Use ptrdiff_t instead of int for intra pred "stride" function parameter.

This way, SIMD-optimized functions don't have to sign-extend their
stride argument manually to be able to do pointer arithmetic.

5 files changed, 264 insertions, 117 deletions

diff --git a/libavcodec/arm/h264pred_init_arm.c b/libavcodec/arm/h264pred_init_arm.c
index 371b1a6733..39c012127d 100644
--- a/libavcodec/arm/h264pred_init_arm.c
+++ b/libavcodec/arm/h264pred_init_arm.c
@@ -23,25 +23,25 @@
 #include "libavutil/arm/cpu.h"
 #include "libavcodec/h264pred.h"
 
-void ff_pred16x16_vert_neon(uint8_t *src, int stride);
-void ff_pred16x16_hor_neon(uint8_t *src, int stride);
-void ff_pred16x16_plane_neon(uint8_t *src, int stride);
-void ff_pred16x16_dc_neon(uint8_t *src, int stride);
-void ff_pred16x16_128_dc_neon(uint8_t *src, int stride);
-void ff_pred16x16_left_dc_neon(uint8_t *src, int stride);
-void ff_pred16x16_top_dc_neon(uint8_t *src, int stride);
+void ff_pred16x16_vert_neon(uint8_t *src, ptrdiff_t stride);
+void ff_pred16x16_hor_neon(uint8_t *src, ptrdiff_t stride);
+void ff_pred16x16_plane_neon(uint8_t *src, ptrdiff_t stride);
+void ff_pred16x16_dc_neon(uint8_t *src, ptrdiff_t stride);
+void ff_pred16x16_128_dc_neon(uint8_t *src, ptrdiff_t stride);
+void ff_pred16x16_left_dc_neon(uint8_t *src, ptrdiff_t stride);
+void ff_pred16x16_top_dc_neon(uint8_t *src, ptrdiff_t stride);
 
-void ff_pred8x8_vert_neon(uint8_t *src, int stride);
-void ff_pred8x8_hor_neon(uint8_t *src, int stride);
-void ff_pred8x8_plane_neon(uint8_t *src, int stride);
-void ff_pred8x8_dc_neon(uint8_t *src, int stride);
-void ff_pred8x8_128_dc_neon(uint8_t *src, int stride);
-void ff_pred8x8_left_dc_neon(uint8_t *src, int stride);
-void ff_pred8x8_top_dc_neon(uint8_t *src, int stride);
-void ff_pred8x8_l0t_dc_neon(uint8_t *src, int stride);
-void ff_pred8x8_0lt_dc_neon(uint8_t *src, int stride);
-void ff_pred8x8_l00_dc_neon(uint8_t *src, int stride);
-void ff_pred8x8_0l0_dc_neon(uint8_t *src, int stride);
+void ff_pred8x8_vert_neon(uint8_t *src, ptrdiff_t stride);
+void ff_pred8x8_hor_neon(uint8_t *src, ptrdiff_t stride);
+void ff_pred8x8_plane_neon(uint8_t *src, ptrdiff_t stride);
+void ff_pred8x8_dc_neon(uint8_t *src, ptrdiff_t stride);
+void ff_pred8x8_128_dc_neon(uint8_t *src, ptrdiff_t stride);
+void ff_pred8x8_left_dc_neon(uint8_t *src, ptrdiff_t stride);
+void ff_pred8x8_top_dc_neon(uint8_t *src, ptrdiff_t stride);
+void ff_pred8x8_l0t_dc_neon(uint8_t *src, ptrdiff_t stride);
+void ff_pred8x8_0lt_dc_neon(uint8_t *src, ptrdiff_t stride);
+void ff_pred8x8_l00_dc_neon(uint8_t *src, ptrdiff_t stride);
+void ff_pred8x8_0l0_dc_neon(uint8_t *src, ptrdiff_t stride);
 
 static void ff_h264_pred_init_neon(H264PredContext *h, int codec_id, const int bit_depth, const int chroma_format_idc)
 {
diff --git a/libavcodec/h264pred.c b/libavcodec/h264pred.c
index fb44046b2e..94cf9d06d7 100644
--- a/libavcodec/h264pred.c
+++ b/libavcodec/h264pred.c
@@ -39,7 +39,9 @@
 #include "h264pred_template.c"
 #undef BIT_DEPTH
 
-static void pred4x4_vertical_vp8_c(uint8_t *src, const uint8_t *topright, int stride){
+static void pred4x4_vertical_vp8_c(uint8_t *src, const uint8_t *topright,
+                                   ptrdiff_t stride)
+{
     const unsigned lt = src[-1-1*stride];
     LOAD_TOP_EDGE
     LOAD_TOP_RIGHT_EDGE
@@ -54,7 +56,9 @@ static void pred4x4_vertical_vp8_c(uint8_t *src, const uint8_t *topright, int st
     AV_WN32A(src+3*stride, v);
 }
 
-static void pred4x4_horizontal_vp8_c(uint8_t *src, const uint8_t *topright, int stride){
+static void pred4x4_horizontal_vp8_c(uint8_t *src, const uint8_t *topright,
+                                     ptrdiff_t stride)
+{
     const unsigned lt = src[-1-1*stride];
     LOAD_LEFT_EDGE
 
@@ -64,7 +68,9 @@ static void pred4x4_horizontal_vp8_c(uint8_t *src, const uint8_t *topright, int
     AV_WN32A(src+3*stride, ((l2 + 2*l3 + l3 + 2) >> 2)*0x01010101);
 }
 
-static void pred4x4_down_left_svq3_c(uint8_t *src, const uint8_t *topright, int stride){
+static void pred4x4_down_left_svq3_c(uint8_t *src, const uint8_t *topright,
+                                     ptrdiff_t stride)
+{
     LOAD_TOP_EDGE
     LOAD_LEFT_EDGE
 
@@ -86,7 +92,9 @@ static void pred4x4_down_left_svq3_c(uint8_t *src, const uint8_t *topright, int
     src[3+3*stride]=(l3 + t3)>>1;
 }
 
-static void pred4x4_down_left_rv40_c(uint8_t *src, const uint8_t *topright, int stride){
+static void pred4x4_down_left_rv40_c(uint8_t *src, const uint8_t *topright,
+                                     ptrdiff_t stride)
+{
     LOAD_TOP_EDGE
     LOAD_TOP_RIGHT_EDGE
     LOAD_LEFT_EDGE
@@ -110,7 +118,10 @@ static void pred4x4_down_left_rv40_c(uint8_t *src, const uint8_t *topright, int
     src[3+3*stride]=(t6 + t7 + 1 + l6 + l7 + 1)>>2;
 }
 
