Use ptrdiff_t instead of int for {avg, put}_pixels line_size parameter.

This avoids SIMD-optimized functions having to sign-extend their
line size argument manually to be able to do pointer arithmetic.

4 files changed, 60 insertions, 73 deletions

diff --git a/libavcodec/x86/dsputil_avg_template.c b/libavcodec/x86/dsputil_avg_template.c
index 90e4074f7a..2249ce6c97 100644
--- a/libavcodec/x86/dsputil_avg_template.c
+++ b/libavcodec/x86/dsputil_avg_template.c
@@ -27,14 +27,14 @@
 //FIXME the following could be optimized too ...
 static void DEF(ff_put_no_rnd_pixels16_x2)(uint8_t *block,
                                            const uint8_t *pixels,
-                                           int line_size, int h)
+                                           ptrdiff_t line_size, int h)
 {
     DEF(ff_put_no_rnd_pixels8_x2)(block,     pixels,     line_size, h);
     DEF(ff_put_no_rnd_pixels8_x2)(block + 8, pixels + 8, line_size, h);
 }
 
 static void DEF(ff_put_pixels16_y2)(uint8_t *block, const uint8_t *pixels,
-                                    int line_size, int h)
+                                    ptrdiff_t line_size, int h)
 {
     DEF(ff_put_pixels8_y2)(block,     pixels,     line_size, h);
     DEF(ff_put_pixels8_y2)(block + 8, pixels + 8, line_size, h);
@@ -42,35 +42,35 @@ static void DEF(ff_put_pixels16_y2)(uint8_t *block, const uint8_t *pixels,
 
 static void DEF(ff_put_no_rnd_pixels16_y2)(uint8_t *block,
                                            const uint8_t *pixels,
-                                           int line_size, int h)
+                                           ptrdiff_t line_size, int h)
 {
     DEF(ff_put_no_rnd_pixels8_y2)(block,     pixels,     line_size, h);
     DEF(ff_put_no_rnd_pixels8_y2)(block + 8, pixels + 8, line_size, h);
 }
 
 static void DEF(ff_avg_pixels16)(uint8_t *block, const uint8_t *pixels,
-                                 int line_size, int h)
+                                 ptrdiff_t line_size, int h)
 {
     DEF(ff_avg_pixels8)(block,     pixels,     line_size, h);
     DEF(ff_avg_pixels8)(block + 8, pixels + 8, line_size, h);
 }
 
 static void DEF(ff_avg_pixels16_x2)(uint8_t *block, const uint8_t *pixels,
-                                    int line_size, int h)
+                                    ptrdiff_t line_size, int h)
 {
     DEF(ff_avg_pixels8_x2)(block,     pixels,     line_size, h);
     DEF(ff_avg_pixels8_x2)(block + 8, pixels + 8, line_size, h);
 }
 
 static void DEF(ff_avg_pixels16_y2)(uint8_t *block, const uint8_t *pixels,
-                                    int line_size, int h)
+                                    ptrdiff_t line_size, int h)
 {
     DEF(ff_avg_pixels8_y2)(block,     pixels,     line_size, h);
     DEF(ff_avg_pixels8_y2)(block + 8, pixels + 8, line_size, h);
 }
 
 static void DEF(ff_avg_pixels16_xy2)(uint8_t *block, const uint8_t *pixels,
-                                     int line_size, int h)
+                                     ptrdiff_t line_size, int h)
 {
     DEF(ff_avg_pixels8_xy2)(block,     pixels,     line_size, h);
     DEF(ff_avg_pixels8_xy2)(block + 8, pixels + 8, line_size, h);
diff --git a/libavcodec/x86/dsputil_mmx.c b/libavcodec/x86/dsputil_mmx.c
index 3ccef6226d..b882c68a3b 100644
--- a/libavcodec/x86/dsputil_mmx.c
+++ b/libavcodec/x86/dsputil_mmx.c
@@ -83,9 +83,9 @@ DECLARE_ALIGNED(16, const double, ff_pd_2)[2] = { 2.0, 2.0 };
 
