rv34: move inverse transform functions to DSP context
[ffmpeg.git] / libavcodec / rv40dsp.c
index b48c4e8..efc049f 100644 (file)
@@ -2,37 +2,38 @@
  * RV40 decoder motion compensation functions
  * Copyright (c) 2008 Konstantin Shishkov
  *
- * This file is part of FFmpeg.
+ * This file is part of Libav.
  *
- * FFmpeg is free software; you can redistribute it and/or
+ * Libav is free software; you can redistribute it and/or
  * modify it under the terms of the GNU Lesser General Public
  * License as published by the Free Software Foundation; either
  * version 2.1 of the License, or (at your option) any later version.
  *
- * FFmpeg is distributed in the hope that it will be useful,
+ * Libav is distributed in the hope that it will be useful,
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  * Lesser General Public License for more details.
  *
  * You should have received a copy of the GNU Lesser General Public
- * License along with FFmpeg; if not, write to the Free Software
+ * License along with Libav; if not, write to the Free Software
  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  */
 
 /**
- * @file rv40dsp.c
+ * @file
  * RV40 decoder motion compensation functions
  */
 
 #include "avcodec.h"
 #include "dsputil.h"
+#include "rv34dsp.h"
 
 #define RV40_LOWPASS(OPNAME, OP) \
 static av_unused void OPNAME ## rv40_qpel8_h_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride,\
                                                      const int h, const int C1, const int C2, const int SHIFT){\
     uint8_t *cm = ff_cropTbl + MAX_NEG_CROP;\
     int i;\
-    for(i=0; i<h; i++)\
+    for(i = 0; i < h; i++)\
     {\
         OP(dst[0], (src[-2] + src[ 3] - 5*(src[-1]+src[2]) + src[0]*C1 + src[1]*C2 + (1<<(SHIFT-1))) >> SHIFT);\
         OP(dst[1], (src[-1] + src[ 4] - 5*(src[ 0]+src[3]) + src[1]*C1 + src[2]*C2 + (1<<(SHIFT-1))) >> SHIFT);\
@@ -42,8 +43,8 @@ static av_unused void OPNAME ## rv40_qpel8_h_lowpass(uint8_t *dst, uint8_t *src,
         OP(dst[5], (src[ 3] + src[ 8] - 5*(src[ 4]+src[7]) + src[5]*C1 + src[6]*C2 + (1<<(SHIFT-1))) >> SHIFT);\
         OP(dst[6], (src[ 4] + src[ 9] - 5*(src[ 5]+src[8]) + src[6]*C1 + src[7]*C2 + (1<<(SHIFT-1))) >> SHIFT);\
         OP(dst[7], (src[ 5] + src[10] - 5*(src[ 6]+src[9]) + src[7]*C1 + src[8]*C2 + (1<<(SHIFT-1))) >> SHIFT);\
-        dst+=dstStride;\
-        src+=srcStride;\
+        dst += dstStride;\
+        src += srcStride;\
     }\
 }\
 \
@@ -51,21 +52,21 @@ static void OPNAME ## rv40_qpel8_v_lowpass(uint8_t *dst, uint8_t *src, int dstSt
                                            const int w, const int C1, const int C2, const int SHIFT){\
     uint8_t *cm = ff_cropTbl + MAX_NEG_CROP;\
