h264: merge _internal & template files.

seems git missed them and we temporary lost our improvments in them.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

9 files changed, 209 insertions, 3605 deletions

diff --git a/libavcodec/dsputil_internal.h b/libavcodec/dsputil_internal.h
deleted file mode 100644
index f3856bb294..0000000000
--- a/libavcodec/dsputil_internal.h
+++ /dev/null
@@ -1,1393 +0,0 @@
-/*
- * DSP utils
- * Copyright (c) 2000, 2001 Fabrice Bellard
- * Copyright (c) 2002-2004 Michael Niedermayer <michaelni@gmx.at>
- *
- * gmc & q-pel & 32/64 bit based MC by Michael Niedermayer <michaelni@gmx.at>
- *
- * This file is part of FFmpeg.
- *
- * FFmpeg 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,
- * 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
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
- */
-
-/**
- * @file
- * DSP utils
- */
-
-#include "h264_high_depth.h"
-
-static inline void FUNC(copy_block2)(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride, int h)
-{
-    int i;
-    for(i=0; i<h; i++)
-    {
-        AV_WN2P(dst   , AV_RN2P(src   ));
-        dst+=dstStride;
-        src+=srcStride;
-    }
-}
-
-static inline void FUNC(copy_block4)(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride, int h)
-{
-    int i;
-    for(i=0; i<h; i++)
-    {
-        AV_WN4P(dst   , AV_RN4P(src   ));
-        dst+=dstStride;
-        src+=srcStride;
-    }
-}
-
-static inline void FUNC(copy_block8)(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride, int h)
-{
-    int i;
-    for(i=0; i<h; i++)
-    {
-        AV_WN4P(dst                , AV_RN4P(src                ));
-        AV_WN4P(dst+4*sizeof(pixel), AV_RN4P(src+4*sizeof(pixel)));
-        dst+=dstStride;
-        src+=srcStride;
-    }
-}
-
-static inline void FUNC(copy_block16)(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride, int h)
-{
-    int i;
-    for(i=0; i<h; i++)
-    {
-        AV_WN4P(dst                 , AV_RN4P(src                 ));
-        AV_WN4P(dst+ 4*sizeof(pixel), AV_RN4P(src+ 4*sizeof(pixel)));
-        AV_WN4P(dst+ 8*sizeof(pixel), AV_RN4P(src+ 8*sizeof(pixel)));
-        AV_WN4P(dst+12*sizeof(pixel), AV_RN4P(src+12*sizeof(pixel)));
-        dst+=dstStride;
-        src+=srcStride;
-    }
-}
-
-/* draw the edges of width 'w' of an image of size width, height */
-//FIXME check that this is ok for mpeg4 interlaced
-static void FUNCC(draw_edges)(uint8_t *p_buf, int p_wrap, int width, int height, int w, int sides)
-{
-    pixel *buf = (pixel*)p_buf;
-    int wrap = p_wrap / sizeof(pixel);
-    pixel *ptr, *last_line;
-    int i;
-
-    /* left and right */
-    ptr = buf;
-    for(i=0;i<height;i++) {
-#if BIT_DEPTH > 8
-        int j;
-        for (j = 0; j < w; j++) {
-            ptr[j-w] = ptr[0];
-            ptr[j+width] = ptr[width-1];
-        }
-#else
-        memset(ptr - w, ptr[0], w);
-        memset(ptr + width, ptr[width-1], w);
-#endif
-        ptr += wrap;
-    }
-
-    /* top and bottom + corners */
-    buf -= w;
-    last_line = buf + (height - 1) * wrap;
-    if (sides & EDGE_TOP)
-        for(i = 0; i < w; i++)
-            memcpy(buf - (i + 1) * wrap, buf, (width + w + w) * sizeof(pixel)); // top
-    if (sides & EDGE_BOTTOM)
-        for (i = 0; i < w; i++)
-            memcpy(last_line + (i + 1) * wrap, last_line, (width + w + w) * sizeof(pixel)); // bottom
-}
-
-/**
- * Copy a rectangular area of samples to a temporary buffer and replicate the border samples.
- * @param buf destination buffer
- * @param src source buffer
- * @param linesize number of bytes between 2 vertically adjacent samples in both the source and destination buffers
- * @param block_w width of block
- * @param block_h height of block
- * @param src_x x coordinate of the top left sample of the block in the source buffer
- * @param src_y y coordinate of the top left sample of the block in the source buffer
- * @param w width of the source buffer
- * @param h height of the source buffer
- */
-void FUNC(ff_emulated_edge_mc)(uint8_t *buf, const uint8_t *src, int linesize, int block_w, int block_h,
-                                    int src_x, int src_y, int w, int h){
-    int x, y;
-    int start_y, start_x, end_y, end_x;
-
-    if(src_y>= h){
-        src+= (h-1-src_y)*linesize;
-        src_y=h-1;
-    }else if(src_y<=-block_h){
-        src+= (1-block_h-src_y)*linesize;
-        src_y=1-block_h;
-    }
-    if(src_x>= w){
-        src+= (w-1-src_x)*sizeof(pixel);
-        src_x=w-1;
-    }else if(src_x<=-block_w){
-        src+= (1-block_w-src_x)*sizeof(pixel);
-        src_x=1-block_w;
-    }
-
-    start_y= FFMAX(0, -src_y);
-    start_x= FFMAX(0, -src_x);
-    end_y= FFMIN(block_h, h-src_y);
-    end_x= FFMIN(block_w, w-src_x);
-    assert(start_y < end_y && block_h);
-    assert(start_x < end_x && block_w);
-
-    w    = end_x - start_x;
-    src += start_y*linesize + start_x*sizeof(pixel);
-    buf += start_x*sizeof(pixel);
-
-    //top
-    for(y=0; y<start_y; y++){
-        memcpy(buf, src, w*sizeof(pixel));
-        buf += linesize;
-    }
-
-    // copy existing part
-    for(; y<end_y; y++){
-        memcpy(buf, src, w*sizeof(pixel));
-        src += linesize;
-        buf += linesize;
-    }
-
-    //bottom
-    src -= linesize;
-    for(; y<block_h; y++){
-        memcpy(buf, src, w*sizeof(pixel));
-        buf += linesize;
-    }
-
-    buf -= block_h * linesize + start_x*sizeof(pixel);
-    while (block_h--){
-        pixel *bufp = (pixel*)buf;
-       //left
-        for(x=0; x<start_x; x++){
-            bufp[x] = bufp[start_x];
-        }
-
-       //right
-        for(x=end_x; x<block_w; x++){
-            bufp[x] = bufp[end_x - 1];
-        }
-        buf += linesize;
-    }
-}
-
-static void FUNCC(add_pixels8)(uint8_t *restrict p_pixels, DCTELEM *p_block, int line_size)
-{
-    int i;
-    pixel *restrict pixels = (pixel *restrict)p_pixels;
-    dctcoef *block = (dctcoef*)p_block;
-    line_size >>= sizeof(pixel)-1;
-
-    for(i=0;i<8;i++) {
-        pixels[0] += block[0];
-        pixels[1] += block[1];
-        pixels[2] += block[2];
-        pixels[3] += block[3];
-        pixels[4] += block[4];
-        pixels[5] += block[5];
-        pixels[6] += block[6];
-        pixels[7] += block[7];
-        pixels += line_size;
-        block += 8;
-    }
-}
-
