h264chroma: Change type of stride parameters to ptrdiff_t

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

5 files changed, 31 insertions, 50 deletions

diff --git a/libavcodec/x86/h264_chromamc.asm b/libavcodec/x86/h264_chromamc.asm
index 1447940857..a9cac59f0f 100644
--- a/libavcodec/x86/h264_chromamc.asm
+++ b/libavcodec/x86/h264_chromamc.asm
@@ -105,11 +105,8 @@ SECTION .text
 %endif ; rv40
 ; void ff_put/avg_h264_chroma_mc8_*(uint8_t *dst /* align 8 */,
 ;                                   uint8_t *src /* align 1 */,
-;                                   int stride, int h, int mx, int my)
+;                                   ptrdiff_t stride, int h, int mx, int my)
 cglobal %1_%2_chroma_mc8%3, 6, 7 + extra_regs, 0
-%if ARCH_X86_64
-    movsxd        r2, r2d
-%endif
     mov          r6d, r5d
     or           r6d, r4d
     jne .at_least_one_non_zero
@@ -291,9 +288,6 @@ cglobal %1_%2_chroma_mc8%3, 6, 7 + extra_regs, 0
 %endif ; PIC
 %endif ; rv40
 cglobal %1_%2_chroma_mc4, 6, 6 + extra_regs, 0
-%if ARCH_X86_64
-    movsxd        r2, r2d
-%endif
     pxor          m7, m7
     movd          m2, r4d         ; x
     movd          m3, r5d         ; y
@@ -376,10 +370,6 @@ cglobal %1_%2_chroma_mc4, 6, 6 + extra_regs, 0
 
 %macro chroma_mc2_mmx_func 2
 cglobal %1_%2_chroma_mc2, 6, 7, 0
-%if ARCH_X86_64
-    movsxd        r2, r2d
-%endif
-
     mov          r6d, r4d
     shl          r4d, 16
     sub          r4d, r6d
@@ -465,9 +455,6 @@ chroma_mc4_mmx_func avg, rv40
 
 %macro chroma_mc8_ssse3_func 2-3
 cglobal %1_%2_chroma_mc8%3, 6, 7, 8
-%if ARCH_X86_64
-    movsxd        r2, r2d
-%endif
     mov          r6d, r5d
     or           r6d, r4d
     jne .at_least_one_non_zero
@@ -613,9 +600,6 @@ cglobal %1_%2_chroma_mc8%3, 6, 7, 8
 
 %macro chroma_mc4_ssse3_func 2
 cglobal %1_%2_chroma_mc4, 6, 7, 0
-%if ARCH_X86_64
-    movsxd        r2, r2d
-%endif
     mov           r6, r4
     shl          r4d, 8
     sub          r4d, r6d
diff --git a/libavcodec/x86/h264_chromamc_10bit.asm b/libavcodec/x86/h264_chromamc_10bit.asm
index 7b003515cc..ff53b91c42 100644
--- a/libavcodec/x86/h264_chromamc_10bit.asm
+++ b/libavcodec/x86/h264_chromamc_10bit.asm
@@ -57,12 +57,11 @@ SECTION .text
 %endmacro
 
 ;-----------------------------------------------------------------------------
-; void ff_put/avg_h264_chroma_mc8(pixel *dst, pixel *src, int stride, int h,
-;                                 int mx, int my)
+; void ff_put/avg_h264_chroma_mc8(pixel *dst, pixel *src, ptrdiff_t stride,
+;                                 int h, int mx, int my)
 ;-----------------------------------------------------------------------------
 %macro CHROMA_MC8 1
 cglobal %1_h264_chroma_mc8_10, 6,7,8
-    movsxdifnidn  r2, r2d
     mov          r6d, r5d
     or           r6d, r4d
     jne .at_least_one_non_zero
@@ -149,8 +148,8 @@ cglobal %1_h264_chroma_mc8_10, 6,7,8
 %endmacro
 
 ;-----------------------------------------------------------------------------
-; void ff_put/avg_h264_chroma_mc4(pixel *dst, pixel *src, int stride, int h,
-;                                 int mx, int my)
+; void ff_put/avg_h264_chroma_mc4(pixel *dst, pixel *src, ptrdiff_t stride,
+;                                 int h, int mx, int my)
 ;-----------------------------------------------------------------------------
 ;TODO: xmm mc4
 %macro MC4_OP 2
@@ -174,7 +173,6 @@ cglobal %1_h264_chroma_mc8_10, 6,7,8
 