-static void pred4x4_down_left_rv40_nodown_c(uint8_t *src, const uint8_t *topright, int stride){
+static void pred4x4_down_left_rv40_nodown_c(uint8_t *src,
+                                            const uint8_t *topright,
+                                            ptrdiff_t stride)
+{
     LOAD_TOP_EDGE
     LOAD_TOP_RIGHT_EDGE
     LOAD_LEFT_EDGE
@@ -133,8 +144,11 @@ static void pred4x4_down_left_rv40_nodown_c(uint8_t *src, const uint8_t *toprigh
     src[3+3*stride]=(t6 + t7 + 1 + 2*l3 + 1)>>2;
 }
 
-static void pred4x4_vertical_left_rv40(uint8_t *src, const uint8_t *topright, int stride,
-                                       const int l0, const int l1, const int l2, const int l3, const int l4){
+static void pred4x4_vertical_left_rv40(uint8_t *src, const uint8_t *topright,
+                                       ptrdiff_t stride,
+                                       const int l0, const int l1, const int l2,
+                                       const int l3, const int l4)
+{
     LOAD_TOP_EDGE
     LOAD_TOP_RIGHT_EDGE
 
@@ -156,20 +170,27 @@ static void pred4x4_vertical_left_rv40(uint8_t *src, const uint8_t *topright, in
     src[3+3*stride]=(t4 + 2*t5 + t6 + 2)>>2;
 }
 
-static void pred4x4_vertical_left_rv40_c(uint8_t *src, const uint8_t *topright, int stride){
+static void pred4x4_vertical_left_rv40_c(uint8_t *src, const uint8_t *topright,
+                                         ptrdiff_t stride)
+{
     LOAD_LEFT_EDGE
     LOAD_DOWN_LEFT_EDGE
 
     pred4x4_vertical_left_rv40(src, topright, stride, l0, l1, l2, l3, l4);
 }
 
-static void pred4x4_vertical_left_rv40_nodown_c(uint8_t *src, const uint8_t *topright, int stride){
+static void pred4x4_vertical_left_rv40_nodown_c(uint8_t *src,
+                                                const uint8_t *topright,
+                                                ptrdiff_t stride)
+{
     LOAD_LEFT_EDGE
 
     pred4x4_vertical_left_rv40(src, topright, stride, l0, l1, l2, l3, l3);
 }
 
-static void pred4x4_vertical_left_vp8_c(uint8_t *src, const uint8_t *topright, int stride){
+static void pred4x4_vertical_left_vp8_c(uint8_t *src, const uint8_t *topright,
+                                        ptrdiff_t stride)
+{
     LOAD_TOP_EDGE
     LOAD_TOP_RIGHT_EDGE
 
@@ -191,7 +212,9 @@ static void pred4x4_vertical_left_vp8_c(uint8_t *src, const uint8_t *topright, i
     src[3+3*stride]=(t5 + 2*t6 + t7 + 2)>>2;
 }
 
-static void pred4x4_horizontal_up_rv40_c(uint8_t *src, const uint8_t *topright, int stride){
+static void pred4x4_horizontal_up_rv40_c(uint8_t *src, const uint8_t *topright,
+                                         ptrdiff_t stride)
+{
     LOAD_LEFT_EDGE
     LOAD_DOWN_LEFT_EDGE
     LOAD_TOP_EDGE
@@ -215,7 +238,10 @@ static void pred4x4_horizontal_up_rv40_c(uint8_t *src, const uint8_t *topright,
     src[3+3*stride]=(l4 + 2*l5 + l6 + 2)>>2;
 }
 
-static void pred4x4_horizontal_up_rv40_nodown_c(uint8_t *src, const uint8_t *topright, int stride){
+static void pred4x4_horizontal_up_rv40_nodown_c(uint8_t *src,
+                                                const uint8_t *topright,
+                                                ptrdiff_t stride)
+{
     LOAD_LEFT_EDGE
     LOAD_TOP_EDGE
     LOAD_TOP_RIGHT_EDGE
@@ -238,7 +264,9 @@ static void pred4x4_horizontal_up_rv40_nodown_c(uint8_t *src, const uint8_t *top
     src[3+3*stride]=l3;
 }
 
-static void pred4x4_tm_vp8_c(uint8_t *src, const uint8_t *topright, int stride){
+static void pred4x4_tm_vp8_c(uint8_t *src, const uint8_t *topright,
+                             ptrdiff_t stride)
+{
     uint8_t *cm = ff_cropTbl + MAX_NEG_CROP - src[-1-stride];
     uint8_t *top = src-stride;
     int y;
@@ -253,15 +281,18 @@ static void pred4x4_tm_vp8_c(uint8_t *src, const uint8_t *topright, int stride){
     }
 }
 
-static void pred16x16_plane_svq3_c(uint8_t *src, int stride){
+static void pred16x16_plane_svq3_c(uint8_t *src, ptrdiff_t stride)
+{
     pred16x16_plane_compat_8_c(src, stride, 1, 0);
 }
 
-static void pred16x16_plane_rv40_c(uint8_t *src, int stride){
+static void pred16x16_plane_rv40_c(uint8_t *src, ptrdiff_t stride)
+{
     pred16x16_plane_compat_8_c(src, stride, 0, 1);
 }
 
-static void pred16x16_tm_vp8_c(uint8_t *src, int stride){
+static void pred16x16_tm_vp8_c(uint8_t *src, ptrdiff_t stride)
+{
     uint8_t *cm = ff_cropTbl + MAX_NEG_CROP - src[-1-stride];
     uint8_t *top = src-stride;
     int y;
@@ -288,7 +319,8 @@ static void pred16x16_tm_vp8_c(uint8_t *src, int stride){
     }
 }
 
-static void pred8x8_left_dc_rv40_c(uint8_t *src, int stride){
+static void pred8x8_left_dc_rv40_c(uint8_t *src, ptrdiff_t stride)
+{
     int i;
     unsigned dc0;
 
@@ -303,7 +335,8 @@ static void pred8x8_left_dc_rv40_c(uint8_t *src, int stride){
     }
 }
 
-static void pred8x8_top_dc_rv40_c(uint8_t *src, int stride){
+static void pred8x8_top_dc_rv40_c(uint8_t *src, ptrdiff_t stride)
+{
     int i;
     unsigned dc0;
 
@@ -318,7 +351,8 @@ static void pred8x8_top_dc_rv40_c(uint8_t *src, int stride){
     }
 }
 
-static void pred8x8_dc_rv40_c(uint8_t *src, int stride){
+static void pred8x8_dc_rv40_c(uint8_t *src, ptrdiff_t stride)
+{
     int i;
     unsigned dc0 = 0;
 
@@ -339,7 +373,8 @@ static void pred8x8_dc_rv40_c(uint8_t *src, int stride){
     }
 }
 