-static void FUNCC(add_pixels4)(uint8_t *restrict p_pixels, DCTELEM *p_block, int line_size)
-{
-    int i;
-    pixel *restrict pixels = (pixel *restrict)p_pixels;
-    dctcoef *block = (dctcoef*)p_block;
-    line_size >>= sizeof(pixel)-1;
-
-    for(i=0;i<4;i++) {
-        pixels[0] += block[0];
-        pixels[1] += block[1];
-        pixels[2] += block[2];
-        pixels[3] += block[3];
-        pixels += line_size;
-        block += 4;
-    }
-}
-
-#if 0
-
-#define PIXOP2(OPNAME, OP) \
-static void OPNAME ## _pixels(uint8_t *block, const uint8_t *pixels, int line_size, int h)\
-{\
-    int i;\
-    for(i=0; i<h; i++){\
-        OP(*((uint64_t*)block), AV_RN64(pixels));\
-        pixels+=line_size;\
-        block +=line_size;\
-    }\
-}\
-\
-static void OPNAME ## _no_rnd_pixels_x2_c(uint8_t *block, const uint8_t *pixels, int line_size, int h)\
-{\
-    int i;\
-    for(i=0; i<h; i++){\
-        const uint64_t a= AV_RN64(pixels  );\
-        const uint64_t b= AV_RN64(pixels+1);\
-        OP(*((uint64_t*)block), (a&b) + (((a^b)&0xFEFEFEFEFEFEFEFEULL)>>1));\
-        pixels+=line_size;\
-        block +=line_size;\
-    }\
-}\
-\
-static void OPNAME ## _pixels_x2_c(uint8_t *block, const uint8_t *pixels, int line_size, int h)\
-{\
-    int i;\
-    for(i=0; i<h; i++){\
-        const uint64_t a= AV_RN64(pixels  );\
-        const uint64_t b= AV_RN64(pixels+1);\
-        OP(*((uint64_t*)block), (a|b) - (((a^b)&0xFEFEFEFEFEFEFEFEULL)>>1));\
-        pixels+=line_size;\
-        block +=line_size;\
-    }\
-}\
-\
-static void OPNAME ## _no_rnd_pixels_y2_c(uint8_t *block, const uint8_t *pixels, int line_size, int h)\
-{\
-    int i;\
-    for(i=0; i<h; i++){\
-        const uint64_t a= AV_RN64(pixels          );\
-        const uint64_t b= AV_RN64(pixels+line_size);\
-        OP(*((uint64_t*)block), (a&b) + (((a^b)&0xFEFEFEFEFEFEFEFEULL)>>1));\
-        pixels+=line_size;\
-        block +=line_size;\
-    }\
-}\
-\
-static void OPNAME ## _pixels_y2_c(uint8_t *block, const uint8_t *pixels, int line_size, int h)\
-{\
-    int i;\
-    for(i=0; i<h; i++){\
-        const uint64_t a= AV_RN64(pixels          );\
-        const uint64_t b= AV_RN64(pixels+line_size);\
-        OP(*((uint64_t*)block), (a|b) - (((a^b)&0xFEFEFEFEFEFEFEFEULL)>>1));\
-        pixels+=line_size;\
-        block +=line_size;\
-    }\
-}\
-\
-static void OPNAME ## _pixels_xy2_c(uint8_t *block, const uint8_t *pixels, int line_size, int h)\
-{\
-        int i;\
-        const uint64_t a= AV_RN64(pixels  );\
-        const uint64_t b= AV_RN64(pixels+1);\
-        uint64_t l0=  (a&0x0303030303030303ULL)\
-                    + (b&0x0303030303030303ULL)\
-                    + 0x0202020202020202ULL;\
-        uint64_t h0= ((a&0xFCFCFCFCFCFCFCFCULL)>>2)\
-                   + ((b&0xFCFCFCFCFCFCFCFCULL)>>2);\
-        uint64_t l1,h1;\
-\
-        pixels+=line_size;\
-        for(i=0; i<h; i+=2){\
-            uint64_t a= AV_RN64(pixels  );\
-            uint64_t b= AV_RN64(pixels+1);\
-            l1=  (a&0x0303030303030303ULL)\
-               + (b&0x0303030303030303ULL);\
-            h1= ((a&0xFCFCFCFCFCFCFCFCULL)>>2)\
-              + ((b&0xFCFCFCFCFCFCFCFCULL)>>2);\
-            OP(*((uint64_t*)block), h0+h1+(((l0+l1)>>2)&0x0F0F0F0F0F0F0F0FULL));\
-            pixels+=line_size;\
-            block +=line_size;\
-            a= AV_RN64(pixels  );\
-            b= AV_RN64(pixels+1);\
-            l0=  (a&0x0303030303030303ULL)\
-               + (b&0x0303030303030303ULL)\
-               + 0x0202020202020202ULL;\
-            h0= ((a&0xFCFCFCFCFCFCFCFCULL)>>2)\
-              + ((b&0xFCFCFCFCFCFCFCFCULL)>>2);\
-            OP(*((uint64_t*)block), h0+h1+(((l0+l1)>>2)&0x0F0F0F0F0F0F0F0FULL));\
-            pixels+=line_size;\
-            block +=line_size;\
-        }\
-}\
-\
-static void OPNAME ## _no_rnd_pixels_xy2_c(uint8_t *block, const uint8_t *pixels, int line_size, int h)\
-{\
-        int i;\
-        const uint64_t a= AV_RN64(pixels  );\
-        const uint64_t b= AV_RN64(pixels+1);\
-        uint64_t l0=  (a&0x0303030303030303ULL)\
-                    + (b&0x0303030303030303ULL)\
-                    + 0x0101010101010101ULL;\
-        uint64_t h0= ((a&0xFCFCFCFCFCFCFCFCULL)>>2)\
-                   + ((b&0xFCFCFCFCFCFCFCFCULL)>>2);\
-        uint64_t l1,h1;\
-\
-        pixels+=line_size;\
-        for(i=0; i<h; i+=2){\
-            uint64_t a= AV_RN64(pixels  );\
-            uint64_t b= AV_RN64(pixels+1);\
-            l1=  (a&0x0303030303030303ULL)\
-               + (b&0x0303030303030303ULL);\
-            h1= ((a&0xFCFCFCFCFCFCFCFCULL)>>2)\
-              + ((b&0xFCFCFCFCFCFCFCFCULL)>>2);\
-            OP(*((uint64_t*)block), h0+h1+(((l0+l1)>>2)&0x0F0F0F0F0F0F0F0FULL));\
-            pixels+=line_size;\
-            block +=line_size;\
-            a= AV_RN64(pixels  );\
-            b= AV_RN64(pixels+1);\
-            l0=  (a&0x0303030303030303ULL)\
-               + (b&0x0303030303030303ULL)\
-               + 0x0101010101010101ULL;\
-            h0= ((a&0xFCFCFCFCFCFCFCFCULL)>>2)\
-              + ((b&0xFCFCFCFCFCFCFCFCULL)>>2);\
-            OP(*((uint64_t*)block), h0+h1+(((l0+l1)>>2)&0x0F0F0F0F0F0F0F0FULL));\
-            pixels+=line_size;\
-            block +=line_size;\
-        }\
-}\
-\
-CALL_2X_PIXELS(OPNAME ## _pixels16_c    , OPNAME ## _pixels_c    , 8*sizeof(pixel))\
-CALL_2X_PIXELS(OPNAME ## _pixels16_x2_c , OPNAME ## _pixels_x2_c , 8*sizeof(pixel))\
-CALL_2X_PIXELS(OPNAME ## _pixels16_y2_c , OPNAME ## _pixels_y2_c , 8*sizeof(pixel))\
-CALL_2X_PIXELS(OPNAME ## _pixels16_xy2_c, OPNAME ## _pixels_xy2_c, 8*sizeof(pixel))\
-CALL_2X_PIXELS(OPNAME ## _no_rnd_pixels16_x2_c , OPNAME ## _no_rnd_pixels_x2_c , 8*sizeof(pixel))\
-CALL_2X_PIXELS(OPNAME ## _no_rnd_pixels16_y2_c , OPNAME ## _no_rnd_pixels_y2_c , 8*sizeof(pixel))\
-CALL_2X_PIXELS(OPNAME ## _no_rnd_pixels16_xy2_c, OPNAME ## _no_rnd_pixels_xy2_c, 8*sizeof(pixel))
-
-#define op_avg(a, b) a = ( ((a)|(b)) - ((((a)^(b))&0xFEFEFEFEFEFEFEFEULL)>>1) )
-#else // 64 bit variant
-
-#define PIXOP2(OPNAME, OP) \
-static void FUNCC(OPNAME ## _pixels2)(uint8_t *block, const uint8_t *pixels, int line_size, int h){\
-    int i;\
-    for(i=0; i<h; i++){\
-        OP(*((pixel2*)(block  )), AV_RN2P(pixels  ));\
-        pixels+=line_size;\
-        block +=line_size;\
-    }\
-}\
-static void FUNCC(OPNAME ## _pixels4)(uint8_t *block, const uint8_t *pixels, int line_size, int h){\