     int i;\
-    for(i=0; i<w; i++)\
+    for(i = 0; i < w; i++)\
     {\
-        const int srcB = src[-2*srcStride];\
-        const int srcA = src[-1*srcStride];\
-        const int src0 = src[0 *srcStride];\
-        const int src1 = src[1 *srcStride];\
-        const int src2 = src[2 *srcStride];\
-        const int src3 = src[3 *srcStride];\
-        const int src4 = src[4 *srcStride];\
-        const int src5 = src[5 *srcStride];\
-        const int src6 = src[6 *srcStride];\
-        const int src7 = src[7 *srcStride];\
-        const int src8 = src[8 *srcStride];\
-        const int src9 = src[9 *srcStride];\
-        const int src10= src[10*srcStride];\
+        const int srcB  = src[-2*srcStride];\
+        const int srcA  = src[-1*srcStride];\
+        const int src0  = src[0 *srcStride];\
+        const int src1  = src[1 *srcStride];\
+        const int src2  = src[2 *srcStride];\
+        const int src3  = src[3 *srcStride];\
+        const int src4  = src[4 *srcStride];\
+        const int src5  = src[5 *srcStride];\
+        const int src6  = src[6 *srcStride];\
+        const int src7  = src[7 *srcStride];\
+        const int src8  = src[8 *srcStride];\
+        const int src9  = src[9 *srcStride];\
+        const int src10 = src[10*srcStride];\
         OP(dst[0*dstStride], (srcB + src3  - 5*(srcA+src2) + src0*C1 + src1*C2 + (1<<(SHIFT-1))) >> SHIFT);\
         OP(dst[1*dstStride], (srcA + src4  - 5*(src0+src3) + src1*C1 + src2*C2 + (1<<(SHIFT-1))) >> SHIFT);\
         OP(dst[2*dstStride], (src0 + src5  - 5*(src1+src4) + src2*C1 + src3*C2 + (1<<(SHIFT-1))) >> SHIFT);\
@@ -105,10 +106,6 @@ static void OPNAME ## rv40_qpel ## SIZE ## _mc10_c(uint8_t *dst, uint8_t *src, i
     OPNAME ## rv40_qpel ## SIZE ## _h_lowpass(dst, src, stride, stride, SIZE, 52, 20, 6);\
 }\
 \
-static void OPNAME ## rv40_qpel ## SIZE ## _mc20_c(uint8_t *dst, uint8_t *src, int stride){\
-    OPNAME ## rv40_qpel ## SIZE ## _h_lowpass(dst, src, stride, stride, SIZE, 20, 20, 5);\
-}\
-\
 static void OPNAME ## rv40_qpel ## SIZE ## _mc30_c(uint8_t *dst, uint8_t *src, int stride){\
     OPNAME ## rv40_qpel ## SIZE ## _h_lowpass(dst, src, stride, stride, SIZE, 20, 52, 6);\
 }\
@@ -119,46 +116,42 @@ static void OPNAME ## rv40_qpel ## SIZE ## _mc01_c(uint8_t *dst, uint8_t *src, i
 \
 static void OPNAME ## rv40_qpel ## SIZE ## _mc11_c(uint8_t *dst, uint8_t *src, int stride){\
     uint8_t full[SIZE*(SIZE+5)];\
-    uint8_t * const full_mid= full + SIZE*2;\
+    uint8_t * const full_mid = full + SIZE*2;\
     put_rv40_qpel ## SIZE ## _h_lowpass(full, src - 2*stride, SIZE, stride, SIZE+5, 52, 20, 6);\
     OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 52, 20, 6);\
 }\