 %macro CHROMA_MC4 1
 cglobal %1_h264_chroma_mc4_10, 6,6,7
-    movsxdifnidn  r2, r2d
     movd          m2, r4m         ; x
     movd          m3, r5m         ; y
     mova          m4, [pw_8]
@@ -200,12 +198,11 @@ cglobal %1_h264_chroma_mc4_10, 6,6,7
 %endmacro
 
 ;-----------------------------------------------------------------------------
-; void ff_put/avg_h264_chroma_mc2(pixel *dst, pixel *src, int stride, int h,
-;                                 int mx, int my)
+; void ff_put/avg_h264_chroma_mc2(pixel *dst, pixel *src, ptrdiff_t stride,
+;                                 int h, int mx, int my)
 ;-----------------------------------------------------------------------------
 %macro CHROMA_MC2 1
 cglobal %1_h264_chroma_mc2_10, 6,7
-    movsxdifnidn  r2, r2d
     mov          r6d, r4d
     shl          r4d, 16
     sub          r4d, r6d
diff --git a/libavcodec/x86/h264chroma_init.c b/libavcodec/x86/h264chroma_init.c
index 8ec8a79aba..0d5ff3d159 100644
--- a/libavcodec/x86/h264chroma_init.c
+++ b/libavcodec/x86/h264chroma_init.c
@@ -25,38 +25,38 @@
 #include "libavcodec/h264chroma.h"
 
 void ff_put_h264_chroma_mc8_rnd_mmx  (uint8_t *dst, uint8_t *src,
-                                      int stride, int h, int x, int y);
+                                      ptrdiff_t stride, int h, int x, int y);
 void ff_avg_h264_chroma_mc8_rnd_mmxext(uint8_t *dst, uint8_t *src,
-                                       int stride, int h, int x, int y);
+                                       ptrdiff_t stride, int h, int x, int y);
 void ff_avg_h264_chroma_mc8_rnd_3dnow(uint8_t *dst, uint8_t *src,
-                                      int stride, int h, int x, int y);
+                                      ptrdiff_t stride, int h, int x, int y);
 
 void ff_put_h264_chroma_mc4_mmx      (uint8_t *dst, uint8_t *src,
-                                      int stride, int h, int x, int y);
+                                      ptrdiff_t stride, int h, int x, int y);
 void ff_avg_h264_chroma_mc4_mmxext   (uint8_t *dst, uint8_t *src,
-                                      int stride, int h, int x, int y);
+                                      ptrdiff_t stride, int h, int x, int y);
 void ff_avg_h264_chroma_mc4_3dnow    (uint8_t *dst, uint8_t *src,
-                                      int stride, int h, int x, int y);
+                                      ptrdiff_t stride, int h, int x, int y);
 
 void ff_put_h264_chroma_mc2_mmxext   (uint8_t *dst, uint8_t *src,
-                                      int stride, int h, int x, int y);
+                                      ptrdiff_t stride, int h, int x, int y);
 void ff_avg_h264_chroma_mc2_mmxext   (uint8_t *dst, uint8_t *src,
-                                      int stride, int h, int x, int y);
+                                      ptrdiff_t stride, int h, int x, int y);
 
 void ff_put_h264_chroma_mc8_rnd_ssse3(uint8_t *dst, uint8_t *src,
-                                      int stride, int h, int x, int y);
+                                      ptrdiff_t stride, int h, int x, int y);
 void ff_put_h264_chroma_mc4_ssse3    (uint8_t *dst, uint8_t *src,
-                                      int stride, int h, int x, int y);
+                                      ptrdiff_t stride, int h, int x, int y);
 
 void ff_avg_h264_chroma_mc8_rnd_ssse3(uint8_t *dst, uint8_t *src,
-                                      int stride, int h, int x, int y);
+                                      ptrdiff_t stride, int h, int x, int y);
 void ff_avg_h264_chroma_mc4_ssse3    (uint8_t *dst, uint8_t *src,
-                                      int stride, int h, int x, int y);
+                                      ptrdiff_t stride, int h, int x, int y);
 