-static void pred8x8_tm_vp8_c(uint8_t *src, int stride){
+static void pred8x8_tm_vp8_c(uint8_t *src, ptrdiff_t stride)
+{
     uint8_t *cm = ff_cropTbl + MAX_NEG_CROP - src[-1-stride];
     uint8_t *top = src-stride;
     int y;
@@ -361,7 +396,9 @@ static void pred8x8_tm_vp8_c(uint8_t *src, int stride){
 /**
  * Set the intra prediction function pointers.
  */
-void ff_h264_pred_init(H264PredContext *h, int codec_id, const int bit_depth, const int chroma_format_idc){
+void ff_h264_pred_init(H264PredContext *h, int codec_id, const int bit_depth,
+                       const int chroma_format_idc)
+{
 //    MpegEncContext * const s = &h->s;
 
 #undef FUNC
diff --git a/libavcodec/h264pred.h b/libavcodec/h264pred.h
index a964ae394b..a8b3dba28b 100644
--- a/libavcodec/h264pred.h
+++ b/libavcodec/h264pred.h
@@ -90,21 +90,23 @@
  * Context for storing H.264 prediction functions
  */
 typedef struct H264PredContext {
-    void(*pred4x4[9 + 3 + 3])(uint8_t *src, const uint8_t *topright, int stride); //FIXME move to dsp?
-    void(*pred8x8l[9 + 3])(uint8_t *src, int topleft, int topright, int stride);
-    void(*pred8x8[4 + 3 + 4])(uint8_t *src, int stride);
-    void(*pred16x16[4 + 3 + 2])(uint8_t *src, int stride);
+    void(*pred4x4[9 + 3 + 3])(uint8_t *src, const uint8_t *topright,
+                              ptrdiff_t stride);
+    void(*pred8x8l[9 + 3])(uint8_t *src, int topleft, int topright,
+                           ptrdiff_t stride);
+    void(*pred8x8[4 + 3 + 4])(uint8_t *src, ptrdiff_t stride);
+    void(*pred16x16[4 + 3 + 2])(uint8_t *src, ptrdiff_t stride);
 
     void(*pred4x4_add[2])(uint8_t *pix /*align  4*/,
-                          const DCTELEM *block /*align 16*/, int stride);
+                          const DCTELEM *block /*align 16*/, ptrdiff_t stride);
     void(*pred8x8l_add[2])(uint8_t *pix /*align  8*/,
-                           const DCTELEM *block /*align 16*/, int stride);
+                           const DCTELEM *block /*align 16*/, ptrdiff_t stride);
     void(*pred8x8_add[3])(uint8_t *pix /*align  8*/,
                           const int *block_offset,
-                          const DCTELEM *block /*align 16*/, int stride);
+                          const DCTELEM *block /*align 16*/, ptrdiff_t stride);
     void(*pred16x16_add[3])(uint8_t *pix /*align 16*/,
                             const int *block_offset,
-                            const DCTELEM *block /*align 16*/, int stride);
+                            const DCTELEM *block /*align 16*/, ptrdiff_t stride);
 } H264PredContext;
 
 void ff_h264_pred_init(H264PredContext *h, int codec_id,
diff --git a/libavcodec/h264pred_template.c b/libavcodec/h264pred_template.c
index 19100b69a9..8ae13b0df7 100644
--- a/libavcodec/h264pred_template.c
+++ b/libavcodec/h264pred_template.c
@@ -29,7 +29,9 @@
 
 #include "bit_depth_template.c"
 
-static void FUNCC(pred4x4_vertical)(uint8_t *_src, const uint8_t *topright, int _stride){
+static void FUNCC(pred4x4_vertical)(uint8_t *_src, const uint8_t *topright,
+                                    ptrdiff_t _stride)
+{
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
     const pixel4 a= AV_RN4PA(src-stride);
@@ -40,7 +42,9 @@ static void FUNCC(pred4x4_vertical)(uint8_t *_src, const uint8_t *topright, int
     AV_WN4PA(src+3*stride, a);
 }
 
-static void FUNCC(pred4x4_horizontal)(uint8_t *_src, const uint8_t *topright, int _stride){
+static void FUNCC(pred4x4_horizontal)(uint8_t *_src, const uint8_t *topright,
+                                      ptrdiff_t _stride)
+{
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
     AV_WN4PA(src+0*stride, PIXEL_SPLAT_X4(src[-1+0*stride]));
@@ -49,7 +53,9 @@ static void FUNCC(pred4x4_horizontal)(uint8_t *_src, const uint8_t *topright, in
     AV_WN4PA(src+3*stride, PIXEL_SPLAT_X4(src[-1+3*stride]));
 }
 
-static void FUNCC(pred4x4_dc)(uint8_t *_src, const uint8_t *topright, int _stride){
+static void FUNCC(pred4x4_dc)(uint8_t *_src, const uint8_t *topright,
+                              ptrdiff_t _stride)
+{
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
     const int dc= (  src[-stride] + src[1-stride] + src[2-stride] + src[3-stride]
@@ -62,7 +68,9 @@ static void FUNCC(pred4x4_dc)(uint8_t *_src, const uint8_t *topright, int _strid
     AV_WN4PA(src+3*stride, a);
 }
 
-static void FUNCC(pred4x4_left_dc)(uint8_t *_src, const uint8_t *topright, int _stride){
+static void FUNCC(pred4x4_left_dc)(uint8_t *_src, const uint8_t *topright,
+                                   ptrdiff_t _stride)
+{
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
     const int dc= (  src[-1+0*stride] + src[-1+1*stride] + src[-1+2*stride] + src[-1+3*stride] + 2) >>2;
@@ -74,7 +82,9 @@ static void FUNCC(pred4x4_left_dc)(uint8_t *_src, const uint8_t *topright, int _
     AV_WN4PA(src+3*stride, a);
 }
 
-static void FUNCC(pred4x4_top_dc)(uint8_t *_src, const uint8_t *topright, int _stride){
+static void FUNCC(pred4x4_top_dc)(uint8_t *_src, const uint8_t *topright,
+                                  ptrdiff_t _stride)
+{
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
     const int dc= (  src[-stride] + src[1-stride] + src[2-stride] + src[3-stride] + 2) >>2;
@@ -86,7 +96,9 @@ static void FUNCC(pred4x4_top_dc)(uint8_t *_src, const uint8_t *topright, int _s
     AV_WN4PA(src+3*stride, a);
 }
 
-static void FUNCC(pred4x4_128_dc)(uint8_t *_src, const uint8_t *topright, int _stride){
+static void FUNCC(pred4x4_128_dc)(uint8_t *_src, const uint8_t *topright,
+                                  ptrdiff_t _stride)
+{
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
     const pixel4 a = PIXEL_SPLAT_X4(1<<(BIT_DEPTH-1));
@@ -97,7 +109,9 @@ static void FUNCC(pred4x4_128_dc)(uint8_t *_src, const uint8_t *topright, int _s
     AV_WN4PA(src+3*stride, a);
 }
 
