Merge commit '25841dfe806a13de526ae09c11149ab1f83555a8'

* commit '25841dfe806a13de526ae09c11149ab1f83555a8':
  Use ptrdiff_t instead of int for {avg, put}_pixels line_size parameter.

Conflicts:
	libavcodec/alpha/dsputil_alpha.c
	libavcodec/dsputil_template.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>

17 files changed, 185 insertions, 198 deletions

diff --git a/libavcodec/alpha/dsputil_alpha.c b/libavcodec/alpha/dsputil_alpha.c
index 80d218aa54..0d856315a5 100644
--- a/libavcodec/alpha/dsputil_alpha.c
+++ b/libavcodec/alpha/dsputil_alpha.c
@@ -32,7 +32,7 @@ void (*add_pixels_clamped_axp_p)(const int16_t *block, uint8_t *pixels,
 /* These functions were the base for the optimized assembler routines,
    and remain here for documentation purposes.  */
 static void put_pixels_clamped_mvi(const int16_t *block, uint8_t *pixels,
-                                   int line_size)
+                                   ptrdiff_t line_size)
 {
     int i = 8;
     uint64_t clampmask = zap(-1, 0xaa); /* 0x00ff00ff00ff00ff */
@@ -56,7 +56,7 @@ static void put_pixels_clamped_mvi(const int16_t *block, uint8_t *pixels,
 }
 
 void add_pixels_clamped_mvi(const int16_t *block, uint8_t *pixels,
-                            int line_size)
+                            ptrdiff_t line_size)
 {
     int h = 8;
     /* Keep this function a leaf function by generating the constants
@@ -211,8 +211,8 @@ static inline uint64_t avg4(uint64_t l1, uint64_t l2, uint64_t l3, uint64_t l4)
 
 #define MAKE_OP(OPNAME, SUFF, OPKIND, STORE)                                \
 static void OPNAME ## _pixels ## SUFF ## _axp                               \
-        (uint8_t *av_restrict block, const uint8_t *av_restrict pixels,           \
-         int line_size, int h)                                              \
+        (uint8_t *restrict block, const uint8_t *restrict pixels,           \
+         ptrdiff_t line_size, int h)                                        \
 {                                                                           \
     if ((size_t) pixels & 0x7) {                                            \
         OPKIND(uldq, STORE);                                                \
@@ -222,8 +222,8 @@ static void OPNAME ## _pixels ## SUFF ## _axp                               \
 }                                                                           \
                                                                             \
 static void OPNAME ## _pixels16 ## SUFF ## _axp                             \
-        (uint8_t *av_restrict block, const uint8_t *av_restrict pixels,           \
-         int line_size, int h)                                              \
+        (uint8_t *restrict block, const uint8_t *restrict pixels,           \
+         ptrdiff_t line_size, int h)                                        \
 {                                                                           \
     OPNAME ## _pixels ## SUFF ## _axp(block,     pixels,     line_size, h); \
     OPNAME ## _pixels ## SUFF ## _axp(block + 8, pixels + 8, line_size, h); \
@@ -262,7 +262,7 @@ PIXOP(put_no_rnd, STORE);
 PIXOP(avg_no_rnd, STORE);
 
 static void put_pixels16_axp_asm(uint8_t *block, const uint8_t *pixels,
-                                 int line_size, int h)
+                                 ptrdiff_t line_size, int h)
 {
     put_pixels_axp_asm(block,     pixels,     line_size, h);
     put_pixels_axp_asm(block + 8, pixels + 8, line_size, h);
diff --git a/libavcodec/alpha/dsputil_alpha.h b/libavcodec/alpha/dsputil_alpha.h
index cdb51358de..16f2d5e6b3 100644
--- a/libavcodec/alpha/dsputil_alpha.h
+++ b/libavcodec/alpha/dsputil_alpha.h
@@ -26,7 +26,7 @@ void ff_simple_idct_put_axp(uint8_t *dest, int line_size, int16_t *block);
 void ff_simple_idct_add_axp(uint8_t *dest, int line_size, int16_t *block);
 
 void put_pixels_axp_asm(uint8_t *block, const uint8_t *pixels,
-                        int line_size, int h);
+                        ptrdiff_t line_size, int h);
 void put_pixels_clamped_mvi_asm(const int16_t *block, uint8_t *pixels,
                                 int line_size);
 void add_pixels_clamped_mvi_asm(const int16_t *block, uint8_t *pixels,
diff --git a/libavcodec/arm/dsputil_init_arm.c b/libavcodec/arm/dsputil_init_arm.c
index ea75c3cc12..8a24c57467 100644
--- a/libavcodec/arm/dsputil_init_arm.c
+++ b/libavcodec/arm/dsputil_init_arm.c
@@ -30,16 +30,16 @@ void ff_simple_idct_arm(int16_t *data);
 static void (*ff_put_pixels_clamped)(const int16_t *block, uint8_t *pixels, int line_size);
 static void (*ff_add_pixels_clamped)(const int16_t *block, uint8_t *pixels, int line_size);
 
-void ff_put_pixels8_arm(uint8_t *block, const uint8_t *pixels, int line_size, int h);
-void ff_put_pixels8_x2_arm(uint8_t *block, const uint8_t *pixels, int line_size, int h);
-void ff_put_pixels8_y2_arm(uint8_t *block, const uint8_t *pixels, int line_size, int h);
-void ff_put_pixels8_xy2_arm(uint8_t *block, const uint8_t *pixels, int line_size, int h);
+void ff_put_pixels8_arm(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h);
+void ff_put_pixels8_x2_arm(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h);
+void ff_put_pixels8_y2_arm(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h);
+void ff_put_pixels8_xy2_arm(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h);
 