-    int i;\
-    for(i=0; i<h; i++){\
-        OP(*((pixel4*)(block  )), AV_RN4P(pixels  ));\
-        pixels+=line_size;\
-        block +=line_size;\
-    }\
-}\
-static void FUNCC(OPNAME ## _pixels8)(uint8_t *block, const uint8_t *pixels, int line_size, int h){\
-    int i;\
-    for(i=0; i<h; i++){\
-        OP(*((pixel4*)(block                )), AV_RN4P(pixels                ));\
-        OP(*((pixel4*)(block+4*sizeof(pixel))), AV_RN4P(pixels+4*sizeof(pixel)));\
-        pixels+=line_size;\
-        block +=line_size;\
-    }\
-}\
-static inline void FUNCC(OPNAME ## _no_rnd_pixels8)(uint8_t *block, const uint8_t *pixels, int line_size, int h){\
-    FUNCC(OPNAME ## _pixels8)(block, pixels, line_size, h);\
-}\
-\
-static inline void FUNC(OPNAME ## _no_rnd_pixels8_l2)(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, \
-                                                int src_stride1, int src_stride2, int h){\
-    int i;\
-    for(i=0; i<h; i++){\
-        pixel4 a,b;\
-        a= AV_RN4P(&src1[i*src_stride1  ]);\
-        b= AV_RN4P(&src2[i*src_stride2  ]);\
-        OP(*((pixel4*)&dst[i*dst_stride  ]), no_rnd_avg_pixel4(a, b));\
-        a= AV_RN4P(&src1[i*src_stride1+4*sizeof(pixel)]);\
-        b= AV_RN4P(&src2[i*src_stride2+4*sizeof(pixel)]);\
-        OP(*((pixel4*)&dst[i*dst_stride+4*sizeof(pixel)]), no_rnd_avg_pixel4(a, b));\
-    }\
-}\
-\
-static inline void FUNC(OPNAME ## _pixels8_l2)(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, \
-                                                int src_stride1, int src_stride2, int h){\
-    int i;\
-    for(i=0; i<h; i++){\
-        pixel4 a,b;\
-        a= AV_RN4P(&src1[i*src_stride1  ]);\
-        b= AV_RN4P(&src2[i*src_stride2  ]);\
-        OP(*((pixel4*)&dst[i*dst_stride  ]), rnd_avg_pixel4(a, b));\
-        a= AV_RN4P(&src1[i*src_stride1+4*sizeof(pixel)]);\
-        b= AV_RN4P(&src2[i*src_stride2+4*sizeof(pixel)]);\
-        OP(*((pixel4*)&dst[i*dst_stride+4*sizeof(pixel)]), rnd_avg_pixel4(a, b));\
-    }\
-}\
-\
-static inline void FUNC(OPNAME ## _pixels4_l2)(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, \
-                                                int src_stride1, int src_stride2, int h){\
-    int i;\
-    for(i=0; i<h; i++){\
-        pixel4 a,b;\
-        a= AV_RN4P(&src1[i*src_stride1  ]);\
-        b= AV_RN4P(&src2[i*src_stride2  ]);\
-        OP(*((pixel4*)&dst[i*dst_stride  ]), rnd_avg_pixel4(a, b));\
-    }\
-}\
-\
-static inline void FUNC(OPNAME ## _pixels2_l2)(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, \
-                                                int src_stride1, int src_stride2, int h){\
-    int i;\
-    for(i=0; i<h; i++){\
-        pixel4 a,b;\
-        a= AV_RN2P(&src1[i*src_stride1  ]);\
-        b= AV_RN2P(&src2[i*src_stride2  ]);\
-        OP(*((pixel2*)&dst[i*dst_stride  ]), rnd_avg_pixel4(a, b));\
-    }\
-}\
-\
-static inline void FUNC(OPNAME ## _pixels16_l2)(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, \
-                                                int src_stride1, int src_stride2, int h){\
-    FUNC(OPNAME ## _pixels8_l2)(dst  , src1  , src2  , dst_stride, src_stride1, src_stride2, h);\
-    FUNC(OPNAME ## _pixels8_l2)(dst+8*sizeof(pixel), src1+8*sizeof(pixel), src2+8*sizeof(pixel), dst_stride, src_stride1, src_stride2, h);\
-}\
-\
-static inline void FUNC(OPNAME ## _no_rnd_pixels16_l2)(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dst_stride, \
-                                                int src_stride1, int src_stride2, int h){\
-    FUNC(OPNAME ## _no_rnd_pixels8_l2)(dst  , src1  , src2  , dst_stride, src_stride1, src_stride2, h);\
-    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){\
-    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){\
-    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){\
-    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){\
-    FUNC(OPNAME ## _pixels8_l2)(block, pixels, pixels+line_size, line_size, line_size, line_size, h);\
-}\
-\
-static inline void FUNC(OPNAME ## _pixels8_l4)(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, const uint8_t *src3, const uint8_t *src4,\
-                 int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\
-    /* FIXME HIGH BIT DEPTH */\
-    int i;\
-    for(i=0; i<h; i++){\
-        uint32_t a, b, c, d, l0, l1, h0, h1;\
-        a= AV_RN32(&src1[i*src_stride1]);\
-        b= AV_RN32(&src2[i*src_stride2]);\
-        c= AV_RN32(&src3[i*src_stride3]);\
-        d= AV_RN32(&src4[i*src_stride4]);\
-        l0=  (a&0x03030303UL)\
-           + (b&0x03030303UL)\
-           + 0x02020202UL;\
-        h0= ((a&0xFCFCFCFCUL)>>2)\
-          + ((b&0xFCFCFCFCUL)>>2);\
-        l1=  (c&0x03030303UL)\
-           + (d&0x03030303UL);\
-        h1= ((c&0xFCFCFCFCUL)>>2)\
-          + ((d&0xFCFCFCFCUL)>>2);\
-        OP(*((uint32_t*)&dst[i*dst_stride]), h0+h1+(((l0+l1)>>2)&0x0F0F0F0FUL));\
-        a= AV_RN32(&src1[i*src_stride1+4]);\
-        b= AV_RN32(&src2[i*src_stride2+4]);\
-        c= AV_RN32(&src3[i*src_stride3+4]);\
-        d= AV_RN32(&src4[i*src_stride4+4]);\
-        l0=  (a&0x03030303UL)\
-           + (b&0x03030303UL)\
-           + 0x02020202UL;\
-        h0= ((a&0xFCFCFCFCUL)>>2)\
-          + ((b&0xFCFCFCFCUL)>>2);\
-        l1=  (c&0x03030303UL)\
-           + (d&0x03030303UL);\
-        h1= ((c&0xFCFCFCFCUL)>>2)\
-          + ((d&0xFCFCFCFCUL)>>2);\
-        OP(*((uint32_t*)&dst[i*dst_stride+4]), h0+h1+(((l0+l1)>>2)&0x0F0F0F0FUL));\
-    }\
-}\
-\
-static inline void FUNCC(OPNAME ## _pixels4_x2)(uint8_t *block, const uint8_t *pixels, int 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){\
-    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){\
-    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){\
-    FUNC(OPNAME ## _pixels2_l2)(block, pixels, pixels+line_size, line_size, line_size, line_size, h);\
-}\
-\
-static inline void FUNC(OPNAME ## _no_rnd_pixels8_l4)(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, const uint8_t *src3, const uint8_t *src4,\
-                 int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\
-    /* FIXME HIGH BIT DEPTH*/\
-    int i;\
-    for(i=0; i<h; i++){\
-        uint32_t a, b, c, d, l0, l1, h0, h1;\
-        a= AV_RN32(&src1[i*src_stride1]);\
-        b= AV_RN32(&src2[i*src_stride2]);\
-        c= AV_RN32(&src3[i*src_stride3]);\