-static void FUNCC(pred4x4_127_dc)(uint8_t *_src, const uint8_t *topright, int _stride){
+static void FUNCC(pred4x4_127_dc)(uint8_t *_src, const uint8_t *topright,
+                                  ptrdiff_t _stride)
+{
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
     const pixel4 a = PIXEL_SPLAT_X4((1<<(BIT_DEPTH-1))-1);
@@ -108,7 +122,9 @@ static void FUNCC(pred4x4_127_dc)(uint8_t *_src, const uint8_t *topright, int _s
     AV_WN4PA(src+3*stride, a);
 }
 
-static void FUNCC(pred4x4_129_dc)(uint8_t *_src, const uint8_t *topright, int _stride){
+static void FUNCC(pred4x4_129_dc)(uint8_t *_src, const uint8_t *topright,
+                                  ptrdiff_t _stride)
+{
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
     const pixel4 a = PIXEL_SPLAT_X4((1<<(BIT_DEPTH-1))+1);
@@ -144,7 +160,9 @@ static void FUNCC(pred4x4_129_dc)(uint8_t *_src, const uint8_t *topright, int _s
     const unsigned av_unused t2 = src[ 2-1*stride];\
     const unsigned av_unused t3 = src[ 3-1*stride];\
 
-static void FUNCC(pred4x4_down_right)(uint8_t *_src, const uint8_t *topright, int _stride){
+static void FUNCC(pred4x4_down_right)(uint8_t *_src, const uint8_t *topright,
+                                      ptrdiff_t _stride)
+{
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
     const int lt= src[-1-1*stride];
@@ -169,7 +187,9 @@ static void FUNCC(pred4x4_down_right)(uint8_t *_src, const uint8_t *topright, in
     src[3+0*stride]=(t1 + 2*t2 + t3 + 2)>>2;
 }
 
-static void FUNCC(pred4x4_down_left)(uint8_t *_src, const uint8_t *_topright, int _stride){
+static void FUNCC(pred4x4_down_left)(uint8_t *_src, const uint8_t *_topright,
+                                     ptrdiff_t _stride)
+{
     pixel *src = (pixel*)_src;
     const pixel *topright = (const pixel*)_topright;
     int stride = _stride/sizeof(pixel);
@@ -195,7 +215,10 @@ static void FUNCC(pred4x4_down_left)(uint8_t *_src, const uint8_t *_topright, in
     src[3+3*stride]=(t6 + 3*t7 + 2)>>2;
 }
 
-static void FUNCC(pred4x4_vertical_right)(uint8_t *_src, const uint8_t *topright, int _stride){
+static void FUNCC(pred4x4_vertical_right)(uint8_t *_src,
+                                          const uint8_t *topright,
+                                          ptrdiff_t _stride)
+{
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
     const int lt= src[-1-1*stride];
@@ -220,7 +243,10 @@ static void FUNCC(pred4x4_vertical_right)(uint8_t *_src, const uint8_t *topright
     src[0+3*stride]=(l0 + 2*l1 + l2 + 2)>>2;
 }
 
-static void FUNCC(pred4x4_vertical_left)(uint8_t *_src, const uint8_t *_topright, int _stride){
+static void FUNCC(pred4x4_vertical_left)(uint8_t *_src,
+                                         const uint8_t *_topright,
+                                         ptrdiff_t _stride)
+{
     pixel *src = (pixel*)_src;
     const pixel *topright = (const pixel*)_topright;
     int stride = _stride/sizeof(pixel);
@@ -245,7 +271,9 @@ static void FUNCC(pred4x4_vertical_left)(uint8_t *_src, const uint8_t *_topright
     src[3+3*stride]=(t4 + 2*t5 + t6 + 2)>>2;
 }
 
-static void FUNCC(pred4x4_horizontal_up)(uint8_t *_src, const uint8_t *topright, int _stride){
+static void FUNCC(pred4x4_horizontal_up)(uint8_t *_src, const uint8_t *topright,
+                                         ptrdiff_t _stride)
+{
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
     LOAD_LEFT_EDGE
@@ -268,7 +296,10 @@ static void FUNCC(pred4x4_horizontal_up)(uint8_t *_src, const uint8_t *topright,
     src[3+3*stride]=l3;
 }
 
-static void FUNCC(pred4x4_horizontal_down)(uint8_t *_src, const uint8_t *topright, int _stride){
+static void FUNCC(pred4x4_horizontal_down)(uint8_t *_src,
+                                           const uint8_t *topright,
+                                           ptrdiff_t _stride)
+{
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
     const int lt= src[-1-1*stride];
@@ -293,7 +324,8 @@ static void FUNCC(pred4x4_horizontal_down)(uint8_t *_src, const uint8_t *toprigh
     src[1+3*stride]=(l1 + 2*l2 + l3 + 2)>>2;
 }
 
-static void FUNCC(pred16x16_vertical)(uint8_t *_src, int _stride){
+static void FUNCC(pred16x16_vertical)(uint8_t *_src, ptrdiff_t _stride)
+{
     int i;
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
@@ -310,7 +342,8 @@ static void FUNCC(pred16x16_vertical)(uint8_t *_src, int _stride){
     }
 }
 
-static void FUNCC(pred16x16_horizontal)(uint8_t *_src, int stride){
+static void FUNCC(pred16x16_horizontal)(uint8_t *_src, ptrdiff_t stride)
+{
     int i;
     pixel *src = (pixel*)_src;
     stride /= sizeof(pixel);
@@ -334,7 +367,8 @@ static void FUNCC(pred16x16_horizontal)(uint8_t *_src, int stride){
         src += stride;\
     }
 
-static void FUNCC(pred16x16_dc)(uint8_t *_src, int stride){
+static void FUNCC(pred16x16_dc)(uint8_t *_src, ptrdiff_t stride)
+{
     int i, dc=0;
     pixel *src = (pixel*)_src;
     pixel4 dcsplat;
@@ -352,7 +386,8 @@ static void FUNCC(pred16x16_dc)(uint8_t *_src, int stride){
     PREDICT_16x16_DC(dcsplat);
 }
 
-static void FUNCC(pred16x16_left_dc)(uint8_t *_src, int stride){
+static void FUNCC(pred16x16_left_dc)(uint8_t *_src, ptrdiff_t stride)
+{
     int i, dc=0;
     pixel *src = (pixel*)_src;
     pixel4 dcsplat;
@@ -366,7 +401,8 @@ static void FUNCC(pred16x16_left_dc)(uint8_t *_src, int stride){
     PREDICT_16x16_DC(dcsplat);
 }
 