-void ff_put_no_rnd_pixels8_x2_arm(uint8_t *block, const uint8_t *pixels, int line_size, int h);
-void ff_put_no_rnd_pixels8_y2_arm(uint8_t *block, const uint8_t *pixels, int line_size, int h);
-void ff_put_no_rnd_pixels8_xy2_arm(uint8_t *block, const uint8_t *pixels, int line_size, int h);
+void ff_put_no_rnd_pixels8_x2_arm(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h);
+void ff_put_no_rnd_pixels8_y2_arm(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h);
+void ff_put_no_rnd_pixels8_xy2_arm(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h);
 
-void ff_put_pixels16_arm(uint8_t *block, const uint8_t *pixels, int line_size, int h);
+void ff_put_pixels16_arm(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h);
 
 CALL_2X_PIXELS(ff_put_pixels16_x2_arm,         ff_put_pixels8_x2_arm,        8)
 CALL_2X_PIXELS(ff_put_pixels16_y2_arm,         ff_put_pixels8_y2_arm,        8)
diff --git a/libavcodec/arm/dsputil_init_armv6.c b/libavcodec/arm/dsputil_init_armv6.c
index dc126fef8a..9fd33f935b 100644
--- a/libavcodec/arm/dsputil_init_armv6.c
+++ b/libavcodec/arm/dsputil_init_armv6.c
@@ -28,23 +28,23 @@ void ff_simple_idct_armv6(int16_t *data);
 void ff_simple_idct_put_armv6(uint8_t *dest, int line_size, int16_t *data);
 void ff_simple_idct_add_armv6(uint8_t *dest, int line_size, int16_t *data);
 
-void ff_put_pixels16_armv6(uint8_t *, const uint8_t *, int, int);
-void ff_put_pixels16_x2_armv6(uint8_t *, const uint8_t *, int, int);
-void ff_put_pixels16_y2_armv6(uint8_t *, const uint8_t *, int, int);
+void ff_put_pixels16_armv6(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_put_pixels16_x2_armv6(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_put_pixels16_y2_armv6(uint8_t *, const uint8_t *, ptrdiff_t, int);
 
-void ff_put_pixels16_x2_no_rnd_armv6(uint8_t *, const uint8_t *, int, int);
-void ff_put_pixels16_y2_no_rnd_armv6(uint8_t *, const uint8_t *, int, int);
+void ff_put_pixels16_x2_no_rnd_armv6(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_put_pixels16_y2_no_rnd_armv6(uint8_t *, const uint8_t *, ptrdiff_t, int);
 
-void ff_avg_pixels16_armv6(uint8_t *, const uint8_t *, int, int);
+void ff_avg_pixels16_armv6(uint8_t *, const uint8_t *, ptrdiff_t, int);
 
-void ff_put_pixels8_armv6(uint8_t *, const uint8_t *, int, int);
-void ff_put_pixels8_x2_armv6(uint8_t *, const uint8_t *, int, int);
-void ff_put_pixels8_y2_armv6(uint8_t *, const uint8_t *, int, int);
+void ff_put_pixels8_armv6(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_put_pixels8_x2_armv6(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_put_pixels8_y2_armv6(uint8_t *, const uint8_t *, ptrdiff_t, int);
 
-void ff_put_pixels8_x2_no_rnd_armv6(uint8_t *, const uint8_t *, int, int);
-void ff_put_pixels8_y2_no_rnd_armv6(uint8_t *, const uint8_t *, int, int);
+void ff_put_pixels8_x2_no_rnd_armv6(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_put_pixels8_y2_no_rnd_armv6(uint8_t *, const uint8_t *, ptrdiff_t, int);
 
-void ff_avg_pixels8_armv6(uint8_t *, const uint8_t *, int, int);
+void ff_avg_pixels8_armv6(uint8_t *, const uint8_t *, ptrdiff_t, int);
 
 void ff_add_pixels_clamped_armv6(const int16_t *block,
                                  uint8_t *restrict pixels,
diff --git a/libavcodec/arm/dsputil_init_neon.c b/libavcodec/arm/dsputil_init_neon.c
index 3050e8cce4..3987199aa9 100644
--- a/libavcodec/arm/dsputil_init_neon.c
+++ b/libavcodec/arm/dsputil_init_neon.c
@@ -32,32 +32,32 @@ void ff_simple_idct_add_neon(uint8_t *dest, int line_size, int16_t *data);
 void ff_clear_block_neon(int16_t *block);
 void ff_clear_blocks_neon(int16_t *blocks);
 