 #define CHROMA_MC(OP, NUM, DEPTH, OPT)                                  \
 void ff_ ## OP ## _h264_chroma_mc ## NUM ## _ ## DEPTH ## _ ## OPT      \
                                       (uint8_t *dst, uint8_t *src,      \
-                                       int stride, int h, int x, int y);
+                                       ptrdiff_t stride, int h, int x, int y);
 
 CHROMA_MC(put, 2, 10, mmxext)
 CHROMA_MC(avg, 2, 10, mmxext)
diff --git a/libavcodec/x86/rv40dsp_init.c b/libavcodec/x86/rv40dsp_init.c
index e006c76584..7bf3ecd1f3 100644
--- a/libavcodec/x86/rv40dsp_init.c
+++ b/libavcodec/x86/rv40dsp_init.c
@@ -34,18 +34,18 @@
 
 #if HAVE_YASM
 void ff_put_rv40_chroma_mc8_mmx  (uint8_t *dst, uint8_t *src,
-                                  int stride, int h, int x, int y);
+                                  ptrdiff_t stride, int h, int x, int y);
 void ff_avg_rv40_chroma_mc8_mmxext(uint8_t *dst, uint8_t *src,
-                                   int stride, int h, int x, int y);
+                                   ptrdiff_t stride, int h, int x, int y);
 void ff_avg_rv40_chroma_mc8_3dnow(uint8_t *dst, uint8_t *src,
-                                  int stride, int h, int x, int y);
+                                  ptrdiff_t stride, int h, int x, int y);
 
 void ff_put_rv40_chroma_mc4_mmx  (uint8_t *dst, uint8_t *src,
-                                  int stride, int h, int x, int y);
+                                  ptrdiff_t stride, int h, int x, int y);
 void ff_avg_rv40_chroma_mc4_mmxext(uint8_t *dst, uint8_t *src,
-                                   int stride, int h, int x, int y);
+                                   ptrdiff_t stride, int h, int x, int y);
 void ff_avg_rv40_chroma_mc4_3dnow(uint8_t *dst, uint8_t *src,
-                                  int stride, int h, int x, int y);
+                                  ptrdiff_t stride, int h, int x, int y);
 
 #define DECLARE_WEIGHT(opt) \
 void ff_rv40_weight_func_rnd_16_##opt(uint8_t *dst, uint8_t *src1, uint8_t *src2, \
diff --git a/libavcodec/x86/vc1dsp_init.c b/libavcodec/x86/vc1dsp_init.c
index aff4b264e3..8982ff908a 100644
--- a/libavcodec/x86/vc1dsp_init.c
+++ b/libavcodec/x86/vc1dsp_init.c
@@ -71,15 +71,15 @@ static void avg_vc1_mspel_mc00_mmxext(uint8_t *dst, const uint8_t *src,
 #endif /* HAVE_YASM */
 
 void ff_put_vc1_chroma_mc8_nornd_mmx  (uint8_t *dst, uint8_t *src,
-                                       int stride, int h, int x, int y);
+                                       ptrdiff_t stride, int h, int x, int y);
 void ff_avg_vc1_chroma_mc8_nornd_mmxext(uint8_t *dst, uint8_t *src,
-                                        int stride, int h, int x, int y);
+                                        ptrdiff_t stride, int h, int x, int y);
 void ff_avg_vc1_chroma_mc8_nornd_3dnow(uint8_t *dst, uint8_t *src,
-                                       int stride, int h, int x, int y);
+                                       ptrdiff_t stride, int h, int x, int y);
 void ff_put_vc1_chroma_mc8_nornd_ssse3(uint8_t *dst, uint8_t *src,
-                                       int stride, int h, int x, int y);
+                                       ptrdiff_t stride, int h, int x, int y);
 void ff_avg_vc1_chroma_mc8_nornd_ssse3(uint8_t *dst, uint8_t *src,
-                                       int stride, int h, int x, int y);
+                                       ptrdiff_t stride, int h, int x, int y);
 
 
 av_cold void ff_vc1dsp_init_x86(VC1DSPContext *dsp)