-static void FUNCC(pred16x16_top_dc)(uint8_t *_src, int stride){
+static void FUNCC(pred16x16_top_dc)(uint8_t *_src, ptrdiff_t stride)
+{
     int i, dc=0;
     pixel *src = (pixel*)_src;
     pixel4 dcsplat;
@@ -381,7 +417,8 @@ static void FUNCC(pred16x16_top_dc)(uint8_t *_src, int stride){
 }
 
 #define PRED16x16_X(n, v) \
-static void FUNCC(pred16x16_##n##_dc)(uint8_t *_src, int stride){\
+static void FUNCC(pred16x16_##n##_dc)(uint8_t *_src, ptrdiff_t stride)\
+{\
     int i;\
     pixel *src = (pixel*)_src;\
     stride /= sizeof(pixel);\
@@ -392,7 +429,11 @@ PRED16x16_X(127, (1<<(BIT_DEPTH-1))-1)
 PRED16x16_X(128, (1<<(BIT_DEPTH-1))+0)
 PRED16x16_X(129, (1<<(BIT_DEPTH-1))+1)
 
-static inline void FUNCC(pred16x16_plane_compat)(uint8_t *_src, int _stride, const int svq3, const int rv40){
+static inline void FUNCC(pred16x16_plane_compat)(uint8_t *_src,
+                                                 ptrdiff_t _stride,
+                                                 const int svq3,
+                                                 const int rv40)
+{
   int i, j, k;
   int a;
   INIT_CLIP
@@ -437,11 +478,13 @@ static inline void FUNCC(pred16x16_plane_compat)(uint8_t *_src, int _stride, con
   }
 }
 
-static void FUNCC(pred16x16_plane)(uint8_t *src, int stride){
+static void FUNCC(pred16x16_plane)(uint8_t *src, ptrdiff_t stride)
+{
     FUNCC(pred16x16_plane_compat)(src, stride, 0, 0);
 }
 
-static void FUNCC(pred8x8_vertical)(uint8_t *_src, int _stride){
+static void FUNCC(pred8x8_vertical)(uint8_t *_src, ptrdiff_t _stride)
+{
     int i;
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
@@ -454,7 +497,8 @@ static void FUNCC(pred8x8_vertical)(uint8_t *_src, int _stride){
     }
 }
 
-static void FUNCC(pred8x16_vertical)(uint8_t *_src, int _stride){
+static void FUNCC(pred8x16_vertical)(uint8_t *_src, ptrdiff_t _stride)
+{
     int i;
     pixel *src = (pixel*)_src;
     int stride = _stride>>(sizeof(pixel)-1);
@@ -467,7 +511,8 @@ static void FUNCC(pred8x16_vertical)(uint8_t *_src, int _stride){
     }
 }
 
-static void FUNCC(pred8x8_horizontal)(uint8_t *_src, int stride){
+static void FUNCC(pred8x8_horizontal)(uint8_t *_src, ptrdiff_t stride)
+{
     int i;
     pixel *src = (pixel*)_src;
     stride /= sizeof(pixel);
@@ -479,7 +524,8 @@ static void FUNCC(pred8x8_horizontal)(uint8_t *_src, int stride){
     }
 }
 
-static void FUNCC(pred8x16_horizontal)(uint8_t *_src, int stride){
+static void FUNCC(pred8x16_horizontal)(uint8_t *_src, ptrdiff_t stride)
+{
     int i;
     pixel *src = (pixel*)_src;
     stride >>= sizeof(pixel)-1;
@@ -491,7 +537,8 @@ static void FUNCC(pred8x16_horizontal)(uint8_t *_src, int stride){
 }
 
 #define PRED8x8_X(n, v)\
-static void FUNCC(pred8x8_##n##_dc)(uint8_t *_src, int stride){\
+static void FUNCC(pred8x8_##n##_dc)(uint8_t *_src, ptrdiff_t stride)\
+{\
     int i;\
     const pixel4 a = PIXEL_SPLAT_X4(v);\
     pixel *src = (pixel*)_src;\
@@ -506,12 +553,14 @@ PRED8x8_X(127, (1<<(BIT_DEPTH-1))-1)
 PRED8x8_X(128, (1<<(BIT_DEPTH-1))+0)
 PRED8x8_X(129, (1<<(BIT_DEPTH-1))+1)
 
-static void FUNCC(pred8x16_128_dc)(uint8_t *_src, int stride){
+static void FUNCC(pred8x16_128_dc)(uint8_t *_src, ptrdiff_t stride)
+{
     FUNCC(pred8x8_128_dc)(_src, stride);
     FUNCC(pred8x8_128_dc)(_src+8*stride, stride);
 }
 
-static void FUNCC(pred8x8_left_dc)(uint8_t *_src, int stride){
+static void FUNCC(pred8x8_left_dc)(uint8_t *_src, ptrdiff_t stride)
+{
     int i;
     int dc0, dc2;
     pixel4 dc0splat, dc2splat;
@@ -536,12 +585,14 @@ static void FUNCC(pred8x8_left_dc)(uint8_t *_src, int stride){
     }
 }
 
-static void FUNCC(pred8x16_left_dc)(uint8_t *_src, int stride){
+static void FUNCC(pred8x16_left_dc)(uint8_t *_src, ptrdiff_t stride)
+{
     FUNCC(pred8x8_left_dc)(_src, stride);
     FUNCC(pred8x8_left_dc)(_src+8*stride, stride);
 }
 
-static void FUNCC(pred8x8_top_dc)(uint8_t *_src, int stride){
+static void FUNCC(pred8x8_top_dc)(uint8_t *_src, ptrdiff_t stride)
+{
     int i;
     int dc0, dc1;
     pixel4 dc0splat, dc1splat;
@@ -566,7 +617,8 @@ static void FUNCC(pred8x8_top_dc)(uint8_t *_src, int stride){
     }
 }
 
-static void FUNCC(pred8x16_top_dc)(uint8_t *_src, int stride){
+static void FUNCC(pred8x16_top_dc)(uint8_t *_src, ptrdiff_t stride)
+{
     int i;
     int dc0, dc1;
     pixel4 dc0splat, dc1splat;
@@ -587,7 +639,8 @@ static void FUNCC(pred8x16_top_dc)(uint8_t *_src, int stride){
     }
 }
 
-static void FUNCC(pred8x8_dc)(uint8_t *_src, int stride){
+static void FUNCC(pred8x8_dc)(uint8_t *_src, ptrdiff_t stride)
+{
     int i;
     int dc0, dc1, dc2;
     pixel4 dc0splat, dc1splat, dc2splat, dc3splat;
@@ -615,7 +668,8 @@ static void FUNCC(pred8x8_dc)(uint8_t *_src, int stride){
     }
 }
 