-        d= AV_RN32(&src4[i*src_stride4]);\
-        l0=  (a&0x03030303UL)\
-           + (b&0x03030303UL)\
-           + 0x01010101UL;\
-        h0= ((a&0xFCFCFCFCUL)>>2)\
-          + ((b&0xFCFCFCFCUL)>>2);\
-        l1=  (c&0x03030303UL)\
-           + (d&0x03030303UL);\
-        h1= ((c&0xFCFCFCFCUL)>>2)\
-          + ((d&0xFCFCFCFCUL)>>2);\
-        OP(*((uint32_t*)&dst[i*dst_stride]), h0+h1+(((l0+l1)>>2)&0x0F0F0F0FUL));\
-        a= AV_RN32(&src1[i*src_stride1+4]);\
-        b= AV_RN32(&src2[i*src_stride2+4]);\
-        c= AV_RN32(&src3[i*src_stride3+4]);\
-        d= AV_RN32(&src4[i*src_stride4+4]);\
-        l0=  (a&0x03030303UL)\
-           + (b&0x03030303UL)\
-           + 0x01010101UL;\
-        h0= ((a&0xFCFCFCFCUL)>>2)\
-          + ((b&0xFCFCFCFCUL)>>2);\
-        l1=  (c&0x03030303UL)\
-           + (d&0x03030303UL);\
-        h1= ((c&0xFCFCFCFCUL)>>2)\
-          + ((d&0xFCFCFCFCUL)>>2);\
-        OP(*((uint32_t*)&dst[i*dst_stride+4]), h0+h1+(((l0+l1)>>2)&0x0F0F0F0FUL));\
-    }\
-}\
-static inline void FUNC(OPNAME ## _pixels16_l4)(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, const uint8_t *src3, const uint8_t *src4,\
-                 int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\
-    FUNC(OPNAME ## _pixels8_l4)(dst  , src1  , src2  , src3  , src4  , dst_stride, src_stride1, src_stride2, src_stride3, src_stride4, h);\
-    FUNC(OPNAME ## _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 FUNC(OPNAME ## _no_rnd_pixels16_l4)(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, const uint8_t *src3, const uint8_t *src4,\
-                 int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\
-    FUNC(OPNAME ## _no_rnd_pixels8_l4)(dst  , src1  , src2  , src3  , src4  , dst_stride, src_stride1, src_stride2, src_stride3, src_stride4, h);\
-    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)\
-{\
-        int i, a0, b0, a1, b1;\
-        pixel *block = (pixel*)p_block;\
-        const pixel *pixels = (const pixel*)p_pixels;\
-        line_size >>= sizeof(pixel)-1;\
-        a0= pixels[0];\
-        b0= pixels[1] + 2;\
-        a0 += b0;\
-        b0 += pixels[2];\
-\
-        pixels+=line_size;\
-        for(i=0; i<h; i+=2){\
-            a1= pixels[0];\
-            b1= pixels[1];\
-            a1 += b1;\
-            b1 += pixels[2];\
-\
-            block[0]= (a1+a0)>>2; /* FIXME non put */\
-            block[1]= (b1+b0)>>2;\
-\
-            pixels+=line_size;\
-            block +=line_size;\
-\
-            a0= pixels[0];\
-            b0= pixels[1] + 2;\
-            a0 += b0;\
-            b0 += pixels[2];\
-\
-            block[0]= (a1+a0)>>2;\
-            block[1]= (b1+b0)>>2;\
-            pixels+=line_size;\
-            block +=line_size;\
-        }\
-}\
-\
-static inline void FUNCC(OPNAME ## _pixels4_xy2)(uint8_t *block, const uint8_t *pixels, int line_size, int h)\
-{\
-        /* FIXME HIGH BIT DEPTH */\
-        int i;\
-        const uint32_t a= AV_RN32(pixels  );\
-        const uint32_t b= AV_RN32(pixels+1);\
-        uint32_t l0=  (a&0x03030303UL)\
-                    + (b&0x03030303UL)\
-                    + 0x02020202UL;\
-        uint32_t h0= ((a&0xFCFCFCFCUL)>>2)\
-                   + ((b&0xFCFCFCFCUL)>>2);\
-        uint32_t l1,h1;\
-\
-        pixels+=line_size;\
-        for(i=0; i<h; i+=2){\
-            uint32_t a= AV_RN32(pixels  );\
-            uint32_t b= AV_RN32(pixels+1);\
-            l1=  (a&0x03030303UL)\
-               + (b&0x03030303UL);\
-            h1= ((a&0xFCFCFCFCUL)>>2)\
-              + ((b&0xFCFCFCFCUL)>>2);\
-            OP(*((uint32_t*)block), h0+h1+(((l0+l1)>>2)&0x0F0F0F0FUL));\
-            pixels+=line_size;\
-            block +=line_size;\
-            a= AV_RN32(pixels  );\
-            b= AV_RN32(pixels+1);\
-            l0=  (a&0x03030303UL)\
-               + (b&0x03030303UL)\
-               + 0x02020202UL;\
-            h0= ((a&0xFCFCFCFCUL)>>2)\
-              + ((b&0xFCFCFCFCUL)>>2);\
-            OP(*((uint32_t*)block), h0+h1+(((l0+l1)>>2)&0x0F0F0F0FUL));\
-            pixels+=line_size;\
-            block +=line_size;\
-        }\
-}\
-\
-static inline void FUNCC(OPNAME ## _pixels8_xy2)(uint8_t *block, const uint8_t *pixels, int line_size, int h)\
-{\
-    /* FIXME HIGH BIT DEPTH */\
-    int j;\
-    for(j=0; j<2; j++){\
-        int i;\
-        const uint32_t a= AV_RN32(pixels  );\
-        const uint32_t b= AV_RN32(pixels+1);\
-        uint32_t l0=  (a&0x03030303UL)\
-                    + (b&0x03030303UL)\
-                    + 0x02020202UL;\
-        uint32_t h0= ((a&0xFCFCFCFCUL)>>2)\
-                   + ((b&0xFCFCFCFCUL)>>2);\
-        uint32_t l1,h1;\
-\
-        pixels+=line_size;\
-        for(i=0; i<h; i+=2){\
-            uint32_t a= AV_RN32(pixels  );\
-            uint32_t b= AV_RN32(pixels+1);\
-            l1=  (a&0x03030303UL)\
-               + (b&0x03030303UL);\
-            h1= ((a&0xFCFCFCFCUL)>>2)\
-              + ((b&0xFCFCFCFCUL)>>2);\
-            OP(*((uint32_t*)block), h0+h1+(((l0+l1)>>2)&0x0F0F0F0FUL));\
-            pixels+=line_size;\
-            block +=line_size;\
-            a= AV_RN32(pixels  );\
-            b= AV_RN32(pixels+1);\
-            l0=  (a&0x03030303UL)\
-               + (b&0x03030303UL)\
-               + 0x02020202UL;\
-            h0= ((a&0xFCFCFCFCUL)>>2)\
-              + ((b&0xFCFCFCFCUL)>>2);\
-            OP(*((uint32_t*)block), h0+h1+(((l0+l1)>>2)&0x0F0F0F0FUL));\
-            pixels+=line_size;\
-            block +=line_size;\
-        }\
-        pixels+=4-line_size*(h+1);\
-        block +=4-line_size*h;\
-    }\
-}\
-\
-static inline void FUNCC(OPNAME ## _no_rnd_pixels8_xy2)(uint8_t *block, const uint8_t *pixels, int line_size, int h)\
-{\
-    /* FIXME HIGH BIT DEPTH */\
-    int j;\
-    for(j=0; j<2; j++){\
-        int i;\
-        const uint32_t a= AV_RN32(pixels  );\
-        const uint32_t b= AV_RN32(pixels+1);\
-        uint32_t l0=  (a&0x03030303UL)\
-                    + (b&0x03030303UL)\
-                    + 0x01010101UL;\
-        uint32_t h0= ((a&0xFCFCFCFCUL)>>2)\
-                   + ((b&0xFCFCFCFCUL)>>2);\
-        uint32_t l1,h1;\
-\
-        pixels+=line_size;\
-        for(i=0; i<h; i+=2){\
-            uint32_t a= AV_RN32(pixels  );\
-            uint32_t b= AV_RN32(pixels+1);\
-            l1=  (a&0x03030303UL)\