 #if HAVE_YASM
 void ff_put_pixels8_x2_mmxext(uint8_t *block, const uint8_t *pixels,
-                              int line_size, int h);
+                              ptrdiff_t line_size, int h);
 void ff_put_pixels8_x2_3dnow(uint8_t *block, const uint8_t *pixels,
-                             int line_size, int h);
+                             ptrdiff_t line_size, int h);
 void ff_put_pixels8_l2_mmxext(uint8_t *dst, uint8_t *src1, uint8_t *src2,
                               int dstStride, int src1Stride, int h);
 void ff_put_no_rnd_pixels8_l2_mmxext(uint8_t *dst, uint8_t *src1,
@@ -94,9 +94,9 @@ void ff_put_no_rnd_pixels8_l2_mmxext(uint8_t *dst, uint8_t *src1,
 void ff_avg_pixels8_l2_mmxext(uint8_t *dst, uint8_t *src1, uint8_t *src2,
                               int dstStride, int src1Stride, int h);
 void ff_put_pixels16_x2_mmxext(uint8_t *block, const uint8_t *pixels,
-                               int line_size, int h);
+                               ptrdiff_t line_size, int h);
 void ff_put_pixels16_x2_3dnow(uint8_t *block, const uint8_t *pixels,
-                              int line_size, int h);
+                              ptrdiff_t line_size, int h);
 void ff_put_pixels16_l2_mmxext(uint8_t *dst, uint8_t *src1, uint8_t *src2,
                                int dstStride, int src1Stride, int h);
 void ff_avg_pixels16_l2_mmxext(uint8_t *dst, uint8_t *src1, uint8_t *src2,
@@ -104,47 +104,47 @@ void ff_avg_pixels16_l2_mmxext(uint8_t *dst, uint8_t *src1, uint8_t *src2,
 void ff_put_no_rnd_pixels16_l2_mmxext(uint8_t *dst, uint8_t *src1, uint8_t *src2,
                                       int dstStride, int src1Stride, int h);
 void ff_put_no_rnd_pixels8_x2_mmxext(uint8_t *block, const uint8_t *pixels,
-                                     int line_size, int h);
+                                     ptrdiff_t line_size, int h);
 void ff_put_no_rnd_pixels8_x2_3dnow(uint8_t *block, const uint8_t *pixels,
-                                    int line_size, int h);
+                                    ptrdiff_t line_size, int h);
 void ff_put_no_rnd_pixels8_x2_exact_mmxext(uint8_t *block,
                                            const uint8_t *pixels,
-                                           int line_size, int h);
+                                           ptrdiff_t line_size, int h);
 void ff_put_no_rnd_pixels8_x2_exact_3dnow(uint8_t *block,
                                           const uint8_t *pixels,
-                                          int line_size, int h);
+                                          ptrdiff_t line_size, int h);
 void ff_put_pixels8_y2_mmxext(uint8_t *block, const uint8_t *pixels,
-                              int line_size, int h);
+                              ptrdiff_t line_size, int h);
 void ff_put_pixels8_y2_3dnow(uint8_t *block, const uint8_t *pixels,
-                             int line_size, int h);
+                             ptrdiff_t line_size, int h);
 void ff_put_no_rnd_pixels8_y2_mmxext(uint8_t *block, const uint8_t *pixels,
-                                     int line_size, int h);
+                                     ptrdiff_t line_size, int h);
 void ff_put_no_rnd_pixels8_y2_3dnow(uint8_t *block, const uint8_t *pixels,
-                                    int line_size, int h);
+                                    ptrdiff_t line_size, int h);
 void ff_put_no_rnd_pixels8_y2_exact_mmxext(uint8_t *block,
                                            const uint8_t *pixels,
-                                           int line_size, int h);
+                                           ptrdiff_t line_size, int h);
 void ff_put_no_rnd_pixels8_y2_exact_3dnow(uint8_t *block,
                                           const uint8_t *pixels,
-                                          int line_size, int h);
+                                          ptrdiff_t line_size, int h);
 void ff_avg_pixels8_mmxext(uint8_t *block, const uint8_t *pixels,
-                           int line_size, int h);
+                           ptrdiff_t line_size, int h);
 void ff_avg_pixels8_3dnow(uint8_t *block, const uint8_t *pixels,
-                          int line_size, int h);
+                          ptrdiff_t line_size, int h);
 void ff_avg_pixels8_x2_mmxext(uint8_t *block, const uint8_t *pixels,
-                              int line_size, int h);
+                              ptrdiff_t line_size, int h);
 void ff_avg_pixels8_x2_3dnow(uint8_t *block, const uint8_t *pixels,
-                             int line_size, int h);
+                             ptrdiff_t line_size, int h);
 void ff_avg_pixels8_y2_mmxext(uint8_t *block, const uint8_t *pixels,
-                              int line_size, int h);
+                              ptrdiff_t line_size, int h);
 void ff_avg_pixels8_y2_3dnow(uint8_t *block, const uint8_t *pixels,
-                             int line_size, int h);
+                             ptrdiff_t line_size, int h);
 void ff_avg_pixels8_xy2_mmxext(uint8_t *block, const uint8_t *pixels,
-                               int line_size, int h);
+                               ptrdiff_t line_size, int h);
 void ff_avg_pixels8_xy2_3dnow(uint8_t *block, const uint8_t *pixels,
-                              int line_size, int h);
+                              ptrdiff_t line_size, int h);
 