-static void FUNCC(pred8x16_dc)(uint8_t *_src, int stride){
+static void FUNCC(pred8x16_dc)(uint8_t *_src, ptrdiff_t stride)
+{
     int i;
     int dc0, dc1, dc2, dc3, dc4;
     pixel4 dc0splat, dc1splat, dc2splat, dc3splat, dc4splat, dc5splat, dc6splat, dc7splat;
@@ -657,51 +711,60 @@ static void FUNCC(pred8x16_dc)(uint8_t *_src, int stride){
     }
 }
 
-static void FUNC(pred8x8_mad_cow_dc_l0t)(uint8_t *src, int stride){
+static void FUNC(pred8x8_mad_cow_dc_l0t)(uint8_t *src, ptrdiff_t stride)
+{
     FUNCC(pred8x8_top_dc)(src, stride);
     FUNCC(pred4x4_dc)(src, NULL, stride);
 }
 
-static void FUNC(pred8x16_mad_cow_dc_l0t)(uint8_t *src, int stride){
+static void FUNC(pred8x16_mad_cow_dc_l0t)(uint8_t *src, ptrdiff_t stride)
+{
     FUNCC(pred8x16_top_dc)(src, stride);
     FUNCC(pred4x4_dc)(src, NULL, stride);
 }
 
-static void FUNC(pred8x8_mad_cow_dc_0lt)(uint8_t *src, int stride){
+static void FUNC(pred8x8_mad_cow_dc_0lt)(uint8_t *src, ptrdiff_t stride)
+{
     FUNCC(pred8x8_dc)(src, stride);
     FUNCC(pred4x4_top_dc)(src, NULL, stride);
 }
 
-static void FUNC(pred8x16_mad_cow_dc_0lt)(uint8_t *src, int stride){
+static void FUNC(pred8x16_mad_cow_dc_0lt)(uint8_t *src, ptrdiff_t stride)
+{
     FUNCC(pred8x16_dc)(src, stride);
     FUNCC(pred4x4_top_dc)(src, NULL, stride);
 }
 
-static void FUNC(pred8x8_mad_cow_dc_l00)(uint8_t *src, int stride){
+static void FUNC(pred8x8_mad_cow_dc_l00)(uint8_t *src, ptrdiff_t stride)
+{
     FUNCC(pred8x8_left_dc)(src, stride);
     FUNCC(pred4x4_128_dc)(src + 4*stride                  , NULL, stride);
     FUNCC(pred4x4_128_dc)(src + 4*stride + 4*sizeof(pixel), NULL, stride);
 }
 
-static void FUNC(pred8x16_mad_cow_dc_l00)(uint8_t *src, int stride){
+static void FUNC(pred8x16_mad_cow_dc_l00)(uint8_t *src, ptrdiff_t stride)
+{
     FUNCC(pred8x16_left_dc)(src, stride);
     FUNCC(pred4x4_128_dc)(src + 4*stride                  , NULL, stride);
     FUNCC(pred4x4_128_dc)(src + 4*stride + 4*sizeof(pixel), NULL, stride);
 }
 
-static void FUNC(pred8x8_mad_cow_dc_0l0)(uint8_t *src, int stride){
+static void FUNC(pred8x8_mad_cow_dc_0l0)(uint8_t *src, ptrdiff_t stride)
+{
     FUNCC(pred8x8_left_dc)(src, stride);
     FUNCC(pred4x4_128_dc)(src                  , NULL, stride);
     FUNCC(pred4x4_128_dc)(src + 4*sizeof(pixel), NULL, stride);
 }
 
-static void FUNC(pred8x16_mad_cow_dc_0l0)(uint8_t *src, int stride){
+static void FUNC(pred8x16_mad_cow_dc_0l0)(uint8_t *src, ptrdiff_t stride)
+{
     FUNCC(pred8x16_left_dc)(src, stride);
     FUNCC(pred4x4_128_dc)(src                  , NULL, stride);
     FUNCC(pred4x4_128_dc)(src + 4*sizeof(pixel), NULL, stride);
 }
 
-static void FUNCC(pred8x8_plane)(uint8_t *_src, int _stride){
+static void FUNCC(pred8x8_plane)(uint8_t *_src, ptrdiff_t _stride)
+{
   int j, k;
   int a;
   INIT_CLIP
@@ -736,7 +799,8 @@ static void FUNCC(pred8x8_plane)(uint8_t *_src, int _stride){
   }
 }
 
-static void FUNCC(pred8x16_plane)(uint8_t *_src, int _stride){
+static void FUNCC(pred8x16_plane)(uint8_t *_src, ptrdiff_t _stride)
+{
   int j, k;
   int a;
   INIT_CLIP
@@ -815,14 +879,16 @@ static void FUNCC(pred8x16_plane)(uint8_t *_src, int _stride){
         src += stride; \
     }
 
-static void FUNCC(pred8x8l_128_dc)(uint8_t *_src, int has_topleft, int has_topright, int _stride)
+static void FUNCC(pred8x8l_128_dc)(uint8_t *_src, int has_topleft,
+                                   int has_topright, ptrdiff_t _stride)
 {
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
 