-               + (b&0x03030303UL);\
-            h1= ((a&0xFCFCFCFCUL)>>2)\
-              + ((b&0xFCFCFCFCUL)>>2);\
-            OP(*((uint32_t*)block), h0+h1+(((l0+l1)>>2)&0x0F0F0F0FUL));\
-            pixels+=line_size;\
-            block +=line_size;\
-            a= AV_RN32(pixels  );\
-            b= AV_RN32(pixels+1);\
-            l0=  (a&0x03030303UL)\
-               + (b&0x03030303UL)\
-               + 0x01010101UL;\
-            h0= ((a&0xFCFCFCFCUL)>>2)\
-              + ((b&0xFCFCFCFCUL)>>2);\
-            OP(*((uint32_t*)block), h0+h1+(((l0+l1)>>2)&0x0F0F0F0FUL));\
-            pixels+=line_size;\
-            block +=line_size;\
-        }\
-        pixels+=4-line_size*(h+1);\
-        block +=4-line_size*h;\
-    }\
-}\
-\
-CALL_2X_PIXELS(FUNCC(OPNAME ## _pixels16)    , FUNCC(OPNAME ## _pixels8)    , 8*sizeof(pixel))\
-CALL_2X_PIXELS(FUNCC(OPNAME ## _pixels16_x2) , FUNCC(OPNAME ## _pixels8_x2) , 8*sizeof(pixel))\
-CALL_2X_PIXELS(FUNCC(OPNAME ## _pixels16_y2) , FUNCC(OPNAME ## _pixels8_y2) , 8*sizeof(pixel))\
-CALL_2X_PIXELS(FUNCC(OPNAME ## _pixels16_xy2), FUNCC(OPNAME ## _pixels8_xy2), 8*sizeof(pixel))\
-av_unused CALL_2X_PIXELS(FUNCC(OPNAME ## _no_rnd_pixels16)    , FUNCC(OPNAME ## _pixels8) , 8*sizeof(pixel))\
-CALL_2X_PIXELS(FUNCC(OPNAME ## _no_rnd_pixels16_x2) , FUNCC(OPNAME ## _no_rnd_pixels8_x2) , 8*sizeof(pixel))\
-CALL_2X_PIXELS(FUNCC(OPNAME ## _no_rnd_pixels16_y2) , FUNCC(OPNAME ## _no_rnd_pixels8_y2) , 8*sizeof(pixel))\
-CALL_2X_PIXELS(FUNCC(OPNAME ## _no_rnd_pixels16_xy2), FUNCC(OPNAME ## _no_rnd_pixels8_xy2), 8*sizeof(pixel))\
-
-#define op_avg(a, b) a = rnd_avg_pixel4(a, b)
-#endif
-#define op_put(a, b) a = b
-
-PIXOP2(avg, op_avg)
-PIXOP2(put, op_put)
-#undef op_avg
-#undef op_put
-
-#define put_no_rnd_pixels8_c  put_pixels8_c
-#define put_no_rnd_pixels16_c put_pixels16_c
-
-static void FUNCC(put_no_rnd_pixels16_l2)(uint8_t *dst, const uint8_t *a, const uint8_t *b, int stride, int h){
-    FUNC(put_no_rnd_pixels16_l2)(dst, a, b, stride, stride, stride, h);
-}
-
-static void FUNCC(put_no_rnd_pixels8_l2)(uint8_t *dst, const uint8_t *a, const uint8_t *b, int stride, int h){
-    FUNC(put_no_rnd_pixels8_l2)(dst, a, b, stride, stride, stride, h);
-}
-
-#define H264_CHROMA_MC(OPNAME, OP)\
-static void FUNCC(OPNAME ## h264_chroma_mc2)(uint8_t *p_dst/*align 8*/, uint8_t *p_src/*align 1*/, int stride, int h, int x, int y){\
-    pixel *dst = (pixel*)p_dst;\
-    pixel *src = (pixel*)p_src;\
-    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;\
-    stride >>= sizeof(pixel)-1;\
-    \
-    assert(x<8 && y<8 && x>=0 && y>=0);\
-\
-    if(D){\
-        for(i=0; i<h; i++){\
-            OP(dst[0], (A*src[0] + B*src[1] + C*src[stride+0] + D*src[stride+1]));\
-            OP(dst[1], (A*src[1] + B*src[2] + C*src[stride+1] + D*src[stride+2]));\
-            dst+= stride;\
-            src+= stride;\
-        }\
-    }else{\
-        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]));\
-            OP(dst[1], (A*src[1] + E*src[step+1]));\
-            dst+= stride;\
-            src+= stride;\
-        }\
-    }\
-}\
-\
-static void FUNCC(OPNAME ## h264_chroma_mc4)(uint8_t *p_dst/*align 8*/, uint8_t *p_src/*align 1*/, int stride, int h, int x, int y){\
-    pixel *dst = (pixel*)p_dst;\
-    pixel *src = (pixel*)p_src;\
-    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;\
-    stride >>= sizeof(pixel)-1;\
-    \
-    assert(x<8 && y<8 && x>=0 && y>=0);\
-\
-    if(D){\
-        for(i=0; i<h; i++){\
-            OP(dst[0], (A*src[0] + B*src[1] + C*src[stride+0] + D*src[stride+1]));\
-            OP(dst[1], (A*src[1] + B*src[2] + C*src[stride+1] + D*src[stride+2]));\
-            OP(dst[2], (A*src[2] + B*src[3] + C*src[stride+2] + D*src[stride+3]));\
-            OP(dst[3], (A*src[3] + B*src[4] + C*src[stride+3] + D*src[stride+4]));\
-            dst+= stride;\
-            src+= stride;\
-        }\
-    }else{\
-        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]));\
-            OP(dst[1], (A*src[1] + E*src[step+1]));\
-            OP(dst[2], (A*src[2] + E*src[step+2]));\
-            OP(dst[3], (A*src[3] + E*src[step+3]));\
-            dst+= stride;\
-            src+= stride;\
-        }\
-    }\
-}\
-\
-static void FUNCC(OPNAME ## h264_chroma_mc8)(uint8_t *p_dst/*align 8*/, uint8_t *p_src/*align 1*/, int stride, int h, int x, int y){\
-    pixel *dst = (pixel*)p_dst;\
-    pixel *src = (pixel*)p_src;\
-    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;\
-    stride >>= sizeof(pixel)-1;\
-    \
-    assert(x<8 && y<8 && x>=0 && y>=0);\
-\
-    if(D){\
-        for(i=0; i<h; i++){\
-            OP(dst[0], (A*src[0] + B*src[1] + C*src[stride+0] + D*src[stride+1]));\
-            OP(dst[1], (A*src[1] + B*src[2] + C*src[stride+1] + D*src[stride+2]));\
-            OP(dst[2], (A*src[2] + B*src[3] + C*src[stride+2] + D*src[stride+3]));\
-            OP(dst[3], (A*src[3] + B*src[4] + C*src[stride+3] + D*src[stride+4]));\
-            OP(dst[4], (A*src[4] + B*src[5] + C*src[stride+4] + D*src[stride+5]));\
-            OP(dst[5], (A*src[5] + B*src[6] + C*src[stride+5] + D*src[stride+6]));\
-            OP(dst[6], (A*src[6] + B*src[7] + C*src[stride+6] + D*src[stride+7]));\
-            OP(dst[7], (A*src[7] + B*src[8] + C*src[stride+7] + D*src[stride+8]));\
-            dst+= stride;\
-            src+= stride;\
-        }\
-    }else{\
-        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]));\
-            OP(dst[1], (A*src[1] + E*src[step+1]));\
-            OP(dst[2], (A*src[2] + E*src[step+2]));\
-            OP(dst[3], (A*src[3] + E*src[step+3]));\
-            OP(dst[4], (A*src[4] + E*src[step+4]));\
-            OP(dst[5], (A*src[5] + E*src[step+5]));\
-            OP(dst[6], (A*src[6] + E*src[step+6]));\
-            OP(dst[7], (A*src[7] + E*src[step+7]));\
-            dst+= stride;\
-            src+= stride;\
-        }\
-    }\
-}
-
-#define op_avg(a, b) a = (((a)+(((b) + 32)>>6)+1)>>1)
-#define op_put(a, b) a = (((b) + 32)>>6)
-
-H264_CHROMA_MC(put_       , op_put)
-H264_CHROMA_MC(avg_       , op_avg)
-#undef op_avg
-#undef op_put
-
-#if 1
-#define H264_LOWPASS(OPNAME, OP, OP2) \
-static av_unused void FUNC(OPNAME ## h264_qpel2_h_lowpass)(uint8_t *p_dst, uint8_t *p_src, int dstStride, int srcStride){\
-    const int h=2;\
-    INIT_CLIP\
-    int i;\
-    pixel *dst = (pixel*)p_dst;\