 \
 static void OPNAME ## rv40_qpel ## SIZE ## _mc21_c(uint8_t *dst, uint8_t *src, int stride){\
     uint8_t full[SIZE*(SIZE+5)];\
-    uint8_t * const full_mid= full + SIZE*2;\
+    uint8_t * const full_mid = full + SIZE*2;\
     put_rv40_qpel ## SIZE ## _h_lowpass(full, src - 2*stride, SIZE, stride, SIZE+5, 20, 20, 5);\
     OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 52, 20, 6);\
 }\
 \
 static void OPNAME ## rv40_qpel ## SIZE ## _mc31_c(uint8_t *dst, uint8_t *src, int stride){\
     uint8_t full[SIZE*(SIZE+5)];\
-    uint8_t * const full_mid= full + SIZE*2;\
+    uint8_t * const full_mid = full + SIZE*2;\
     put_rv40_qpel ## SIZE ## _h_lowpass(full, src - 2*stride, SIZE, stride, SIZE+5, 20, 52, 6);\
     OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 52, 20, 6);\
 }\
 \
-static void OPNAME ## rv40_qpel ## SIZE ## _mc02_c(uint8_t *dst, uint8_t *src, int stride){\
-    OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, src, stride, stride, SIZE, 20, 20, 5);\
-}\
-\
 static void OPNAME ## rv40_qpel ## SIZE ## _mc12_c(uint8_t *dst, uint8_t *src, int stride){\
     uint8_t full[SIZE*(SIZE+5)];\
-    uint8_t * const full_mid= full + SIZE*2;\
+    uint8_t * const full_mid = full + SIZE*2;\
     put_rv40_qpel ## SIZE ## _h_lowpass(full, src - 2*stride, SIZE, stride, SIZE+5, 52, 20, 6);\
     OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 20, 20, 5);\
 }\
 \
 static void OPNAME ## rv40_qpel ## SIZE ## _mc22_c(uint8_t *dst, uint8_t *src, int stride){\
     uint8_t full[SIZE*(SIZE+5)];\
-    uint8_t * const full_mid= full + SIZE*2;\
+    uint8_t * const full_mid = full + SIZE*2;\
     put_rv40_qpel ## SIZE ## _h_lowpass(full, src - 2*stride, SIZE, stride, SIZE+5, 20, 20, 5);\
     OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 20, 20, 5);\
 }\
 \
 static void OPNAME ## rv40_qpel ## SIZE ## _mc32_c(uint8_t *dst, uint8_t *src, int stride){\
     uint8_t full[SIZE*(SIZE+5)];\
-    uint8_t * const full_mid= full + SIZE*2;\
+    uint8_t * const full_mid = full + SIZE*2;\
     put_rv40_qpel ## SIZE ## _h_lowpass(full, src - 2*stride, SIZE, stride, SIZE+5, 20, 52, 6);\
     OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 20, 20, 5);\
 }\
@@ -169,14 +162,14 @@ static void OPNAME ## rv40_qpel ## SIZE ## _mc03_c(uint8_t *dst, uint8_t *src, i
 \
 static void OPNAME ## rv40_qpel ## SIZE ## _mc13_c(uint8_t *dst, uint8_t *src, int stride){\
     uint8_t full[SIZE*(SIZE+5)];\
-    uint8_t * const full_mid= full + SIZE*2;\
+    uint8_t * const full_mid = full + SIZE*2;\
     put_rv40_qpel ## SIZE ## _h_lowpass(full, src - 2*stride, SIZE, stride, SIZE+5, 52, 20, 6);\
     OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 20, 52, 6);\
