x86: rnd_template: Eliminate pointless OP_AVG macro indirection

[ffmpeg] / libavcodec / x86 / hpeldsp.asm

diff --git a/libavcodec/x86/hpeldsp.asm b/libavcodec/x86/hpeldsp.asm
index 8afd955bf046f49160661e9c037542d4a360ca07..ec04d99e777f11f1df458bc641b2d6b86e53c170 100644 (file)
--- a/libavcodec/x86/hpeldsp.asm
+++ b/libavcodec/x86/hpeldsp.asm
@@ -25,10 +25,9 @@ cextern pb_1
 
 SECTION_TEXT
 
-; put_pixels8_x2(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+; put_pixels8_x2(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 %macro PUT_PIXELS8_X2 0
 cglobal put_pixels8_x2, 4,5
-    movsxdifnidn r2, r2d
     lea          r4, [r2*2]
 .loop:
     mova         m0, [r1]
@@ -58,10 +57,9 @@ INIT_MMX 3dnow
 PUT_PIXELS8_X2
 
 
-; put_pixels16_x2(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+; put_pixels16_x2(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 %macro PUT_PIXELS_16 0
 cglobal put_pixels16_x2, 4,5
-    movsxdifnidn r2, r2d
     lea          r4, [r2*2]
 .loop:
     mova         m0, [r1]
@@ -103,11 +101,10 @@ INIT_MMX 3dnow
 PUT_PIXELS_16
 
 
-; put_no_rnd_pixels8_x2(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+; put_no_rnd_pixels8_x2(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 %macro PUT_NO_RND_PIXELS8_X2 0
 cglobal put_no_rnd_pixels8_x2, 4,5
     mova         m6, [pb_1]
-    movsxdifnidn r2, r2d
     lea          r4, [r2*2]
 .loop:
     mova         m0, [r1]
@@ -145,10 +142,9 @@ INIT_MMX 3dnow
 PUT_NO_RND_PIXELS8_X2
 
 
-; put_no_rnd_pixels8_x2_exact(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+; put_no_rnd_pixels8_x2_exact(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 %macro PUT_NO_RND_PIXELS8_X2_EXACT 0
 cglobal put_no_rnd_pixels8_x2_exact, 4,5
-    movsxdifnidn r2, r2d
     lea          r4, [r2*3]
     pcmpeqb      m6, m6
 .loop:
@@ -193,10 +189,9 @@ INIT_MMX 3dnow
 PUT_NO_RND_PIXELS8_X2_EXACT
 
 
-; put_pixels8_y2(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+; put_pixels8_y2(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 %macro PUT_PIXELS8_Y2 0
 cglobal put_pixels8_y2, 4,5
-    movsxdifnidn r2, r2d
     lea          r4, [r2*2]
     mova         m0, [r1]
     sub          r0, r2
@@ -228,11 +223,10 @@ INIT_MMX 3dnow
 PUT_PIXELS8_Y2
 
 
-; put_no_rnd_pixels8_y2(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+; put_no_rnd_pixels8_y2(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 %macro PUT_NO_RND_PIXELS8_Y2 0
 cglobal put_no_rnd_pixels8_y2, 4,5
     mova         m6, [pb_1]
-    movsxdifnidn r2, r2d
     lea          r4, [r2+r2]
     mova         m0, [r1]
     sub          r0, r2
@@ -266,10 +260,9 @@ INIT_MMX 3dnow
 PUT_NO_RND_PIXELS8_Y2
 
 
-; put_no_rnd_pixels8_y2_exact(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+; put_no_rnd_pixels8_y2_exact(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 %macro PUT_NO_RND_PIXELS8_Y2_EXACT 0
 cglobal put_no_rnd_pixels8_y2_exact, 4,5
-    movsxdifnidn r2, r2d
     lea          r4, [r2*3]
     mova         m0, [r1]
     pcmpeqb      m6, m6
@@ -309,11 +302,10 @@ INIT_MMX 3dnow
 PUT_NO_RND_PIXELS8_Y2_EXACT
 
 
-; avg_pixels8(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+; avg_pixels8(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 %macro AVG_PIXELS8 0
 cglobal avg_pixels8, 4,5
-    movsxdifnidn r2, edx
-    lea          r4, [r2+r2]
+    lea          r4, [r2*2]
 .loop:
     mova         m0, [r0]
     mova         m1, [r0+r2]
@@ -340,10 +332,9 @@ INIT_MMX 3dnow
 AVG_PIXELS8
 
 
-; avg_pixels8_x2(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+; avg_pixels8_x2(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 %macro AVG_PIXELS8_X2 0
 cglobal avg_pixels8_x2, 4,5
-    movsxdifnidn r2, edx
     lea          r4, [r2*2]
 .loop:
     mova         m0, [r1]
@@ -377,10 +368,9 @@ INIT_MMX 3dnow
 AVG_PIXELS8_X2
 
 
-; avg_pixels8_y2(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+; avg_pixels8_y2(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 %macro AVG_PIXELS8_Y2 0
 cglobal avg_pixels8_y2, 4,5
-    movsxdifnidn r2, r2d
     lea          r4, [r2*2]
     mova         m0, [r1]
     sub          r0, r2
@@ -420,11 +410,10 @@ INIT_MMX 3dnow
 AVG_PIXELS8_Y2
 
 
-; avg_pixels8_xy2(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+; avg_pixels8_xy2(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 %macro AVG_PIXELS8_XY2 0
 cglobal avg_pixels8_xy2, 4,5
     mova         m6, [pb_1]
-    movsxdifnidn r2, r2d
     lea          r4, [r2*2]
     mova         m0, [r1]
     pavgb        m0, [r1+1]
@@ -452,7 +441,7 @@ cglobal avg_pixels8_xy2, 4,5
     pavgb        m2, [r0]
     pavgb        m1, [r0+r2]
     mova       [r0], m2
-    mova    [r0+r2], m2
+    mova    [r0+r2], m1
     add          r0, r4
     sub         r3d, 4
     jne .loop