     PREDICT_8x8_DC(PIXEL_SPLAT_X4(1<<(BIT_DEPTH-1)));
 }
-static void FUNCC(pred8x8l_left_dc)(uint8_t *_src, int has_topleft, int has_topright, int _stride)
+static void FUNCC(pred8x8l_left_dc)(uint8_t *_src, int has_topleft,
+                                    int has_topright, ptrdiff_t _stride)
 {
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
@@ -831,7 +897,8 @@ static void FUNCC(pred8x8l_left_dc)(uint8_t *_src, int has_topleft, int has_topr
     const pixel4 dc = PIXEL_SPLAT_X4((l0+l1+l2+l3+l4+l5+l6+l7+4) >> 3);
     PREDICT_8x8_DC(dc);
 }
-static void FUNCC(pred8x8l_top_dc)(uint8_t *_src, int has_topleft, int has_topright, int _stride)
+static void FUNCC(pred8x8l_top_dc)(uint8_t *_src, int has_topleft,
+                                   int has_topright, ptrdiff_t _stride)
 {
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
@@ -840,7 +907,8 @@ static void FUNCC(pred8x8l_top_dc)(uint8_t *_src, int has_topleft, int has_topri
     const pixel4 dc = PIXEL_SPLAT_X4((t0+t1+t2+t3+t4+t5+t6+t7+4) >> 3);
     PREDICT_8x8_DC(dc);
 }
-static void FUNCC(pred8x8l_dc)(uint8_t *_src, int has_topleft, int has_topright, int _stride)
+static void FUNCC(pred8x8l_dc)(uint8_t *_src, int has_topleft,
+                               int has_topright, ptrdiff_t _stride)
 {
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
@@ -851,7 +919,8 @@ static void FUNCC(pred8x8l_dc)(uint8_t *_src, int has_topleft, int has_topright,
                                      +t0+t1+t2+t3+t4+t5+t6+t7+8) >> 4);
     PREDICT_8x8_DC(dc);
 }
-static void FUNCC(pred8x8l_horizontal)(uint8_t *_src, int has_topleft, int has_topright, int _stride)
+static void FUNCC(pred8x8l_horizontal)(uint8_t *_src, int has_topleft,
+                                       int has_topright, ptrdiff_t _stride)
 {
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
@@ -864,7 +933,8 @@ static void FUNCC(pred8x8l_horizontal)(uint8_t *_src, int has_topleft, int has_t
     ROW(0); ROW(1); ROW(2); ROW(3); ROW(4); ROW(5); ROW(6); ROW(7);
 #undef ROW
 }
-static void FUNCC(pred8x8l_vertical)(uint8_t *_src, int has_topleft, int has_topright, int _stride)
+static void FUNCC(pred8x8l_vertical)(uint8_t *_src, int has_topleft,
+                                     int has_topright, ptrdiff_t _stride)
 {
     int y;
     pixel *src = (pixel*)_src;
@@ -887,7 +957,8 @@ static void FUNCC(pred8x8l_vertical)(uint8_t *_src, int has_topleft, int has_top
         AV_WN4PA(((pixel4*)(src+y*stride))+1, b);
     }
 }
-static void FUNCC(pred8x8l_down_left)(uint8_t *_src, int has_topleft, int has_topright, int _stride)
+static void FUNCC(pred8x8l_down_left)(uint8_t *_src, int has_topleft,
+                                      int has_topright, ptrdiff_t _stride)
 {
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
@@ -909,7 +980,8 @@ static void FUNCC(pred8x8l_down_left)(uint8_t *_src, int has_topleft, int has_to
     SRC(6,7)=SRC(7,6)= (t13 + 2*t14 + t15 + 2) >> 2;
     SRC(7,7)= (t14 + 3*t15 + 2) >> 2;
 }
-static void FUNCC(pred8x8l_down_right)(uint8_t *_src, int has_topleft, int has_topright, int _stride)
+static void FUNCC(pred8x8l_down_right)(uint8_t *_src, int has_topleft,
+                                       int has_topright, ptrdiff_t _stride)
 {
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
@@ -932,7 +1004,8 @@ static void FUNCC(pred8x8l_down_right)(uint8_t *_src, int has_topleft, int has_t
     SRC(6,0)=SRC(7,1)= (t4 + 2*t5 + t6 + 2) >> 2;
     SRC(7,0)= (t5 + 2*t6 + t7 + 2) >> 2;
 }
-static void FUNCC(pred8x8l_vertical_right)(uint8_t *_src, int has_topleft, int has_topright, int _stride)
+static void FUNCC(pred8x8l_vertical_right)(uint8_t *_src, int has_topleft,
+                                           int has_topright, ptrdiff_t _stride)
 {
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
@@ -962,7 +1035,8 @@ static void FUNCC(pred8x8l_vertical_right)(uint8_t *_src, int has_topleft, int h
     SRC(7,1)= (t5 + 2*t6 + t7 + 2) >> 2;
     SRC(7,0)= (t6 + t7 + 1) >> 1;
 }
-static void FUNCC(pred8x8l_horizontal_down)(uint8_t *_src, int has_topleft, int has_topright, int _stride)
+static void FUNCC(pred8x8l_horizontal_down)(uint8_t *_src, int has_topleft,
+                                            int has_topright, ptrdiff_t _stride)
 {
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
@@ -992,7 +1066,8 @@ static void FUNCC(pred8x8l_horizontal_down)(uint8_t *_src, int has_topleft, int
     SRC(6,0)= (t5 + 2*t4 + t3 + 2) >> 2;
     SRC(7,0)= (t6 + 2*t5 + t4 + 2) >> 2;
 }
-static void FUNCC(pred8x8l_vertical_left)(uint8_t *_src, int has_topleft, int has_topright, int _stride)
+static void FUNCC(pred8x8l_vertical_left)(uint8_t *_src, int has_topleft,
+                                          int has_topright, ptrdiff_t _stride)
 {
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
@@ -1021,7 +1096,8 @@ static void FUNCC(pred8x8l_vertical_left)(uint8_t *_src, int has_topleft, int ha
     SRC(7,6)= (t10 + t11 + 1) >> 1;
     SRC(7,7)= (t10 + 2*t11 + t12 + 2) >> 2;
 }
-static void FUNCC(pred8x8l_horizontal_up)(uint8_t *_src, int has_topleft, int has_topright, int _stride)
+static void FUNCC(pred8x8l_horizontal_up)(uint8_t *_src, int has_topleft,
+                                          int has_topright, ptrdiff_t _stride)
 {
     pixel *src = (pixel*)_src;
     int stride = _stride/sizeof(pixel);
@@ -1055,7 +1131,9 @@ static void FUNCC(pred8x8l_horizontal_up)(uint8_t *_src, int has_topleft, int ha
 #undef PL
 #undef SRC
 
-static void FUNCC(pred4x4_vertical_add)(uint8_t *_pix, const DCTELEM *_block, int stride){
+static void FUNCC(pred4x4_vertical_add)(uint8_t *_pix, const DCTELEM *_block,
+                                        ptrdiff_t stride)
+{
     int i;
     pixel *pix = (pixel*)_pix;
     const dctcoef *block = (const dctcoef*)_block;
@@ -1072,7 +1150,9 @@ static void FUNCC(pred4x4_vertical_add)(uint8_t *_pix, const DCTELEM *_block, in
     }
 }
 
-static void FUNCC(pred4x4_horizontal_add)(uint8_t *_pix, const DCTELEM *_block, int stride){
+static void FUNCC(pred4x4_horizontal_add)(uint8_t *_pix, const DCTELEM *_block,
+                                          ptrdiff_t stride)
+{
     int i;
     pixel *pix = (pixel*)_pix;
     const dctcoef *block = (const dctcoef*)_block;
@@ -1088,7 +1168,9 @@ static void FUNCC(pred4x4_horizontal_add)(uint8_t *_pix, const DCTELEM *_block,
     }
 }
 
-static void FUNCC(pred8x8l_vertical_add)(uint8_t *_pix, const DCTELEM *_block, int stride){
+static void FUNCC(pred8x8l_vertical_add)(uint8_t *_pix, const DCTELEM *_block,
+                                         ptrdiff_t stride)
+{
     int i;
     pixel *pix = (pixel*)_pix;
     const dctcoef *block = (const dctcoef*)_block;
@@ -1109,7 +1191,9 @@ static void FUNCC(pred8x8l_vertical_add)(uint8_t *_pix, const DCTELEM *_block, i
     }
 }
 
