 
 void copy16to16_shl(int16_t *dst, const int16_t *src, intptr_t stride, int shift, int size)
 {
+    X265_CHECK(!(size & 3), "invalid size\n");
     for (int i = 0; i < size; i++)
     {
         for (int j = 0; j < size; j++)
         {
-            dst[i * size + j] = (src[i * stride + j]) << shift;
+            dst[i * size + j] = src[i * stride + j] << shift;
         }
     }
 }
diff -r 2abf89f5c4f2 -r d330d286791b source/common/x86/asm-primitives.cpp
--- a/source/common/x86/asm-primitives.cpp	Thu Nov 20 14:31:04 2014 -0800
+++ b/source/common/x86/asm-primitives.cpp	Fri Nov 21 09:49:14 2014 +0900
@@ -1548,6 +1548,7 @@
         p.frame_init_lowres_core = x265_frame_init_lowres_core_sse2;
         SA8D_INTER_FROM_BLOCK(sse2);
 
+        p.cpy16to16_shl = x265_copy16to16_shl_sse2;
         p.cvt32to16_shl[BLOCK_4x4] = x265_cvt32to16_shl_4_sse2;
         p.cvt32to16_shl[BLOCK_8x8] = x265_cvt32to16_shl_8_sse2;
         p.cvt32to16_shl[BLOCK_16x16] = x265_cvt32to16_shl_16_sse2;
@@ -1614,7 +1615,6 @@
         LUMA_ADDAVG(_sse4);
         CHROMA_ADDAVG(_sse4);
         CHROMA_ADDAVG_422(_sse4);
-        p.cpy16to16_shl = x265_copy16to16_shl_sse4;
         p.cvt16to32_shr[BLOCK_4x4] = x265_cvt16to32_shr_4_sse4;
         p.cvt16to32_shr[BLOCK_8x8] = x265_cvt16to32_shr_8_sse4;
         p.cvt16to32_shr[BLOCK_16x16] = x265_cvt16to32_shr_16_sse4;
diff -r 2abf89f5c4f2 -r d330d286791b source/common/x86/blockcopy8.asm
--- a/source/common/x86/blockcopy8.asm	Thu Nov 20 14:31:04 2014 -0800
+++ b/source/common/x86/blockcopy8.asm	Fri Nov 21 09:49:14 2014 +0900
@@ -3672,37 +3672,35 @@
 ;--------------------------------------------------------------------------------------
 ; void copy16to16_shl(int16_t *dst, int16_t *src, intptr_t stride, int shift, int size);
 ;--------------------------------------------------------------------------------------
-INIT_XMM sse4
-cglobal copy16to16_shl, 5, 7, 2, dst, src, stride, shift, size
+INIT_XMM sse2
+cglobal copy16to16_shl, 5, 6, 2, dst, src, stride, shift, size
 %define shift       m1
 
     ; make shift
-    mov             r5d,      r3m
-    movd            shift,    r5d
+    movd            shift,    r3d
 
     ; register alloc
     ; r0 - dst
     ; r1 - src
     ; r2 - stride
-    ; r3 - shift
     ; r4 - size
 
     sub             r2d,      r4d
     add             r2d,      r2d
     mov             r5d,      r4d
-    shr             r4d,      3
+    shr             r4d,      2
 .loop_row:
-    mov             r6d,      r4d
+    mov             r3d,      r4d
 
 .loop_col:
-    movu            m0,       [r1]
+    movh            m0,       [r1]
     psllw           m0,       shift
-    movu            [r0],     m0
-
-    add             r1,       16
-    add             r0,       16
-
-    dec             r6d
+    movh            [r0],     m0
+
+    add             r1,       8
+    add             r0,       8
+
+    dec             r3d
     jnz             .loop_col
 
     add             r1,       r2
diff -r 2abf89f5c4f2 -r d330d286791b source/common/x86/blockcopy8.h
--- a/source/common/x86/blockcopy8.h	Thu Nov 20 14:31:04 2014 -0800
+++ b/source/common/x86/blockcopy8.h	Fri Nov 21 09:49:14 2014 +0900
@@ -32,7 +32,7 @@
 void x265_cvt32to16_shl_8_avx2(int16_t* dst, const int* src, intptr_t, int);
 void x265_cvt32to16_shl_16_avx2(int16_t* dst, const int* src, intptr_t, int);
 void x265_cvt32to16_shl_32_avx2(int16_t* dst, const int* src, intptr_t, int);
-void x265_copy16to16_shl_sse4(int16_t* dst, const int16_t* src, intptr_t, int32_t, int32_t);
+void x265_copy16to16_shl_sse2(int16_t* dst, const int16_t* src, intptr_t, int32_t, int32_t);
 void x265_cvt16to32_shr_4_sse4(int32_t* dst, const int16_t* src, intptr_t, int32_t, int32_t);
 void x265_cvt16to32_shr_8_sse4(int32_t* dst, const int16_t* src, intptr_t, int32_t, int32_t);
 void x265_cvt16to32_shr_16_sse4(int32_t* dst, const int16_t* src, intptr_t, int32_t, int32_t);