-    pixel *src = (pixel*)p_src;\
-    dstStride >>= sizeof(pixel)-1;\
-    srcStride >>= sizeof(pixel)-1;\
-    for(i=0; i<h; i++)\
-    {\
-        OP(dst[0], (src[0]+src[1])*20 - (src[-1]+src[2])*5 + (src[-2]+src[3]));\
-        OP(dst[1], (src[1]+src[2])*20 - (src[0 ]+src[3])*5 + (src[-1]+src[4]));\
-        dst+=dstStride;\
-        src+=srcStride;\
-    }\
-}\
-\
-static av_unused void FUNC(OPNAME ## h264_qpel2_v_lowpass)(uint8_t *p_dst, uint8_t *p_src, int dstStride, int srcStride){\
-    const int w=2;\
-    INIT_CLIP\
-    int i;\
-    pixel *dst = (pixel*)p_dst;\
-    pixel *src = (pixel*)p_src;\
-    dstStride >>= sizeof(pixel)-1;\
-    srcStride >>= sizeof(pixel)-1;\
-    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];\
-        OP(dst[0*dstStride], (src0+src1)*20 - (srcA+src2)*5 + (srcB+src3));\
-        OP(dst[1*dstStride], (src1+src2)*20 - (src0+src3)*5 + (srcA+src4));\
-        dst++;\
-        src++;\
-    }\
-}\
-\
-static av_unused void FUNC(OPNAME ## h264_qpel2_hv_lowpass)(uint8_t *p_dst, int16_t *tmp, uint8_t *p_src, int dstStride, int tmpStride, int srcStride){\
-    const int h=2;\
-    const int w=2;\
-    const int pad = (BIT_DEPTH > 9) ? (-10 * ((1<<BIT_DEPTH)-1)) : 0;\
-    INIT_CLIP\
-    int i;\
-    pixel *dst = (pixel*)p_dst;\
-    pixel *src = (pixel*)p_src;\
-    dstStride >>= sizeof(pixel)-1;\
-    srcStride >>= sizeof(pixel)-1;\
-    src -= 2*srcStride;\
-    for(i=0; i<h+5; i++)\
-    {\
-        tmp[0]= (src[0]+src[1])*20 - (src[-1]+src[2])*5 + (src[-2]+src[3]) + pad;\
-        tmp[1]= (src[1]+src[2])*20 - (src[0 ]+src[3])*5 + (src[-1]+src[4]) + pad;\
-        tmp+=tmpStride;\
-        src+=srcStride;\
-    }\
-    tmp -= tmpStride*(h+5-2);\
-    for(i=0; i<w; i++)\
-    {\
-        const int tmpB= tmp[-2*tmpStride] - pad;\
-        const int tmpA= tmp[-1*tmpStride] - pad;\
-        const int tmp0= tmp[0 *tmpStride] - pad;\
-        const int tmp1= tmp[1 *tmpStride] - pad;\
-        const int tmp2= tmp[2 *tmpStride] - pad;\
-        const int tmp3= tmp[3 *tmpStride] - pad;\
-        const int tmp4= tmp[4 *tmpStride] - pad;\
-        OP2(dst[0*dstStride], (tmp0+tmp1)*20 - (tmpA+tmp2)*5 + (tmpB+tmp3));\
-        OP2(dst[1*dstStride], (tmp1+tmp2)*20 - (tmp0+tmp3)*5 + (tmpA+tmp4));\
-        dst++;\
-        tmp++;\
-    }\
-}\
-static void FUNC(OPNAME ## h264_qpel4_h_lowpass)(uint8_t *p_dst, uint8_t *p_src, int dstStride, int srcStride){\
-    const int h=4;\
-    INIT_CLIP\
-    int i;\
-    pixel *dst = (pixel*)p_dst;\
-    pixel *src = (pixel*)p_src;\
-    dstStride >>= sizeof(pixel)-1;\
-    srcStride >>= sizeof(pixel)-1;\
-    for(i=0; i<h; i++)\
-    {\
-        OP(dst[0], (src[0]+src[1])*20 - (src[-1]+src[2])*5 + (src[-2]+src[3]));\
-        OP(dst[1], (src[1]+src[2])*20 - (src[0 ]+src[3])*5 + (src[-1]+src[4]));\
-        OP(dst[2], (src[2]+src[3])*20 - (src[1 ]+src[4])*5 + (src[0 ]+src[5]));\
-        OP(dst[3], (src[3]+src[4])*20 - (src[2 ]+src[5])*5 + (src[1 ]+src[6]));\
-        dst+=dstStride;\
-        src+=srcStride;\
-    }\
-}\
-\
-static void FUNC(OPNAME ## h264_qpel4_v_lowpass)(uint8_t *p_dst, uint8_t *p_src, int dstStride, int srcStride){\
-    const int w=4;\
-    INIT_CLIP\
-    int i;\
-    pixel *dst = (pixel*)p_dst;\
-    pixel *src = (pixel*)p_src;\
-    dstStride >>= sizeof(pixel)-1;\
-    srcStride >>= sizeof(pixel)-1;\
-    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];\
-        OP(dst[0*dstStride], (src0+src1)*20 - (srcA+src2)*5 + (srcB+src3));\
-        OP(dst[1*dstStride], (src1+src2)*20 - (src0+src3)*5 + (srcA+src4));\
-        OP(dst[2*dstStride], (src2+src3)*20 - (src1+src4)*5 + (src0+src5));\
-        OP(dst[3*dstStride], (src3+src4)*20 - (src2+src5)*5 + (src1+src6));\
-        dst++;\
-        src++;\
-    }\
-}\
-\
-static void FUNC(OPNAME ## h264_qpel4_hv_lowpass)(uint8_t *p_dst, int16_t *tmp, uint8_t *p_src, int dstStride, int tmpStride, int srcStride){\
-    const int h=4;\
-    const int w=4;\
-    const int pad = (BIT_DEPTH > 9) ? (-10 * ((1<<BIT_DEPTH)-1)) : 0;\
-    INIT_CLIP\
-    int i;\
-    pixel *dst = (pixel*)p_dst;\
-    pixel *src = (pixel*)p_src;\
-    dstStride >>= sizeof(pixel)-1;\
-    srcStride >>= sizeof(pixel)-1;\
-    src -= 2*srcStride;\
-    for(i=0; i<h+5; i++)\
-    {\
-        tmp[0]= (src[0]+src[1])*20 - (src[-1]+src[2])*5 + (src[-2]+src[3]) + pad;\
-        tmp[1]= (src[1]+src[2])*20 - (src[0 ]+src[3])*5 + (src[-1]+src[4]) + pad;\
-        tmp[2]= (src[2]+src[3])*20 - (src[1 ]+src[4])*5 + (src[0 ]+src[5]) + pad;\
-        tmp[3]= (src[3]+src[4])*20 - (src[2 ]+src[5])*5 + (src[1 ]+src[6]) + pad;\
-        tmp+=tmpStride;\
-        src+=srcStride;\
-    }\
-    tmp -= tmpStride*(h+5-2);\
-    for(i=0; i<w; i++)\
-    {\
-        const int tmpB= tmp[-2*tmpStride] - pad;\
-        const int tmpA= tmp[-1*tmpStride] - pad;\
-        const int tmp0= tmp[0 *tmpStride] - pad;\
-        const int tmp1= tmp[1 *tmpStride] - pad;\
-        const int tmp2= tmp[2 *tmpStride] - pad;\
-        const int tmp3= tmp[3 *tmpStride] - pad;\
-        const int tmp4= tmp[4 *tmpStride] - pad;\
-        const int tmp5= tmp[5 *tmpStride] - pad;\
-        const int tmp6= tmp[6 *tmpStride] - pad;\
-        OP2(dst[0*dstStride], (tmp0+tmp1)*20 - (tmpA+tmp2)*5 + (tmpB+tmp3));\
-        OP2(dst[1*dstStride], (tmp1+tmp2)*20 - (tmp0+tmp3)*5 + (tmpA+tmp4));\
-        OP2(dst[2*dstStride], (tmp2+tmp3)*20 - (tmp1+tmp4)*5 + (tmp0+tmp5));\
-        OP2(dst[3*dstStride], (tmp3+tmp4)*20 - (tmp2+tmp5)*5 + (tmp1+tmp6));\
-        dst++;\
-        tmp++;\
-    }\
-}\
-\
-static void FUNC(OPNAME ## h264_qpel8_h_lowpass)(uint8_t *p_dst, uint8_t *p_src, int dstStride, int srcStride){\
-    const int h=8;\
-    INIT_CLIP\
-    int i;\
-    pixel *dst = (pixel*)p_dst;\
-    pixel *src = (pixel*)p_src;\
-    dstStride >>= sizeof(pixel)-1;\
-    srcStride >>= sizeof(pixel)-1;\
-    for(i=0; i<h; i++)\
-    {\
-        OP(dst[0], (src[0]+src[1])*20 - (src[-1]+src[2])*5 + (src[-2]+src[3 ]));\
-        OP(dst[1], (src[1]+src[2])*20 - (src[0 ]+src[3])*5 + (src[-1]+src[4 ]));\
-        OP(dst[2], (src[2]+src[3])*20 - (src[1 ]+src[4])*5 + (src[0 ]+src[5 ]));\
-        OP(dst[3], (src[3]+src[4])*20 - (src[2 ]+src[5])*5 + (src[1 ]+src[6 ]));\
-        OP(dst[4], (src[4]+src[5])*20 - (src[3 ]+src[6])*5 + (src[2 ]+src[7 ]));\
-        OP(dst[5], (src[5]+src[6])*20 - (src[4 ]+src[7])*5 + (src[3 ]+src[8 ]));\