-static void FUNCC(pred8x8l_horizontal_add)(uint8_t *_pix, const DCTELEM *_block, int stride){
+static void FUNCC(pred8x8l_horizontal_add)(uint8_t *_pix, const DCTELEM *_block,
+                                           ptrdiff_t stride)
+{
     int i;
     pixel *pix = (pixel*)_pix;
     const dctcoef *block = (const dctcoef*)_block;
@@ -1129,25 +1213,36 @@ static void FUNCC(pred8x8l_horizontal_add)(uint8_t *_pix, const DCTELEM *_block,
     }
 }
 
-static void FUNCC(pred16x16_vertical_add)(uint8_t *pix, const int *block_offset, const DCTELEM *block, int stride){
+static void FUNCC(pred16x16_vertical_add)(uint8_t *pix, const int *block_offset,
+                                          const DCTELEM *block,
+                                          ptrdiff_t stride)
+{
     int i;
     for(i=0; i<16; i++)
         FUNCC(pred4x4_vertical_add)(pix + block_offset[i], block + i*16*sizeof(pixel), stride);
 }
 
-static void FUNCC(pred16x16_horizontal_add)(uint8_t *pix, const int *block_offset, const DCTELEM *block, int stride){
+static void FUNCC(pred16x16_horizontal_add)(uint8_t *pix,
+                                            const int *block_offset,
+                                            const DCTELEM *block,
+                                            ptrdiff_t stride)
+{
     int i;
     for(i=0; i<16; i++)
         FUNCC(pred4x4_horizontal_add)(pix + block_offset[i], block + i*16*sizeof(pixel), stride);
 }
 
-static void FUNCC(pred8x8_vertical_add)(uint8_t *pix, const int *block_offset, const DCTELEM *block, int stride){
+static void FUNCC(pred8x8_vertical_add)(uint8_t *pix, const int *block_offset,
+                                        const DCTELEM *block, ptrdiff_t stride)
+{
     int i;
     for(i=0; i<4; i++)
         FUNCC(pred4x4_vertical_add)(pix + block_offset[i], block + i*16*sizeof(pixel), stride);
 }
 
-static void FUNCC(pred8x16_vertical_add)(uint8_t *pix, const int *block_offset, const DCTELEM *block, int stride){
+static void FUNCC(pred8x16_vertical_add)(uint8_t *pix, const int *block_offset,
+                                         const DCTELEM *block, ptrdiff_t stride)
+{
     int i;
     for(i=0; i<4; i++)
         FUNCC(pred4x4_vertical_add)(pix + block_offset[i], block + i*16*sizeof(pixel), stride);
@@ -1155,13 +1250,19 @@ static void FUNCC(pred8x16_vertical_add)(uint8_t *pix, const int *block_offset,
         FUNCC(pred4x4_vertical_add)(pix + block_offset[i+4], block + i*16*sizeof(pixel), stride);
 }
 
-static void FUNCC(pred8x8_horizontal_add)(uint8_t *pix, const int *block_offset, const DCTELEM *block, int stride){
+static void FUNCC(pred8x8_horizontal_add)(uint8_t *pix, const int *block_offset,
+                                          const DCTELEM *block,
+                                          ptrdiff_t stride)
+{
     int i;
     for(i=0; i<4; i++)
         FUNCC(pred4x4_horizontal_add)(pix + block_offset[i], block + i*16*sizeof(pixel), stride);
 }
 
-static void FUNCC(pred8x16_horizontal_add)(uint8_t *pix, const int *block_offset, const DCTELEM *block, int stride){
+static void FUNCC(pred8x16_horizontal_add)(uint8_t *pix,
+                                           const int *block_offset,
+                                           const DCTELEM *block, ptrdiff_t stride)
+{
     int i;
     for(i=0; i<4; i++)
         FUNCC(pred4x4_horizontal_add)(pix + block_offset[i], block + i*16*sizeof(pixel), stride);
diff --git a/libavcodec/x86/h264_intrapred_init.c b/libavcodec/x86/h264_intrapred_init.c
index a686473d75..eebd137a6e 100644
--- a/libavcodec/x86/h264_intrapred_init.c
+++ b/libavcodec/x86/h264_intrapred_init.c
@@ -23,7 +23,9 @@
 #include "libavcodec/h264pred.h"
 
 #define PRED4x4(TYPE, DEPTH, OPT) \
-void ff_pred4x4_ ## TYPE ## _ ## DEPTH ## _ ## OPT (uint8_t *src, const uint8_t *topright, int stride);
+void ff_pred4x4_ ## TYPE ## _ ## DEPTH ## _ ## OPT (uint8_t *src, \
+                                                    const uint8_t *topright, \
+                                                    ptrdiff_t stride);
 
 PRED4x4(dc, 10, mmx2)
 PRED4x4(down_left, 10, sse2)
@@ -42,7 +44,8 @@ PRED4x4(horizontal_down, 10, ssse3)
 PRED4x4(horizontal_down, 10, avx)
 
 #define PRED8x8(TYPE, DEPTH, OPT) \
-void ff_pred8x8_ ## TYPE ## _ ## DEPTH ## _ ## OPT (uint8_t *src, int stride);
+void ff_pred8x8_ ## TYPE ## _ ## DEPTH ## _ ## OPT (uint8_t *src, \
+                                                    ptrdiff_t stride);
 
 PRED8x8(dc, 10, mmx2)
 PRED8x8(dc, 10, sse2)
@@ -52,7 +55,10 @@ PRED8x8(vertical, 10, sse2)
 PRED8x8(horizontal, 10, sse2)
 
 #define PRED8x8L(TYPE, DEPTH, OPT)\
-void ff_pred8x8l_ ## TYPE ## _ ## DEPTH ## _ ## OPT (uint8_t *src, int has_topleft, int has_topright, int stride);
+void ff_pred8x8l_ ## TYPE ## _ ## DEPTH ## _ ## OPT (uint8_t *src, \
+                                                     int has_topleft, \
+                                                     int has_topright, \
+                                                     ptrdiff_t stride);
 
 PRED8x8L(dc, 10, sse2)
 PRED8x8L(dc, 10, avx)
@@ -79,7 +85,8 @@ PRED8x8L(horizontal_up, 10, ssse3)
 PRED8x8L(horizontal_up, 10, avx)
 
 #define PRED16x16(TYPE, DEPTH, OPT)\
-void ff_pred16x16_ ## TYPE ## _ ## DEPTH ## _ ## OPT (uint8_t *src, int stride);
+void ff_pred16x16_ ## TYPE ## _ ## DEPTH ## _ ## OPT (uint8_t *src, \
+                                                      ptrdiff_t stride);
 
 PRED16x16(dc, 10, mmx2)
 PRED16x16(dc, 10, sse2)