 }\
 \
 static void OPNAME ## rv40_qpel ## SIZE ## _mc23_c(uint8_t *dst, uint8_t *src, int stride){\
     uint8_t full[SIZE*(SIZE+5)];\
-    uint8_t * const full_mid= full + SIZE*2;\
+    uint8_t * const full_mid = full + SIZE*2;\
     put_rv40_qpel ## SIZE ## _h_lowpass(full, src - 2*stride, SIZE, stride, SIZE+5, 20, 20, 5);\
     OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 20, 52, 6);\
 }\
@@ -205,50 +198,50 @@ static const int rv40_bias[4][4] = {
 
 #define RV40_CHROMA_MC(OPNAME, OP)\
 static void OPNAME ## rv40_chroma_mc4_c(uint8_t *dst/*align 8*/, uint8_t *src/*align 1*/, int stride, int h, int x, int y){\
-    const int A=(8-x)*(8-y);\
-    const int B=(  x)*(8-y);\
-    const int C=(8-x)*(  y);\
-    const int D=(  x)*(  y);\
+    const int A = (8-x) * (8-y);\
+    const int B = (  x) * (8-y);\
+    const int C = (8-x) * (  y);\
+    const int D = (  x) * (  y);\
     int i;\
     int bias = rv40_bias[y>>1][x>>1];\
     \
     assert(x<8 && y<8 && x>=0 && y>=0);\
 \
     if(D){\
-        for(i=0; i<h; i++){\
+        for(i = 0; i < h; i++){\
             OP(dst[0], (A*src[0] + B*src[1] + C*src[stride+0] + D*src[stride+1] + bias));\
             OP(dst[1], (A*src[1] + B*src[2] + C*src[stride+1] + D*src[stride+2] + bias));\
             OP(dst[2], (A*src[2] + B*src[3] + C*src[stride+2] + D*src[stride+3] + bias));\
             OP(dst[3], (A*src[3] + B*src[4] + C*src[stride+3] + D*src[stride+4] + bias));\
-            dst+= stride;\
-            src+= stride;\
+            dst += stride;\
+            src += stride;\
         }\
     }else{\
-        const int E= B+C;\
-        const int step= C ? stride : 1;\
-        for(i=0; i<h; i++){\
+        const int E = B + C;\
+        const int step = C ? stride : 1;\
+        for(i = 0; i < h; i++){\
             OP(dst[0], (A*src[0] + E*src[step+0] + bias));\
             OP(dst[1], (A*src[1] + E*src[step+1] + bias));\
             OP(dst[2], (A*src[2] + E*src[step+2] + bias));\
             OP(dst[3], (A*src[3] + E*src[step+3] + bias));\
-            dst+= stride;\
-            src+= stride;\
+            dst += stride;\
+            src += stride;\
         }\
     }\
 }\
 \
 static void OPNAME ## rv40_chroma_mc8_c(uint8_t *dst/*align 8*/, uint8_t *src/*align 1*/, int stride, int h, int x, int y){\
-    const int A=(8-x)*(8-y);\
-    const int B=(  x)*(8-y);\
-    const int C=(8-x)*(  y);\
-    const int D=(  x)*(  y);\
+    const int A = (8-x) * (8-y);\
+    const int B = (  x) * (8-y);\
+    const int C = (8-x) * (  y);\
+    const int D = (  x) * (  y);\
     int i;\
     int bias = rv40_bias[y>>1][x>>1];\
     \
     assert(x<8 && y<8 && x>=0 && y>=0);\
 \
     if(D){\