-void ff_put_pixels16_neon(uint8_t *, const uint8_t *, int, int);
-void ff_put_pixels16_x2_neon(uint8_t *, const uint8_t *, int, int);
-void ff_put_pixels16_y2_neon(uint8_t *, const uint8_t *, int, int);
-void ff_put_pixels16_xy2_neon(uint8_t *, const uint8_t *, int, int);
-void ff_put_pixels8_neon(uint8_t *, const uint8_t *, int, int);
-void ff_put_pixels8_x2_neon(uint8_t *, const uint8_t *, int, int);
-void ff_put_pixels8_y2_neon(uint8_t *, const uint8_t *, int, int);
-void ff_put_pixels8_xy2_neon(uint8_t *, const uint8_t *, int, int);
-void ff_put_pixels16_x2_no_rnd_neon(uint8_t *, const uint8_t *, int, int);
-void ff_put_pixels16_y2_no_rnd_neon(uint8_t *, const uint8_t *, int, int);
-void ff_put_pixels16_xy2_no_rnd_neon(uint8_t *, const uint8_t *, int, int);
-void ff_put_pixels8_x2_no_rnd_neon(uint8_t *, const uint8_t *, int, int);
-void ff_put_pixels8_y2_no_rnd_neon(uint8_t *, const uint8_t *, int, int);
-void ff_put_pixels8_xy2_no_rnd_neon(uint8_t *, const uint8_t *, int, int);
-
-void ff_avg_pixels16_neon(uint8_t *, const uint8_t *, int, int);
-void ff_avg_pixels16_x2_neon(uint8_t *, const uint8_t *, int, int);
-void ff_avg_pixels16_y2_neon(uint8_t *, const uint8_t *, int, int);
-void ff_avg_pixels16_xy2_neon(uint8_t *, const uint8_t *, int, int);
-void ff_avg_pixels8_neon(uint8_t *, const uint8_t *, int, int);
-void ff_avg_pixels8_x2_neon(uint8_t *, const uint8_t *, int, int);
-void ff_avg_pixels8_y2_neon(uint8_t *, const uint8_t *, int, int);
-void ff_avg_pixels8_xy2_neon(uint8_t *, const uint8_t *, int, int);
-void ff_avg_pixels16_x2_no_rnd_neon(uint8_t *, const uint8_t *, int, int);
-void ff_avg_pixels16_y2_no_rnd_neon(uint8_t *, const uint8_t *, int, int);
-void ff_avg_pixels16_xy2_no_rnd_neon(uint8_t *, const uint8_t *, int, int);
+void ff_put_pixels16_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_put_pixels16_x2_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_put_pixels16_y2_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_put_pixels16_xy2_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_put_pixels8_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_put_pixels8_x2_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_put_pixels8_y2_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_put_pixels8_xy2_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_put_pixels16_x2_no_rnd_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_put_pixels16_y2_no_rnd_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_put_pixels16_xy2_no_rnd_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_put_pixels8_x2_no_rnd_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_put_pixels8_y2_no_rnd_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_put_pixels8_xy2_no_rnd_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+
+void ff_avg_pixels16_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_avg_pixels16_x2_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_avg_pixels16_y2_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_avg_pixels16_xy2_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_avg_pixels8_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_avg_pixels8_x2_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_avg_pixels8_y2_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_avg_pixels8_xy2_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_avg_pixels16_x2_no_rnd_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_avg_pixels16_y2_no_rnd_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
+void ff_avg_pixels16_xy2_no_rnd_neon(uint8_t *, const uint8_t *, ptrdiff_t, int);
 
 void ff_add_pixels_clamped_neon(const int16_t *, uint8_t *, int);
 void ff_put_pixels_clamped_neon(const int16_t *, uint8_t *, int);
diff --git a/libavcodec/bfin/dsputil_bfin.c b/libavcodec/bfin/dsputil_bfin.c
index da6edb5e5e..656fe0e00a 100644
--- a/libavcodec/bfin/dsputil_bfin.c
+++ b/libavcodec/bfin/dsputil_bfin.c
@@ -55,73 +55,73 @@ static void bfin_clear_blocks (int16_t *blocks)
 
 
 
-static void bfin_put_pixels8 (uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void bfin_put_pixels8 (uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     ff_bfin_put_pixels8uc (block, pixels, pixels, line_size, line_size, h);
 }
 
-static void bfin_put_pixels8_x2(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void bfin_put_pixels8_x2(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     ff_bfin_put_pixels8uc (block, pixels, pixels+1, line_size, line_size, h);
 }
 
-static void bfin_put_pixels8_y2 (uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void bfin_put_pixels8_y2 (uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     ff_bfin_put_pixels8uc (block, pixels, pixels+line_size, line_size, line_size, h);
 }
 
-static void bfin_put_pixels8_xy2 (uint8_t *block, const uint8_t *s0, int line_size, int h)
+static void bfin_put_pixels8_xy2 (uint8_t *block, const uint8_t *s0, ptrdiff_t line_size, int h)
 {
     ff_bfin_z_put_pixels8_xy2 (block,s0,line_size, line_size, h);
 }
 
-static void bfin_put_pixels16 (uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void bfin_put_pixels16 (uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     ff_bfin_put_pixels16uc (block, pixels, pixels, line_size, line_size, h);
 }
 
-static void bfin_put_pixels16_x2 (uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void bfin_put_pixels16_x2 (uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     ff_bfin_put_pixels16uc (block, pixels, pixels+1, line_size, line_size, h);
 }
 
-static void bfin_put_pixels16_y2 (uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void bfin_put_pixels16_y2 (uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     ff_bfin_put_pixels16uc (block, pixels, pixels+line_size, line_size, line_size, h);
 }
 
-static void bfin_put_pixels16_xy2 (uint8_t *block, const uint8_t *s0, int line_size, int h)
+static void bfin_put_pixels16_xy2 (uint8_t *block, const uint8_t *s0, ptrdiff_t line_size, int h)
 {
     ff_bfin_z_put_pixels16_xy2 (block,s0,line_size, line_size, h);
 }
 
-static void bfin_put_pixels8_nornd (uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void bfin_put_pixels8_nornd (uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     ff_bfin_put_pixels8uc_nornd (block, pixels, pixels, line_size, h);
 }
 