-void ff_put_pixels8_mmxext(uint8_t *block, const uint8_t *pixels, int line_size, int h);
+void ff_put_pixels8_mmxext(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h);
 static void ff_put_pixels16_mmxext(uint8_t *block, const uint8_t *pixels,
                                    int line_size, int h)
 {
@@ -1455,9 +1455,9 @@ static void gmc_mmx(uint8_t *dst, uint8_t *src,
 #endif /* HAVE_INLINE_ASM */
 
 void ff_put_pixels16_sse2(uint8_t *block, const uint8_t *pixels,
-                          int line_size, int h);
+                          ptrdiff_t line_size, int h);
 void ff_avg_pixels16_sse2(uint8_t *block, const uint8_t *pixels,
-                          int line_size, int h);
+                          ptrdiff_t line_size, int h);
 
 void ff_put_h264_chroma_mc8_rnd_mmx  (uint8_t *dst, uint8_t *src,
                                       int stride, int h, int x, int y);
diff --git a/libavcodec/x86/dsputil_rnd_template.c b/libavcodec/x86/dsputil_rnd_template.c
index 7fcc6b7533..6ce926c6d1 100644
--- a/libavcodec/x86/dsputil_rnd_template.c
+++ b/libavcodec/x86/dsputil_rnd_template.c
@@ -25,7 +25,7 @@
  */
 
 // put_pixels
-static void DEF(put, pixels8_x2)(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void DEF(put, pixels8_x2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     MOVQ_BFE(mm6);
     __asm__ volatile(
@@ -107,7 +107,7 @@ static void av_unused DEF(put, pixels8_l2)(uint8_t *dst, uint8_t *src1, uint8_t
         :"memory");
 }
 
-static void DEF(put, pixels16_x2)(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void DEF(put, pixels16_x2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     MOVQ_BFE(mm6);
     __asm__ volatile(
@@ -202,7 +202,7 @@ static void av_unused DEF(put, pixels16_l2)(uint8_t *dst, uint8_t *src1, uint8_t
         :"memory");
 }
 
-static void DEF(put, pixels8_y2)(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void DEF(put, pixels8_y2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     MOVQ_BFE(mm6);
     __asm__ volatile(
@@ -231,7 +231,7 @@ static void DEF(put, pixels8_y2)(uint8_t *block, const uint8_t *pixels, int line
         :REG_a, "memory");
 }
 
-static void DEF(put, pixels8_xy2)(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void DEF(put, pixels8_xy2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     MOVQ_ZERO(mm7);
     SET_RND(mm6); // =2 for rnd  and  =1 for no_rnd version
@@ -298,7 +298,7 @@ static void DEF(put, pixels8_xy2)(uint8_t *block, const uint8_t *pixels, int lin
 }
 
 // avg_pixels
-static void av_unused DEF(avg, pixels4)(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void av_unused DEF(avg, pixels4)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     MOVQ_BFE(mm6);
     JUMPALIGN();
@@ -319,7 +319,7 @@ static void av_unused DEF(avg, pixels4)(uint8_t *block, const uint8_t *pixels, i
 
 #ifndef NO_RND
 // in case more speed is needed - unroling would certainly help
-static void DEF(avg, pixels8)(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void DEF(avg, pixels8)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     MOVQ_BFE(mm6);
     JUMPALIGN();
@@ -339,7 +339,7 @@ static void DEF(avg, pixels8)(uint8_t *block, const uint8_t *pixels, int line_si
 }
 #endif // NO_RND
 