-        for(i=0; i<h; i++){\
+        for(i = 0; i < h; i++){\
             OP(dst[0], (A*src[0] + B*src[1] + C*src[stride+0] + D*src[stride+1] + bias));\
             OP(dst[1], (A*src[1] + B*src[2] + C*src[stride+1] + D*src[stride+2] + bias));\
             OP(dst[2], (A*src[2] + B*src[3] + C*src[stride+2] + D*src[stride+3] + bias));\
@@ -257,13 +250,13 @@ static void OPNAME ## rv40_chroma_mc8_c(uint8_t *dst/*align 8*/, uint8_t *src/*a
             OP(dst[5], (A*src[5] + B*src[6] + C*src[stride+5] + D*src[stride+6] + bias));\
             OP(dst[6], (A*src[6] + B*src[7] + C*src[stride+6] + D*src[stride+7] + bias));\
             OP(dst[7], (A*src[7] + B*src[8] + C*src[stride+7] + D*src[stride+8] + bias));\
-            dst+= stride;\
-            src+= stride;\
+            dst += stride;\
+            src += stride;\
         }\
     }else{\
-        const int E= B+C;\
-        const int step= C ? stride : 1;\
-        for(i=0; i<h; i++){\
+        const int E = B + C;\
+        const int step = C ? stride : 1;\
+        for(i = 0; i < h; i++){\
             OP(dst[0], (A*src[0] + E*src[step+0] + bias));\
             OP(dst[1], (A*src[1] + E*src[step+1] + bias));\
             OP(dst[2], (A*src[2] + E*src[step+2] + bias));\
@@ -272,8 +265,8 @@ static void OPNAME ## rv40_chroma_mc8_c(uint8_t *dst/*align 8*/, uint8_t *src/*a
             OP(dst[5], (A*src[5] + E*src[step+5] + bias));\
             OP(dst[6], (A*src[6] + E*src[step+6] + bias));\
             OP(dst[7], (A*src[7] + E*src[step+7] + bias));\
-            dst+= stride;\
-            src+= stride;\
+            dst += stride;\
+            src += stride;\
         }\
     }\
 }
@@ -284,70 +277,100 @@ static void OPNAME ## rv40_chroma_mc8_c(uint8_t *dst/*align 8*/, uint8_t *src/*a
 RV40_CHROMA_MC(put_, op_put)
 RV40_CHROMA_MC(avg_, op_avg)
 
-void ff_rv40dsp_init(DSPContext* c, AVCodecContext *avctx) {
-    c->put_rv40_qpel_pixels_tab[0][ 0] = c->put_h264_qpel_pixels_tab[0][0];
-    c->put_rv40_qpel_pixels_tab[0][ 1] = put_rv40_qpel16_mc10_c;
-    c->put_rv40_qpel_pixels_tab[0][ 2] = put_rv40_qpel16_mc20_c;
-    c->put_rv40_qpel_pixels_tab[0][ 3] = put_rv40_qpel16_mc30_c;
-    c->put_rv40_qpel_pixels_tab[0][ 4] = put_rv40_qpel16_mc01_c;
-    c->put_rv40_qpel_pixels_tab[0][ 5] = put_rv40_qpel16_mc11_c;
-    c->put_rv40_qpel_pixels_tab[0][ 6] = put_rv40_qpel16_mc21_c;
-    c->put_rv40_qpel_pixels_tab[0][ 7] = put_rv40_qpel16_mc31_c;
-    c->put_rv40_qpel_pixels_tab[0][ 8] = put_rv40_qpel16_mc02_c;
-    c->put_rv40_qpel_pixels_tab[0][ 9] = put_rv40_qpel16_mc12_c;
-    c->put_rv40_qpel_pixels_tab[0][10] = put_rv40_qpel16_mc22_c;
-    c->put_rv40_qpel_pixels_tab[0][11] = put_rv40_qpel16_mc32_c;