-static void bfin_put_pixels8_x2_nornd (uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void bfin_put_pixels8_x2_nornd (uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     ff_bfin_put_pixels8uc_nornd (block, pixels, pixels+1, line_size, h);
 }
 
-static void bfin_put_pixels8_y2_nornd (uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void bfin_put_pixels8_y2_nornd (uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     ff_bfin_put_pixels8uc_nornd (block, pixels, pixels+line_size, line_size, h);
 }
 
 
-static void bfin_put_pixels16_nornd (uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void bfin_put_pixels16_nornd (uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     ff_bfin_put_pixels16uc_nornd (block, pixels, pixels, line_size, h);
 }
 
-static void bfin_put_pixels16_x2_nornd (uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void bfin_put_pixels16_x2_nornd (uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     ff_bfin_put_pixels16uc_nornd (block, pixels, pixels+1, line_size, h);
 }
 
-static void bfin_put_pixels16_y2_nornd (uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void bfin_put_pixels16_y2_nornd (uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     ff_bfin_put_pixels16uc_nornd (block, pixels, pixels+line_size, line_size, h);
 }
diff --git a/libavcodec/dsputil.h b/libavcodec/dsputil.h
index b1eba49915..4fea90f666 100644
--- a/libavcodec/dsputil.h
+++ b/libavcodec/dsputil.h
@@ -144,7 +144,7 @@ void clear_blocks_c(int16_t *blocks);
 /* add and put pixel (decoding) */
 // blocksizes for op_pixels_func are 8x4,8x8 16x8 16x16
 //h for op_pixels_func is limited to {width/2, width} but never larger than 16 and never smaller than 4
-typedef void (*op_pixels_func)(uint8_t *block/*align width (8 or 16)*/, const uint8_t *pixels/*align 1*/, int line_size, int h);
+typedef void (*op_pixels_func)(uint8_t *block/*align width (8 or 16)*/, const uint8_t *pixels/*align 1*/, ptrdiff_t line_size, int h);
 typedef void (*tpel_mc_func)(uint8_t *block/*align width (8 or 16)*/, const uint8_t *pixels/*align 1*/, int line_size, int w, int h);
 typedef void (*qpel_mc_func)(uint8_t *dst/*align width (8 or 16)*/, uint8_t *src/*align 1*/, int stride);
 typedef void (*h264_chroma_mc_func)(uint8_t *dst/*align 8*/, uint8_t *src/*align 1*/, int srcStride, int h, int x, int y);
diff --git a/libavcodec/dsputil_template.c b/libavcodec/dsputil_template.c
index 7e20995a1d..ceaebea439 100644
--- a/libavcodec/dsputil_template.c
+++ b/libavcodec/dsputil_template.c
@@ -172,19 +172,19 @@ static inline void FUNC(OPNAME ## _no_rnd_pixels16_l2)(uint8_t *dst, const uint8
     FUNC(OPNAME ## _no_rnd_pixels8_l2)(dst+8*sizeof(pixel), src1+8*sizeof(pixel), src2+8*sizeof(pixel), dst_stride, src_stride1, src_stride2, h);\
 }\
 \
-static inline void FUNCC(OPNAME ## _no_rnd_pixels8_x2)(uint8_t *block, const uint8_t *pixels, int line_size, int h){\
+static inline void FUNCC(OPNAME ## _no_rnd_pixels8_x2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h){\
     FUNC(OPNAME ## _no_rnd_pixels8_l2)(block, pixels, pixels+sizeof(pixel), line_size, line_size, line_size, h);\
 }\
 \
-static inline void FUNCC(OPNAME ## _pixels8_x2)(uint8_t *block, const uint8_t *pixels, int line_size, int h){\
+static inline void FUNCC(OPNAME ## _pixels8_x2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h){\
     FUNC(OPNAME ## _pixels8_l2)(block, pixels, pixels+sizeof(pixel), line_size, line_size, line_size, h);\
 }\
 \
-static inline void FUNCC(OPNAME ## _no_rnd_pixels8_y2)(uint8_t *block, const uint8_t *pixels, int line_size, int h){\
+static inline void FUNCC(OPNAME ## _no_rnd_pixels8_y2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h){\
     FUNC(OPNAME ## _no_rnd_pixels8_l2)(block, pixels, pixels+line_size, line_size, line_size, line_size, h);\
 }\
 \
-static inline void FUNCC(OPNAME ## _pixels8_y2)(uint8_t *block, const uint8_t *pixels, int line_size, int h){\
+static inline void FUNCC(OPNAME ## _pixels8_y2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h){\
     FUNC(OPNAME ## _pixels8_l2)(block, pixels, pixels+line_size, line_size, line_size, line_size, h);\
 }\
 \
@@ -225,19 +225,19 @@ static inline void FUNC(OPNAME ## _pixels8_l4)(uint8_t *dst, const uint8_t *src1
     }\
 }\
 \
-static inline void FUNCC(OPNAME ## _pixels4_x2)(uint8_t *block, const uint8_t *pixels, int line_size, int h){\
+static inline void FUNCC(OPNAME ## _pixels4_x2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h){\
     FUNC(OPNAME ## _pixels4_l2)(block, pixels, pixels+sizeof(pixel), line_size, line_size, line_size, h);\
 }\
 \
-static inline void FUNCC(OPNAME ## _pixels4_y2)(uint8_t *block, const uint8_t *pixels, int line_size, int h){\
+static inline void FUNCC(OPNAME ## _pixels4_y2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h){\
     FUNC(OPNAME ## _pixels4_l2)(block, pixels, pixels+line_size, line_size, line_size, line_size, h);\
 }\
 \
-static inline void FUNCC(OPNAME ## _pixels2_x2)(uint8_t *block, const uint8_t *pixels, int line_size, int h){\
+static inline void FUNCC(OPNAME ## _pixels2_x2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h){\
     FUNC(OPNAME ## _pixels2_l2)(block, pixels, pixels+sizeof(pixel), line_size, line_size, line_size, h);\
 }\
 \
-static inline void FUNCC(OPNAME ## _pixels2_y2)(uint8_t *block, const uint8_t *pixels, int line_size, int h){\
+static inline void FUNCC(OPNAME ## _pixels2_y2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h){\
     FUNC(OPNAME ## _pixels2_l2)(block, pixels, pixels+line_size, line_size, line_size, line_size, h);\
 }\
 \
@@ -288,11 +288,11 @@ static inline void FUNC(OPNAME ## _no_rnd_pixels16_l4)(uint8_t *dst, const uint8
     FUNC(OPNAME ## _no_rnd_pixels8_l4)(dst+8*sizeof(pixel), src1+8*sizeof(pixel), src2+8*sizeof(pixel), src3+8*sizeof(pixel), src4+8*sizeof(pixel), dst_stride, src_stride1, src_stride2, src_stride3, src_stride4, h);\
 }\
 \
-static inline void FUNCC(OPNAME ## _pixels2_xy2)(uint8_t *p_block, const uint8_t *p_pixels, int line_size, int h)\
+static inline void FUNCC(OPNAME ## _pixels2_xy2)(uint8_t *_block, const uint8_t *_pixels, ptrdiff_t line_size, int h)\
 {\
         int i, a0, b0, a1, b1;\
-        pixel *block = (pixel*)p_block;\
-        const pixel *pixels = (const pixel*)p_pixels;\
+        pixel *block = (pixel*)_block;\
+        const pixel *pixels = (const pixel*)_pixels;\
         line_size >>= sizeof(pixel)-1;\
         a0= pixels[0];\
         b0= pixels[1] + 2;\
@@ -324,7 +324,7 @@ static inline void FUNCC(OPNAME ## _pixels2_xy2)(uint8_t *p_block, const uint8_t
         }\
 }\
 \
-static inline void FUNCC(OPNAME ## _pixels4_xy2)(uint8_t *block, const uint8_t *pixels, int line_size, int h)\
+static inline void FUNCC(OPNAME ## _pixels4_xy2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)\
 {\
         /* FIXME HIGH BIT DEPTH */\
         int i;\
@@ -403,7 +403,7 @@ static inline void FUNCC(OPNAME ## _pixels8_xy2)(uint8_t *block, const uint8_t *
     }\
 }\
 \
-static inline void FUNCC(OPNAME ## _no_rnd_pixels8_xy2)(uint8_t *block, const uint8_t *pixels, int line_size, int h)\
+static inline void FUNCC(OPNAME ## _no_rnd_pixels8_xy2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)\
 {\
     /* FIXME HIGH BIT DEPTH */\
     int j;\
diff --git a/libavcodec/motion_est.c b/libavcodec/motion_est.c
index ce61cab15b..5fafb1bb13 100644
--- a/libavcodec/motion_est.c
+++ b/libavcodec/motion_est.c
@@ -290,7 +290,7 @@ static int zero_cmp(void *s, uint8_t *a, uint8_t *b, int stride, int h){
     return 0;
 }
 
