rv40dsp x86: use only one register, for both increment and loop counter

[ffmpeg] / libavcodec / vp8dsp.h

diff --git a/libavcodec/vp8dsp.h b/libavcodec/vp8dsp.h
index 81e19f435b032cc7bbb57d3876cc5fabe2c4a9f3..62cc0109899b96106bf203bed4d68ba33b66ed78 100644 (file)
--- a/libavcodec/vp8dsp.h
+++ b/libavcodec/vp8dsp.h
@@ -29,40 +29,44 @@
 
 #include "dsputil.h"
 
-typedef void (*vp8_mc_func)(uint8_t *dst/*align 8*/, int dstStride,
-                            uint8_t *src/*align 1*/, int srcStride,
+typedef void (*vp8_mc_func)(uint8_t *dst/*align 8*/, ptrdiff_t dstStride,
+                            uint8_t *src/*align 1*/, ptrdiff_t srcStride,
                             int h, int x, int y);
 
 typedef struct VP8DSPContext {
     void (*vp8_luma_dc_wht)(DCTELEM block[4][4][16], DCTELEM dc[16]);
     void (*vp8_luma_dc_wht_dc)(DCTELEM block[4][4][16], DCTELEM dc[16]);
-    void (*vp8_idct_add)(uint8_t *dst, DCTELEM block[16], int stride);
-    void (*vp8_idct_dc_add)(uint8_t *dst, DCTELEM block[16], int stride);
-    void (*vp8_idct_dc_add4y)(uint8_t *dst, DCTELEM block[4][16], int stride);
-    void (*vp8_idct_dc_add4uv)(uint8_t *dst, DCTELEM block[4][16], int stride);
+    void (*vp8_idct_add)(uint8_t *dst, DCTELEM block[16], ptrdiff_t stride);
+    void (*vp8_idct_dc_add)(uint8_t *dst, DCTELEM block[16], ptrdiff_t stride);
+    void (*vp8_idct_dc_add4y)(uint8_t *dst, DCTELEM block[4][16],
+                              ptrdiff_t stride);
+    void (*vp8_idct_dc_add4uv)(uint8_t *dst, DCTELEM block[4][16],
+                               ptrdiff_t stride);
 
     // loop filter applied to edges between macroblocks
-    void (*vp8_v_loop_filter16y)(uint8_t *dst, int stride,
+    void (*vp8_v_loop_filter16y)(uint8_t *dst, ptrdiff_t stride,
                                  int flim_E, int flim_I, int hev_thresh);
-    void (*vp8_h_loop_filter16y)(uint8_t *dst, int stride,
+    void (*vp8_h_loop_filter16y)(uint8_t *dst, ptrdiff_t stride,
                                  int flim_E, int flim_I, int hev_thresh);
-    void (*vp8_v_loop_filter8uv)(uint8_t *dstU, uint8_t *dstV, int stride,
+    void (*vp8_v_loop_filter8uv)(uint8_t *dstU, uint8_t *dstV, ptrdiff_t stride,
                                  int flim_E, int flim_I, int hev_thresh);
-    void (*vp8_h_loop_filter8uv)(uint8_t *dstU, uint8_t *dstV, int stride,
+    void (*vp8_h_loop_filter8uv)(uint8_t *dstU, uint8_t *dstV, ptrdiff_t stride,
                                  int flim_E, int flim_I, int hev_thresh);
 
     // loop filter applied to inner macroblock edges
-    void (*vp8_v_loop_filter16y_inner)(uint8_t *dst, int stride,
+    void (*vp8_v_loop_filter16y_inner)(uint8_t *dst, ptrdiff_t stride,
                                        int flim_E, int flim_I, int hev_thresh);
-    void (*vp8_h_loop_filter16y_inner)(uint8_t *dst, int stride,
+    void (*vp8_h_loop_filter16y_inner)(uint8_t *dst, ptrdiff_t stride,
                                        int flim_E, int flim_I, int hev_thresh);
-    void (*vp8_v_loop_filter8uv_inner)(uint8_t *dstU, uint8_t *dstV, int stride,
+    void (*vp8_v_loop_filter8uv_inner)(uint8_t *dstU, uint8_t *dstV,
+                                       ptrdiff_t stride,
                                        int flim_E, int flim_I, int hev_thresh);
-    void (*vp8_h_loop_filter8uv_inner)(uint8_t *dstU, uint8_t *dstV, int stride,
+    void (*vp8_h_loop_filter8uv_inner)(uint8_t *dstU, uint8_t *dstV,
+                                       ptrdiff_t stride,
                                        int flim_E, int flim_I, int hev_thresh);
 
-    void (*vp8_v_loop_filter_simple)(uint8_t *dst, int stride, int flim);
-    void (*vp8_h_loop_filter_simple)(uint8_t *dst, int stride, int flim);
+    void (*vp8_v_loop_filter_simple)(uint8_t *dst, ptrdiff_t stride, int flim);
+    void (*vp8_h_loop_filter_simple)(uint8_t *dst, ptrdiff_t stride, int flim);
 
     /**
      * first dimension: width>>3, height is assumed equal to width
@@ -76,9 +80,12 @@ typedef struct VP8DSPContext {
     vp8_mc_func put_vp8_bilinear_pixels_tab[3][3][3];
 } VP8DSPContext;
 
-void ff_put_vp8_pixels16_c(uint8_t *dst, uint8_t *src, int stride, int h, int x, int y);
-void ff_put_vp8_pixels8_c(uint8_t *dst, uint8_t *src, int stride, int h, int x, int y);
-void ff_put_vp8_pixels4_c(uint8_t *dst, uint8_t *src, int stride, int h, int x, int y);
+void ff_put_vp8_pixels16_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride,
+                           int h, int x, int y);
+void ff_put_vp8_pixels8_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride,
+                          int h, int x, int y);
+void ff_put_vp8_pixels4_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride,
+                          int h, int x, int y);
 
 void ff_vp8dsp_init(VP8DSPContext *c);
 void ff_vp8dsp_init_x86(VP8DSPContext *c);