-    c->put_rv40_qpel_pixels_tab[0][12] = put_rv40_qpel16_mc03_c;
-    c->put_rv40_qpel_pixels_tab[0][13] = put_rv40_qpel16_mc13_c;
-    c->put_rv40_qpel_pixels_tab[0][14] = put_rv40_qpel16_mc23_c;
-    c->avg_rv40_qpel_pixels_tab[0][ 0] = c->avg_h264_qpel_pixels_tab[0][0];
-    c->avg_rv40_qpel_pixels_tab[0][ 1] = avg_rv40_qpel16_mc10_c;
-    c->avg_rv40_qpel_pixels_tab[0][ 2] = avg_rv40_qpel16_mc20_c;
-    c->avg_rv40_qpel_pixels_tab[0][ 3] = avg_rv40_qpel16_mc30_c;
-    c->avg_rv40_qpel_pixels_tab[0][ 4] = avg_rv40_qpel16_mc01_c;
-    c->avg_rv40_qpel_pixels_tab[0][ 5] = avg_rv40_qpel16_mc11_c;
-    c->avg_rv40_qpel_pixels_tab[0][ 6] = avg_rv40_qpel16_mc21_c;
-    c->avg_rv40_qpel_pixels_tab[0][ 7] = avg_rv40_qpel16_mc31_c;
-    c->avg_rv40_qpel_pixels_tab[0][ 8] = avg_rv40_qpel16_mc02_c;
-    c->avg_rv40_qpel_pixels_tab[0][ 9] = avg_rv40_qpel16_mc12_c;
-    c->avg_rv40_qpel_pixels_tab[0][10] = avg_rv40_qpel16_mc22_c;
-    c->avg_rv40_qpel_pixels_tab[0][11] = avg_rv40_qpel16_mc32_c;
-    c->avg_rv40_qpel_pixels_tab[0][12] = avg_rv40_qpel16_mc03_c;
-    c->avg_rv40_qpel_pixels_tab[0][13] = avg_rv40_qpel16_mc13_c;
-    c->avg_rv40_qpel_pixels_tab[0][14] = avg_rv40_qpel16_mc23_c;
-    c->put_rv40_qpel_pixels_tab[1][ 0] = c->put_h264_qpel_pixels_tab[1][0];
-    c->put_rv40_qpel_pixels_tab[1][ 1] = put_rv40_qpel8_mc10_c;
-    c->put_rv40_qpel_pixels_tab[1][ 2] = put_rv40_qpel8_mc20_c;
-    c->put_rv40_qpel_pixels_tab[1][ 3] = put_rv40_qpel8_mc30_c;
-    c->put_rv40_qpel_pixels_tab[1][ 4] = put_rv40_qpel8_mc01_c;
-    c->put_rv40_qpel_pixels_tab[1][ 5] = put_rv40_qpel8_mc11_c;
-    c->put_rv40_qpel_pixels_tab[1][ 6] = put_rv40_qpel8_mc21_c;
-    c->put_rv40_qpel_pixels_tab[1][ 7] = put_rv40_qpel8_mc31_c;
-    c->put_rv40_qpel_pixels_tab[1][ 8] = put_rv40_qpel8_mc02_c;
-    c->put_rv40_qpel_pixels_tab[1][ 9] = put_rv40_qpel8_mc12_c;
-    c->put_rv40_qpel_pixels_tab[1][10] = put_rv40_qpel8_mc22_c;
-    c->put_rv40_qpel_pixels_tab[1][11] = put_rv40_qpel8_mc32_c;
-    c->put_rv40_qpel_pixels_tab[1][12] = put_rv40_qpel8_mc03_c;
-    c->put_rv40_qpel_pixels_tab[1][13] = put_rv40_qpel8_mc13_c;
-    c->put_rv40_qpel_pixels_tab[1][14] = put_rv40_qpel8_mc23_c;
-    c->avg_rv40_qpel_pixels_tab[1][ 0] = c->avg_h264_qpel_pixels_tab[1][0];
-    c->avg_rv40_qpel_pixels_tab[1][ 1] = avg_rv40_qpel8_mc10_c;
-    c->avg_rv40_qpel_pixels_tab[1][ 2] = avg_rv40_qpel8_mc20_c;
-    c->avg_rv40_qpel_pixels_tab[1][ 3] = avg_rv40_qpel8_mc30_c;
-    c->avg_rv40_qpel_pixels_tab[1][ 4] = avg_rv40_qpel8_mc01_c;
-    c->avg_rv40_qpel_pixels_tab[1][ 5] = avg_rv40_qpel8_mc11_c;
-    c->avg_rv40_qpel_pixels_tab[1][ 6] = avg_rv40_qpel8_mc21_c;
-    c->avg_rv40_qpel_pixels_tab[1][ 7] = avg_rv40_qpel8_mc31_c;