-        OP(dst[6], (src[6]+src[7])*20 - (src[5 ]+src[8])*5 + (src[4 ]+src[9 ]));\
-        OP(dst[7], (src[7]+src[8])*20 - (src[6 ]+src[9])*5 + (src[5 ]+src[10]));\
-        dst+=dstStride;\
-        src+=srcStride;\
-    }\
-}\
-\
-static void FUNC(OPNAME ## h264_qpel8_v_lowpass)(uint8_t *p_dst, uint8_t *p_src, int dstStride, int srcStride){\
-    const int w=8;\
-    INIT_CLIP\
-    int i;\
-    pixel *dst = (pixel*)p_dst;\
-    pixel *src = (pixel*)p_src;\
-    dstStride >>= sizeof(pixel)-1;\
-    srcStride >>= sizeof(pixel)-1;\
-    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];\
-        OP(dst[0*dstStride], (src0+src1)*20 - (srcA+src2)*5 + (srcB+src3));\
-        OP(dst[1*dstStride], (src1+src2)*20 - (src0+src3)*5 + (srcA+src4));\
-        OP(dst[2*dstStride], (src2+src3)*20 - (src1+src4)*5 + (src0+src5));\
-        OP(dst[3*dstStride], (src3+src4)*20 - (src2+src5)*5 + (src1+src6));\
-        OP(dst[4*dstStride], (src4+src5)*20 - (src3+src6)*5 + (src2+src7));\
-        OP(dst[5*dstStride], (src5+src6)*20 - (src4+src7)*5 + (src3+src8));\
-        OP(dst[6*dstStride], (src6+src7)*20 - (src5+src8)*5 + (src4+src9));\
-        OP(dst[7*dstStride], (src7+src8)*20 - (src6+src9)*5 + (src5+src10));\
-        dst++;\
-        src++;\
-    }\
-}\
-\
-static void FUNC(OPNAME ## h264_qpel8_hv_lowpass)(uint8_t *p_dst, int16_t *tmp, uint8_t *p_src, int dstStride, int tmpStride, int srcStride){\
-    const int h=8;\
-    const int w=8;\
-    const int pad = (BIT_DEPTH > 9) ? (-10 * ((1<<BIT_DEPTH)-1)) : 0;\
-    INIT_CLIP\
-    int i;\
-    pixel *dst = (pixel*)p_dst;\
-    pixel *src = (pixel*)p_src;\
-    dstStride >>= sizeof(pixel)-1;\
-    srcStride >>= sizeof(pixel)-1;\
-    src -= 2*srcStride;\
-    for(i=0; i<h+5; i++)\
-    {\
-        tmp[0]= (src[0]+src[1])*20 - (src[-1]+src[2])*5 + (src[-2]+src[3 ]) + pad;\
-        tmp[1]= (src[1]+src[2])*20 - (src[0 ]+src[3])*5 + (src[-1]+src[4 ]) + pad;\
-        tmp[2]= (src[2]+src[3])*20 - (src[1 ]+src[4])*5 + (src[0 ]+src[5 ]) + pad;\
-        tmp[3]= (src[3]+src[4])*20 - (src[2 ]+src[5])*5 + (src[1 ]+src[6 ]) + pad;\
-        tmp[4]= (src[4]+src[5])*20 - (src[3 ]+src[6])*5 + (src[2 ]+src[7 ]) + pad;\
-        tmp[5]= (src[5]+src[6])*20 - (src[4 ]+src[7])*5 + (src[3 ]+src[8 ]) + pad;\
-        tmp[6]= (src[6]+src[7])*20 - (src[5 ]+src[8])*5 + (src[4 ]+src[9 ]) + pad;\
-        tmp[7]= (src[7]+src[8])*20 - (src[6 ]+src[9])*5 + (src[5 ]+src[10]) + pad;\
-        tmp+=tmpStride;\
-        src+=srcStride;\
-    }\
-    tmp -= tmpStride*(h+5-2);\
-    for(i=0; i<w; i++)\
-    {\
-        const int tmpB= tmp[-2*tmpStride] - pad;\
-        const int tmpA= tmp[-1*tmpStride] - pad;\
-        const int tmp0= tmp[0 *tmpStride] - pad;\
-        const int tmp1= tmp[1 *tmpStride] - pad;\
-        const int tmp2= tmp[2 *tmpStride] - pad;\
-        const int tmp3= tmp[3 *tmpStride] - pad;\
-        const int tmp4= tmp[4 *tmpStride] - pad;\
-        const int tmp5= tmp[5 *tmpStride] - pad;\
-        const int tmp6= tmp[6 *tmpStride] - pad;\
-        const int tmp7= tmp[7 *tmpStride] - pad;\
-        const int tmp8= tmp[8 *tmpStride] - pad;\
-        const int tmp9= tmp[9 *tmpStride] - pad;\
-        const int tmp10=tmp[10*tmpStride] - pad;\
-        OP2(dst[0*dstStride], (tmp0+tmp1)*20 - (tmpA+tmp2)*5 + (tmpB+tmp3));\
-        OP2(dst[1*dstStride], (tmp1+tmp2)*20 - (tmp0+tmp3)*5 + (tmpA+tmp4));\
-        OP2(dst[2*dstStride], (tmp2+tmp3)*20 - (tmp1+tmp4)*5 + (tmp0+tmp5));\
-        OP2(dst[3*dstStride], (tmp3+tmp4)*20 - (tmp2+tmp5)*5 + (tmp1+tmp6));\
-        OP2(dst[4*dstStride], (tmp4+tmp5)*20 - (tmp3+tmp6)*5 + (tmp2+tmp7));\
-        OP2(dst[5*dstStride], (tmp5+tmp6)*20 - (tmp4+tmp7)*5 + (tmp3+tmp8));\
-        OP2(dst[6*dstStride], (tmp6+tmp7)*20 - (tmp5+tmp8)*5 + (tmp4+tmp9));\
-        OP2(dst[7*dstStride], (tmp7+tmp8)*20 - (tmp6+tmp9)*5 + (tmp5+tmp10));\
-        dst++;\
-        tmp++;\
-    }\
-}\
-\
-static void FUNC(OPNAME ## h264_qpel16_v_lowpass)(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
-    FUNC(OPNAME ## h264_qpel8_v_lowpass)(dst                , src                , dstStride, srcStride);\
-    FUNC(OPNAME ## h264_qpel8_v_lowpass)(dst+8*sizeof(pixel), src+8*sizeof(pixel), dstStride, srcStride);\
-    src += 8*srcStride;\
-    dst += 8*dstStride;\
-    FUNC(OPNAME ## h264_qpel8_v_lowpass)(dst                , src                , dstStride, srcStride);\
-    FUNC(OPNAME ## h264_qpel8_v_lowpass)(dst+8*sizeof(pixel), src+8*sizeof(pixel), dstStride, srcStride);\
-}\
-\
-static void FUNC(OPNAME ## h264_qpel16_h_lowpass)(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
-    FUNC(OPNAME ## h264_qpel8_h_lowpass)(dst                , src                , dstStride, srcStride);\
-    FUNC(OPNAME ## h264_qpel8_h_lowpass)(dst+8*sizeof(pixel), src+8*sizeof(pixel), dstStride, srcStride);\
-    src += 8*srcStride;\
-    dst += 8*dstStride;\
-    FUNC(OPNAME ## h264_qpel8_h_lowpass)(dst                , src                , dstStride, srcStride);\
-    FUNC(OPNAME ## h264_qpel8_h_lowpass)(dst+8*sizeof(pixel), src+8*sizeof(pixel), dstStride, srcStride);\
-}\
-\
-static void FUNC(OPNAME ## h264_qpel16_hv_lowpass)(uint8_t *dst, int16_t *tmp, uint8_t *src, int dstStride, int tmpStride, int srcStride){\
-    FUNC(OPNAME ## h264_qpel8_hv_lowpass)(dst                , tmp  , src                , dstStride, tmpStride, srcStride);\
-    FUNC(OPNAME ## h264_qpel8_hv_lowpass)(dst+8*sizeof(pixel), tmp+8, src+8*sizeof(pixel), dstStride, tmpStride, srcStride);\
-    src += 8*srcStride;\
-    dst += 8*dstStride;\
-    FUNC(OPNAME ## h264_qpel8_hv_lowpass)(dst                , tmp  , src                , dstStride, tmpStride, srcStride);\
-    FUNC(OPNAME ## h264_qpel8_hv_lowpass)(dst+8*sizeof(pixel), tmp+8, src+8*sizeof(pixel), dstStride, tmpStride, srcStride);\
-}\
-
-#define H264_MC(OPNAME, SIZE) \
-static av_unused void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc00)(uint8_t *dst, uint8_t *src, int stride){\
-    FUNCC(OPNAME ## pixels ## SIZE)(dst, src, stride, SIZE);\
-}\
-\
-static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc10)(uint8_t *dst, uint8_t *src, int stride){\
-    uint8_t half[SIZE*SIZE*sizeof(pixel)];\
-    FUNC(put_h264_qpel ## SIZE ## _h_lowpass)(half, src, SIZE*sizeof(pixel), stride);\
-    FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, src, half, stride, stride, SIZE*sizeof(pixel), SIZE);\