-static void zero_hpel(uint8_t *a, const uint8_t *b, int stride, int h){
+static void zero_hpel(uint8_t *a, const uint8_t *b, ptrdiff_t stride, int h){
 }
 
 int ff_init_me(MpegEncContext *s){
diff --git a/libavcodec/ppc/dsputil_altivec.c b/libavcodec/ppc/dsputil_altivec.c
index ab8e3b7fb1..1527bd0245 100644
--- a/libavcodec/ppc/dsputil_altivec.c
+++ b/libavcodec/ppc/dsputil_altivec.c
@@ -607,7 +607,7 @@ static void add_bytes_altivec(uint8_t *dst, uint8_t *src, int w) {
 }
 
 /* next one assumes that ((line_size % 16) == 0) */
-void ff_put_pixels16_altivec(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+void ff_put_pixels16_altivec(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     register vector unsigned char pixelsv1, pixelsv2;
     register vector unsigned char pixelsv1B, pixelsv2B;
@@ -616,9 +616,9 @@ void ff_put_pixels16_altivec(uint8_t *block, const uint8_t *pixels, int line_siz
 
     register vector unsigned char perm = vec_lvsl(0, pixels);
     int i;
-    register int line_size_2 = line_size << 1;
-    register int line_size_3 = line_size + line_size_2;
-    register int line_size_4 = line_size << 2;
+    register ptrdiff_t line_size_2 = line_size << 1;
+    register ptrdiff_t line_size_3 = line_size + line_size_2;
+    register ptrdiff_t line_size_4 = line_size << 2;
 
 // hand-unrolling the loop by 4 gains about 15%
 // mininum execution time goes from 74 to 60 cycles
@@ -649,7 +649,7 @@ void ff_put_pixels16_altivec(uint8_t *block, const uint8_t *pixels, int line_siz
 