-    c->avg_rv40_qpel_pixels_tab[1][ 8] = avg_rv40_qpel8_mc02_c;
-    c->avg_rv40_qpel_pixels_tab[1][ 9] = avg_rv40_qpel8_mc12_c;
-    c->avg_rv40_qpel_pixels_tab[1][10] = avg_rv40_qpel8_mc22_c;
-    c->avg_rv40_qpel_pixels_tab[1][11] = avg_rv40_qpel8_mc32_c;
-    c->avg_rv40_qpel_pixels_tab[1][12] = avg_rv40_qpel8_mc03_c;
-    c->avg_rv40_qpel_pixels_tab[1][13] = avg_rv40_qpel8_mc13_c;
-    c->avg_rv40_qpel_pixels_tab[1][14] = avg_rv40_qpel8_mc23_c;
+#define RV40_WEIGHT_FUNC(size) \
+static void rv40_weight_func_ ## size (uint8_t *dst, uint8_t *src1, uint8_t *src2, int w1, int w2, int stride)\
+{\
+    int i, j;\
+\
+    for (j = 0; j < size; j++) {\
+        for (i = 0; i < size; i++)\
+            dst[i] = (((w2 * src1[i]) >> 9) + ((w1 * src2[i]) >> 9) + 0x10) >> 5;\
+        src1 += stride;\
+        src2 += stride;\
+        dst  += stride;\
+    }\
+}
+
+RV40_WEIGHT_FUNC(16)
+RV40_WEIGHT_FUNC(8)
+
+av_cold void ff_rv40dsp_init(RV34DSPContext *c, DSPContext* dsp) {
+
+    ff_rv34dsp_init(c, dsp);
+
+    c->put_pixels_tab[0][ 0] = dsp->put_h264_qpel_pixels_tab[0][0];
+    c->put_pixels_tab[0][ 1] = put_rv40_qpel16_mc10_c;
+    c->put_pixels_tab[0][ 2] = dsp->put_h264_qpel_pixels_tab[0][2];
+    c->put_pixels_tab[0][ 3] = put_rv40_qpel16_mc30_c;
+    c->put_pixels_tab[0][ 4] = put_rv40_qpel16_mc01_c;
+    c->put_pixels_tab[0][ 5] = put_rv40_qpel16_mc11_c;
+    c->put_pixels_tab[0][ 6] = put_rv40_qpel16_mc21_c;
+    c->put_pixels_tab[0][ 7] = put_rv40_qpel16_mc31_c;
+    c->put_pixels_tab[0][ 8] = dsp->put_h264_qpel_pixels_tab[0][8];
+    c->put_pixels_tab[0][ 9] = put_rv40_qpel16_mc12_c;
+    c->put_pixels_tab[0][10] = put_rv40_qpel16_mc22_c;
+    c->put_pixels_tab[0][11] = put_rv40_qpel16_mc32_c;
+    c->put_pixels_tab[0][12] = put_rv40_qpel16_mc03_c;
+    c->put_pixels_tab[0][13] = put_rv40_qpel16_mc13_c;
+    c->put_pixels_tab[0][14] = put_rv40_qpel16_mc23_c;
+    c->put_pixels_tab[0][15] = ff_put_rv40_qpel16_mc33_c;
+    c->avg_pixels_tab[0][ 0] = dsp->avg_h264_qpel_pixels_tab[0][0];
+    c->avg_pixels_tab[0][ 1] = avg_rv40_qpel16_mc10_c;
+    c->avg_pixels_tab[0][ 2] = dsp->avg_h264_qpel_pixels_tab[0][2];
+    c->avg_pixels_tab[0][ 3] = avg_rv40_qpel16_mc30_c;
+    c->avg_pixels_tab[0][ 4] = avg_rv40_qpel16_mc01_c;
+    c->avg_pixels_tab[0][ 5] = avg_rv40_qpel16_mc11_c;
+    c->avg_pixels_tab[0][ 6] = avg_rv40_qpel16_mc21_c;
+    c->avg_pixels_tab[0][ 7] = avg_rv40_qpel16_mc31_c;
+    c->avg_pixels_tab[0][ 8] = dsp->avg_h264_qpel_pixels_tab[0][8];
+    c->avg_pixels_tab[0][ 9] = avg_rv40_qpel16_mc12_c;
+    c->avg_pixels_tab[0][10] = avg_rv40_qpel16_mc22_c;
+    c->avg_pixels_tab[0][11] = avg_rv40_qpel16_mc32_c;
+    c->avg_pixels_tab[0][12] = avg_rv40_qpel16_mc03_c;