-static void DEF(avg, pixels16)(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void DEF(avg, pixels16)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     MOVQ_BFE(mm6);
     JUMPALIGN();
@@ -363,7 +363,7 @@ static void DEF(avg, pixels16)(uint8_t *block, const uint8_t *pixels, int line_s
 }
 
 #ifndef NO_RND
-static void DEF(avg, pixels8_x2)(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void DEF(avg, pixels8_x2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     MOVQ_BFE(mm6);
     JUMPALIGN();
@@ -405,7 +405,7 @@ static av_unused void DEF(avg, pixels8_l2)(uint8_t *dst, uint8_t *src1, uint8_t
     } while (--h);
 }
 
-static void DEF(avg, pixels16_x2)(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void DEF(avg, pixels16_x2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     MOVQ_BFE(mm6);
     JUMPALIGN();
@@ -458,7 +458,7 @@ static av_unused void DEF(avg, pixels16_l2)(uint8_t *dst, uint8_t *src1, uint8_t
     } while (--h);
 }
 
-static void DEF(avg, pixels8_y2)(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void DEF(avg, pixels8_y2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     MOVQ_BFE(mm6);
     __asm__ volatile(
@@ -498,7 +498,7 @@ static void DEF(avg, pixels8_y2)(uint8_t *block, const uint8_t *pixels, int line
 }
 
 // this routine is 'slightly' suboptimal but mostly unused
-static void DEF(avg, pixels8_xy2)(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void DEF(avg, pixels8_xy2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     MOVQ_ZERO(mm7);
     SET_RND(mm6); // =2 for rnd  and  =1 for no_rnd version
@@ -573,22 +573,22 @@ static void DEF(avg, pixels8_xy2)(uint8_t *block, const uint8_t *pixels, int lin
 }
 
 //FIXME optimize
-static void DEF(put, pixels16_y2)(uint8_t *block, const uint8_t *pixels, int line_size, int h){
+static void DEF(put, pixels16_y2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h){
     DEF(put, pixels8_y2)(block  , pixels  , line_size, h);
     DEF(put, pixels8_y2)(block+8, pixels+8, line_size, h);
 }
 
-static void DEF(put, pixels16_xy2)(uint8_t *block, const uint8_t *pixels, int line_size, int h){
+static void DEF(put, pixels16_xy2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h){
     DEF(put, pixels8_xy2)(block  , pixels  , line_size, h);
     DEF(put, pixels8_xy2)(block+8, pixels+8, line_size, h);
 }
 
-static void DEF(avg, pixels16_y2)(uint8_t *block, const uint8_t *pixels, int line_size, int h){
+static void DEF(avg, pixels16_y2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h){
     DEF(avg, pixels8_y2)(block  , pixels  , line_size, h);
     DEF(avg, pixels8_y2)(block+8, pixels+8, line_size, h);
 }
 
-static void DEF(avg, pixels16_xy2)(uint8_t *block, const uint8_t *pixels, int line_size, int h){
+static void DEF(avg, pixels16_xy2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h){
     DEF(avg, pixels8_xy2)(block  , pixels  , line_size, h);
     DEF(avg, pixels8_xy2)(block+8, pixels+8, line_size, h);
 }
diff --git a/libavcodec/x86/hpeldsp.asm b/libavcodec/x86/hpeldsp.asm
index 920ae67630..d38186c857 100644
--- 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
@@ -309,10 +302,9 @@ 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, r2d
     lea          r4, [r2*2]
 .loop:
     mova         m0, [r0]
@@ -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, r2d
     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]
@@ -465,9 +454,8 @@ INIT_MMX 3dnow
 AVG_PIXELS8_XY2
 
 INIT_XMM sse2
-; void put_pixels16_sse2(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+; void put_pixels16_sse2(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 cglobal put_pixels16, 4,5,4
-    movsxdifnidn r2, r2d
     lea          r4, [r2*3]
 .loop:
     movu         m0, [r1]
@@ -484,9 +472,8 @@ cglobal put_pixels16, 4,5,4
     jnz       .loop
     REP_RET
 
-; void avg_pixels16_sse2(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+; void avg_pixels16_sse2(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 cglobal avg_pixels16, 4,5,4
-    movsxdifnidn r2, r2d
     lea          r4, [r2*3]
 .loop:
     movu         m0, [r1]