 /* next one assumes that ((line_size % 16) == 0) */
 #define op_avg(a,b)  a = ( ((a)|(b)) - ((((a)^(b))&0xFEFEFEFEUL)>>1) )
-void ff_avg_pixels16_altivec(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+void ff_avg_pixels16_altivec(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     register vector unsigned char pixelsv1, pixelsv2, pixelsv, blockv;
     register vector unsigned char perm = vec_lvsl(0, pixels);
@@ -668,7 +668,7 @@ void ff_avg_pixels16_altivec(uint8_t *block, const uint8_t *pixels, int line_siz
 }
 
 /* next one assumes that ((line_size % 8) == 0) */
-static void avg_pixels8_altivec(uint8_t * block, const uint8_t * pixels, int line_size, int h)
+static void avg_pixels8_altivec(uint8_t * block, const uint8_t * pixels, ptrdiff_t line_size, int h)
 {
     register vector unsigned char pixelsv1, pixelsv2, pixelsv, blockv;
     int i;
@@ -699,7 +699,7 @@ static void avg_pixels8_altivec(uint8_t * block, const uint8_t * pixels, int lin
 }
 
 /* next one assumes that ((line_size % 8) == 0) */
-static void put_pixels8_xy2_altivec(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void put_pixels8_xy2_altivec(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     register int i;
     register vector unsigned char pixelsv1, pixelsv2, pixelsavg;
@@ -758,7 +758,7 @@ static void put_pixels8_xy2_altivec(uint8_t *block, const uint8_t *pixels, int l
 }
 
 /* next one assumes that ((line_size % 8) == 0) */
-static void put_no_rnd_pixels8_xy2_altivec(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void put_no_rnd_pixels8_xy2_altivec(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     register int i;
     register vector unsigned char pixelsv1, pixelsv2, pixelsavg;
@@ -818,7 +818,7 @@ static void put_no_rnd_pixels8_xy2_altivec(uint8_t *block, const uint8_t *pixels
 }
 
 /* next one assumes that ((line_size % 16) == 0) */
-static void put_pixels16_xy2_altivec(uint8_t * block, const uint8_t * pixels, int line_size, int h)
+static void put_pixels16_xy2_altivec(uint8_t * block, const uint8_t * pixels, ptrdiff_t line_size, int h)
 {
     register int i;
     register vector unsigned char pixelsv1, pixelsv2, pixelsv3, pixelsv4;
@@ -886,7 +886,7 @@ static void put_pixels16_xy2_altivec(uint8_t * block, const uint8_t * pixels, in
 }
 
 /* next one assumes that ((line_size % 16) == 0) */
-static void put_no_rnd_pixels16_xy2_altivec(uint8_t * block, const uint8_t * pixels, int line_size, int h)
+static void put_no_rnd_pixels16_xy2_altivec(uint8_t * block, const uint8_t * pixels, ptrdiff_t line_size, int h)
 {
     register int i;
     register vector unsigned char pixelsv1, pixelsv2, pixelsv3, pixelsv4;
@@ -1284,7 +1284,7 @@ static int hadamard8_diff16_altivec(/*MpegEncContext*/ void *s, uint8_t *dst, ui
 }
 
 /* next one assumes that ((line_size % 8) == 0) */
-static void avg_pixels8_xy2_altivec(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+static void avg_pixels8_xy2_altivec(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     register int i;
     register vector unsigned char pixelsv1, pixelsv2, pixelsavg;
diff --git a/libavcodec/ppc/dsputil_altivec.h b/libavcodec/ppc/dsputil_altivec.h
index 0b5e404225..f23ceaf337 100644
--- a/libavcodec/ppc/dsputil_altivec.h
+++ b/libavcodec/ppc/dsputil_altivec.h
@@ -26,9 +26,9 @@
 #include <stdint.h>
 #include "libavcodec/dsputil.h"
 
-void ff_put_pixels16_altivec(uint8_t *block, const uint8_t *pixels, int line_size, int h);
+void ff_put_pixels16_altivec(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h);
 
-void ff_avg_pixels16_altivec(uint8_t *block, const uint8_t *pixels, int line_size, int h);
+void ff_avg_pixels16_altivec(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h);
 
 void ff_fdct_altivec(int16_t *block);
 void ff_gmc1_altivec(uint8_t *dst, uint8_t *src, int stride, int h,
diff --git a/libavcodec/sh4/dsputil_align.c b/libavcodec/sh4/dsputil_align.c
index fcea3c346b..14274091c0 100644
--- a/libavcodec/sh4/dsputil_align.c
+++ b/libavcodec/sh4/dsputil_align.c
@@ -262,7 +262,7 @@ if (sz==16) { \
 
 #define         DEFFUNC(op,rnd,xy,sz,OP_N,avgfunc) \
 static void op##_##rnd##_pixels##sz##_##xy (uint8_t * dest, const uint8_t * ref, \
-                                const int stride, int height) \
+                                const int ptrdiff_t, int height) \
 { \
         switch((int)ref&3) { \
         case 0:OP_N##0(sz,rnd##_##avgfunc); return; \
diff --git a/libavcodec/sparc/dsputil_vis.c b/libavcodec/sparc/dsputil_vis.c
index 74c5203215..9da80ca778 100644
--- a/libavcodec/sparc/dsputil_vis.c
+++ b/libavcodec/sparc/dsputil_vis.c
@@ -120,7 +120,7 @@ DECLARE_ALIGNED(8, static const int16_t, constants256_1024)[] =
 #define TMP32           58
 
 static void MC_put_o_16_vis (uint8_t * dest, const uint8_t * ref,
-                             const int stride, int height)
+                             const ptrdiff_t line_size, int height)
 {
         ref = vis_alignaddr(ref);
         do {    /* 5 cycles */
@@ -141,7 +141,7 @@ static void MC_put_o_16_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_put_o_8_vis (uint8_t * dest, const uint8_t * ref,
-                            const int stride, int height)
+                            const ptrdiff_t line_size, int height)
 {
         ref = vis_alignaddr(ref);
         do {    /* 4 cycles */
@@ -160,7 +160,7 @@ static void MC_put_o_8_vis (uint8_t * dest, const uint8_t * ref,
 
 
 static void MC_avg_o_16_vis (uint8_t * dest, const uint8_t * ref,
-                             const int stride, int height)
+                             const ptrdiff_t line_size, int height)
 {
         int stride_8 = stride + 8;
 