+    c->avg_pixels_tab[0][13] = avg_rv40_qpel16_mc13_c;
+    c->avg_pixels_tab[0][14] = avg_rv40_qpel16_mc23_c;
+    c->avg_pixels_tab[0][15] = ff_avg_rv40_qpel16_mc33_c;
+    c->put_pixels_tab[1][ 0] = dsp->put_h264_qpel_pixels_tab[1][0];
+    c->put_pixels_tab[1][ 1] = put_rv40_qpel8_mc10_c;
+    c->put_pixels_tab[1][ 2] = dsp->put_h264_qpel_pixels_tab[1][2];
+    c->put_pixels_tab[1][ 3] = put_rv40_qpel8_mc30_c;
+    c->put_pixels_tab[1][ 4] = put_rv40_qpel8_mc01_c;
+    c->put_pixels_tab[1][ 5] = put_rv40_qpel8_mc11_c;
+    c->put_pixels_tab[1][ 6] = put_rv40_qpel8_mc21_c;
+    c->put_pixels_tab[1][ 7] = put_rv40_qpel8_mc31_c;
+    c->put_pixels_tab[1][ 8] = dsp->put_h264_qpel_pixels_tab[1][8];
+    c->put_pixels_tab[1][ 9] = put_rv40_qpel8_mc12_c;
+    c->put_pixels_tab[1][10] = put_rv40_qpel8_mc22_c;
+    c->put_pixels_tab[1][11] = put_rv40_qpel8_mc32_c;
+    c->put_pixels_tab[1][12] = put_rv40_qpel8_mc03_c;
+    c->put_pixels_tab[1][13] = put_rv40_qpel8_mc13_c;
+    c->put_pixels_tab[1][14] = put_rv40_qpel8_mc23_c;
+    c->put_pixels_tab[1][15] = ff_put_rv40_qpel8_mc33_c;
+    c->avg_pixels_tab[1][ 0] = dsp->avg_h264_qpel_pixels_tab[1][0];
+    c->avg_pixels_tab[1][ 1] = avg_rv40_qpel8_mc10_c;
+    c->avg_pixels_tab[1][ 2] = dsp->avg_h264_qpel_pixels_tab[1][2];
+    c->avg_pixels_tab[1][ 3] = avg_rv40_qpel8_mc30_c;
+    c->avg_pixels_tab[1][ 4] = avg_rv40_qpel8_mc01_c;
+    c->avg_pixels_tab[1][ 5] = avg_rv40_qpel8_mc11_c;
+    c->avg_pixels_tab[1][ 6] = avg_rv40_qpel8_mc21_c;
+    c->avg_pixels_tab[1][ 7] = avg_rv40_qpel8_mc31_c;
+    c->avg_pixels_tab[1][ 8] = dsp->avg_h264_qpel_pixels_tab[1][8];
+    c->avg_pixels_tab[1][ 9] = avg_rv40_qpel8_mc12_c;
+    c->avg_pixels_tab[1][10] = avg_rv40_qpel8_mc22_c;
+    c->avg_pixels_tab[1][11] = avg_rv40_qpel8_mc32_c;
+    c->avg_pixels_tab[1][12] = avg_rv40_qpel8_mc03_c;
+    c->avg_pixels_tab[1][13] = avg_rv40_qpel8_mc13_c;
+    c->avg_pixels_tab[1][14] = avg_rv40_qpel8_mc23_c;
+    c->avg_pixels_tab[1][15] = ff_avg_rv40_qpel8_mc33_c;
+
+    c->put_chroma_pixels_tab[0] = put_rv40_chroma_mc8_c;
+    c->put_chroma_pixels_tab[1] = put_rv40_chroma_mc4_c;
+    c->avg_chroma_pixels_tab[0] = avg_rv40_chroma_mc8_c;
+    c->avg_chroma_pixels_tab[1] = avg_rv40_chroma_mc4_c;
+
+    c->rv40_weight_pixels_tab[0] = rv40_weight_func_16;
+    c->rv40_weight_pixels_tab[1] = rv40_weight_func_8;
 
-    c->put_rv40_chroma_pixels_tab[0]= put_rv40_chroma_mc8_c;
-    c->put_rv40_chroma_pixels_tab[1]= put_rv40_chroma_mc4_c;
-    c->avg_rv40_chroma_pixels_tab[0]= avg_rv40_chroma_mc8_c;
-    c->avg_rv40_chroma_pixels_tab[1]= avg_rv40_chroma_mc4_c;
+    if (HAVE_MMX)
+        ff_rv40dsp_init_x86(c, dsp);
 }