@@ -320,7 +320,7 @@ static void MC_avg_o_16_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_avg_o_8_vis (uint8_t * dest, const uint8_t * ref,
-                            const int stride, int height)
+                            const ptrdiff_t line_size, int height)
 {
         ref = vis_alignaddr(ref);
 
@@ -412,7 +412,7 @@ static void MC_avg_o_8_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_put_x_16_vis (uint8_t * dest, const uint8_t * ref,
-                             const int stride, int height)
+                             const ptrdiff_t line_size, int height)
 {
         unsigned long off = (unsigned long) ref & 0x7;
         unsigned long off_plus_1 = off + 1;
@@ -604,7 +604,7 @@ static void MC_put_x_16_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_put_x_8_vis (uint8_t * dest, const uint8_t * ref,
-                            const int stride, int height)
+                            const ptrdiff_t line_size, int height)
 {
         unsigned long off = (unsigned long) ref & 0x7;
         unsigned long off_plus_1 = off + 1;
@@ -727,7 +727,7 @@ static void MC_put_x_8_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_avg_x_16_vis (uint8_t * dest, const uint8_t * ref,
-                             const int stride, int height)
+                             const ptrdiff_t line_size, int height)
 {
         unsigned long off = (unsigned long) ref & 0x7;
         unsigned long off_plus_1 = off + 1;
@@ -817,7 +817,7 @@ static void MC_avg_x_16_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_avg_x_8_vis (uint8_t * dest, const uint8_t * ref,
-                            const int stride, int height)
+                            const ptrdiff_t line_size, int height)
 {
         unsigned long off = (unsigned long) ref & 0x7;
         unsigned long off_plus_1 = off + 1;
@@ -982,7 +982,7 @@ static void MC_avg_x_8_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_put_y_16_vis (uint8_t * dest, const uint8_t * ref,
-                             const int stride, int height)
+                             const ptrdiff_t line_size, int height)
 {
         ref = vis_alignaddr(ref);
         vis_ld64(ref[0], TMP0);
@@ -1136,7 +1136,7 @@ static void MC_put_y_16_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_put_y_8_vis (uint8_t * dest, const uint8_t * ref,
-                            const int stride, int height)
+                            const ptrdiff_t line_size, int height)
 {
         ref = vis_alignaddr(ref);
         vis_ld64(ref[0], TMP0);
@@ -1226,7 +1226,7 @@ static void MC_put_y_8_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_avg_y_16_vis (uint8_t * dest, const uint8_t * ref,
-                             const int stride, int height)
+                             const ptrdiff_t line_size, int height)
 {
         int stride_8 = stride + 8;
         int stride_16 = stride + 16;
@@ -1354,7 +1354,7 @@ static void MC_avg_y_16_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_avg_y_8_vis (uint8_t * dest, const uint8_t * ref,
-                            const int stride, int height)
+                            const ptrdiff_t line_size, int height)
 {
         int stride_8 = stride + 8;
 
@@ -1433,7 +1433,7 @@ static void MC_avg_y_8_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_put_xy_16_vis (uint8_t * dest, const uint8_t * ref,
-                              const int stride, int height)
+                              const ptrdiff_t line_size, int height)
 {
         unsigned long off = (unsigned long) ref & 0x7;
         unsigned long off_plus_1 = off + 1;
@@ -1597,7 +1597,7 @@ static void MC_put_xy_16_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_put_xy_8_vis (uint8_t * dest, const uint8_t * ref,
-                             const int stride, int height)
+                             const ptrdiff_t line_size, int height)
 {
         unsigned long off = (unsigned long) ref & 0x7;
         unsigned long off_plus_1 = off + 1;
@@ -1701,7 +1701,7 @@ static void MC_put_xy_8_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_avg_xy_16_vis (uint8_t * dest, const uint8_t * ref,
-                              const int stride, int height)
+                              const ptrdiff_t line_size, int height)
 {
         unsigned long off = (unsigned long) ref & 0x7;
         unsigned long off_plus_1 = off + 1;
@@ -1897,7 +1897,7 @@ static void MC_avg_xy_16_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_avg_xy_8_vis (uint8_t * dest, const uint8_t * ref,
-                             const int stride, int height)
+                             const ptrdiff_t line_size, int height)
 {
         unsigned long off = (unsigned long) ref & 0x7;
         unsigned long off_plus_1 = off + 1;
@@ -2040,7 +2040,7 @@ static void MC_avg_xy_8_vis (uint8_t * dest, const uint8_t * ref,
  */
 
 static void MC_put_no_round_o_16_vis (uint8_t * dest, const uint8_t * ref,
-                                      const int stride, int height)
+                                      const ptrdiff_t line_size, int height)
 {
         ref = vis_alignaddr(ref);
         do {    /* 5 cycles */
@@ -2061,7 +2061,7 @@ static void MC_put_no_round_o_16_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_put_no_round_o_8_vis (uint8_t * dest, const uint8_t * ref,
-                            const int stride, int height)
+                            const ptrdiff_t line_size, int height)
 {
         ref = vis_alignaddr(ref);
         do {    /* 4 cycles */
@@ -2080,7 +2080,7 @@ static void MC_put_no_round_o_8_vis (uint8_t * dest, const uint8_t * ref,
 
 
 static void MC_avg_no_round_o_16_vis (uint8_t * dest, const uint8_t * ref,
-                             const int stride, int height)
+                             const ptrdiff_t line_size, int height)
 {
         int stride_8 = stride + 8;
 
@@ -2240,7 +2240,7 @@ static void MC_avg_no_round_o_16_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_put_no_round_x_16_vis (uint8_t * dest, const uint8_t * ref,
-                             const int stride, int height)
+                             const ptrdiff_t line_size, int height)
 {
         unsigned long off = (unsigned long) ref & 0x7;
         unsigned long off_plus_1 = off + 1;
@@ -2432,7 +2432,7 @@ static void MC_put_no_round_x_16_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_put_no_round_x_8_vis (uint8_t * dest, const uint8_t * ref,
-                            const int stride, int height)
+                            const ptrdiff_t line_size, int height)
 {
         unsigned long off = (unsigned long) ref & 0x7;
         unsigned long off_plus_1 = off + 1;
@@ -2555,7 +2555,7 @@ static void MC_put_no_round_x_8_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_avg_no_round_x_16_vis (uint8_t * dest, const uint8_t * ref,
-                             const int stride, int height)
+                             const ptrdiff_t line_size, int height)
 {
         unsigned long off = (unsigned long) ref & 0x7;
         unsigned long off_plus_1 = off + 1;
@@ -2645,7 +2645,7 @@ static void MC_avg_no_round_x_16_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_put_no_round_y_16_vis (uint8_t * dest, const uint8_t * ref,
-                             const int stride, int height)
+                             const ptrdiff_t line_size, int height)
 {
         ref = vis_alignaddr(ref);
         vis_ld64(ref[0], TMP0);
@@ -2799,7 +2799,7 @@ static void MC_put_no_round_y_16_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_put_no_round_y_8_vis (uint8_t * dest, const uint8_t * ref,
-                            const int stride, int height)
+                            const ptrdiff_t line_size, int height)
 {
         ref = vis_alignaddr(ref);
         vis_ld64(ref[0], TMP0);
@@ -2889,7 +2889,7 @@ static void MC_put_no_round_y_8_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_avg_no_round_y_16_vis (uint8_t * dest, const uint8_t * ref,
-                             const int stride, int height)
+                             const ptrdiff_t line_size, int height)
 {
         int stride_8 = stride + 8;
         int stride_16 = stride + 16;
@@ -3017,7 +3017,7 @@ static void MC_avg_no_round_y_16_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_put_no_round_xy_16_vis (uint8_t * dest, const uint8_t * ref,
-                                       const int stride, int height)
+                                       const ptrdiff_t line_size, int height)
 {
         unsigned long off = (unsigned long) ref & 0x7;
         unsigned long off_plus_1 = off + 1;
@@ -3181,7 +3181,7 @@ static void MC_put_no_round_xy_16_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_put_no_round_xy_8_vis (uint8_t * dest, const uint8_t * ref,
-                                      const int stride, int height)
+                                      const ptrdiff_t line_size, int height)
 {
         unsigned long off = (unsigned long) ref & 0x7;
         unsigned long off_plus_1 = off + 1;
@@ -3285,7 +3285,7 @@ static void MC_put_no_round_xy_8_vis (uint8_t * dest, const uint8_t * ref,
 }
 
 static void MC_avg_no_round_xy_16_vis (uint8_t * dest, const uint8_t * ref,
-                                       const int stride, int height)
+                                       const ptrdiff_t line_size, int height)
 {
         unsigned long off = (unsigned long) ref & 0x7;
         unsigned long off_plus_1 = off + 1;
diff --git a/libavcodec/x86/dsputil_avg_template.c b/libavcodec/x86/dsputil_avg_template.c
index 414667c609..b9a8f83400 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 2e8300a788..271d2b3818 100644
--- a/libavcodec/x86/dsputil_mmx.c
+++ b/libavcodec/x86/dsputil_mmx.c
@@ -84,9 +84,9 @@ 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,
@@ -95,9 +95,9 @@ 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,
@@ -105,47 +105,47 @@ 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)
 {
@@ -1522,9 +1522,9 @@ 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 c4790af5b7..4568207a2e 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 7f0c285fa3..c83c38870b 100644
--- a/libavcodec/x86/hpeldsp.asm
+++ b/libavcodec/x86/hpeldsp.asm
@@ -32,10 +32,9 @@ 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]
@@ -65,10 +64,9 @@ 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]
@@ -110,11 +108,10 @@ 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]
@@ -152,10 +149,9 @@ 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:
@@ -200,10 +196,9 @@ 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
@@ -235,11 +230,10 @@ 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
@@ -273,10 +267,9 @@ 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
@@ -316,10 +309,9 @@ 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]
@@ -347,10 +339,9 @@ 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]
@@ -384,10 +375,9 @@ 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
@@ -427,11 +417,10 @@ 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]
@@ -472,9 +461,8 @@ 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]
@@ -491,9 +479,8 @@ 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]