-}\
-\
-static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc20)(uint8_t *dst, uint8_t *src, int stride){\
-    FUNC(OPNAME ## h264_qpel ## SIZE ## _h_lowpass)(dst, src, stride, stride);\
-}\
-\
-static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc30)(uint8_t *dst, uint8_t *src, int stride){\
-    uint8_t half[SIZE*SIZE*sizeof(pixel)];\
-    FUNC(put_h264_qpel ## SIZE ## _h_lowpass)(half, src, SIZE*sizeof(pixel), stride);\
-    FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, src+sizeof(pixel), half, stride, stride, SIZE*sizeof(pixel), SIZE);\
-}\
-\
-static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc01)(uint8_t *dst, uint8_t *src, int stride){\
-    uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\
-    uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\
-    uint8_t half[SIZE*SIZE*sizeof(pixel)];\
-    FUNC(copy_block ## SIZE )(full, src - stride*2, SIZE*sizeof(pixel),  stride, SIZE + 5);\
-    FUNC(put_h264_qpel ## SIZE ## _v_lowpass)(half, full_mid, SIZE*sizeof(pixel), SIZE*sizeof(pixel));\
-    FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, full_mid, half, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\
-}\
-\
-static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc02)(uint8_t *dst, uint8_t *src, int stride){\
-    uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\
-    uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\
-    FUNC(copy_block ## SIZE )(full, src - stride*2, SIZE*sizeof(pixel),  stride, SIZE + 5);\
-    FUNC(OPNAME ## h264_qpel ## SIZE ## _v_lowpass)(dst, full_mid, stride, SIZE*sizeof(pixel));\
-}\
-\
-static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc03)(uint8_t *dst, uint8_t *src, int stride){\
-    uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\
-    uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\
-    uint8_t half[SIZE*SIZE*sizeof(pixel)];\
-    FUNC(copy_block ## SIZE )(full, src - stride*2, SIZE*sizeof(pixel),  stride, SIZE + 5);\
-    FUNC(put_h264_qpel ## SIZE ## _v_lowpass)(half, full_mid, SIZE*sizeof(pixel), SIZE*sizeof(pixel));\
-    FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, full_mid+SIZE*sizeof(pixel), half, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\
-}\
-\
-static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc11)(uint8_t *dst, uint8_t *src, int stride){\
-    uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\
-    uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\
-    uint8_t halfH[SIZE*SIZE*sizeof(pixel)];\
-    uint8_t halfV[SIZE*SIZE*sizeof(pixel)];\
-    FUNC(put_h264_qpel ## SIZE ## _h_lowpass)(halfH, src, SIZE*sizeof(pixel), stride);\
-    FUNC(copy_block ## SIZE )(full, src - stride*2, SIZE*sizeof(pixel),  stride, SIZE + 5);\
-    FUNC(put_h264_qpel ## SIZE ## _v_lowpass)(halfV, full_mid, SIZE*sizeof(pixel), SIZE*sizeof(pixel));\
-    FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfH, halfV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\
-}\
-\
-static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc31)(uint8_t *dst, uint8_t *src, int stride){\
-    uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\
-    uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\
-    uint8_t halfH[SIZE*SIZE*sizeof(pixel)];\
-    uint8_t halfV[SIZE*SIZE*sizeof(pixel)];\
-    FUNC(put_h264_qpel ## SIZE ## _h_lowpass)(halfH, src, SIZE*sizeof(pixel), stride);\
-    FUNC(copy_block ## SIZE )(full, src - stride*2 + sizeof(pixel), SIZE*sizeof(pixel),  stride, SIZE + 5);\
-    FUNC(put_h264_qpel ## SIZE ## _v_lowpass)(halfV, full_mid, SIZE*sizeof(pixel), SIZE*sizeof(pixel));\
-    FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfH, halfV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\
-}\
-\
-static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc13)(uint8_t *dst, uint8_t *src, int stride){\
-    uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\
-    uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\
-    uint8_t halfH[SIZE*SIZE*sizeof(pixel)];\
-    uint8_t halfV[SIZE*SIZE*sizeof(pixel)];\
-    FUNC(put_h264_qpel ## SIZE ## _h_lowpass)(halfH, src + stride, SIZE*sizeof(pixel), stride);\
-    FUNC(copy_block ## SIZE )(full, src - stride*2, SIZE*sizeof(pixel),  stride, SIZE + 5);\
-    FUNC(put_h264_qpel ## SIZE ## _v_lowpass)(halfV, full_mid, SIZE*sizeof(pixel), SIZE*sizeof(pixel));\
-    FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfH, halfV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\
-}\
-\
-static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc33)(uint8_t *dst, uint8_t *src, int stride){\
-    uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\
-    uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\
-    uint8_t halfH[SIZE*SIZE*sizeof(pixel)];\
-    uint8_t halfV[SIZE*SIZE*sizeof(pixel)];\
-    FUNC(put_h264_qpel ## SIZE ## _h_lowpass)(halfH, src + stride, SIZE*sizeof(pixel), stride);\
-    FUNC(copy_block ## SIZE )(full, src - stride*2 + sizeof(pixel), SIZE*sizeof(pixel),  stride, SIZE + 5);\
-    FUNC(put_h264_qpel ## SIZE ## _v_lowpass)(halfV, full_mid, SIZE*sizeof(pixel), SIZE*sizeof(pixel));\
-    FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfH, halfV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\
-}\
-\
-static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc22)(uint8_t *dst, uint8_t *src, int stride){\
-    int16_t tmp[SIZE*(SIZE+5)*sizeof(pixel)];\
-    FUNC(OPNAME ## h264_qpel ## SIZE ## _hv_lowpass)(dst, tmp, src, stride, SIZE*sizeof(pixel), stride);\
-}\
-\
-static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc21)(uint8_t *dst, uint8_t *src, int stride){\
-    int16_t tmp[SIZE*(SIZE+5)*sizeof(pixel)];\
-    uint8_t halfH[SIZE*SIZE*sizeof(pixel)];\
-    uint8_t halfHV[SIZE*SIZE*sizeof(pixel)];\
-    FUNC(put_h264_qpel ## SIZE ## _h_lowpass)(halfH, src, SIZE*sizeof(pixel), stride);\
-    FUNC(put_h264_qpel ## SIZE ## _hv_lowpass)(halfHV, tmp, src, SIZE*sizeof(pixel), SIZE*sizeof(pixel), stride);\
-    FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfH, halfHV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\
-}\
-\
-static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc23)(uint8_t *dst, uint8_t *src, int stride){\
-    int16_t tmp[SIZE*(SIZE+5)*sizeof(pixel)];\
-    uint8_t halfH[SIZE*SIZE*sizeof(pixel)];\
-    uint8_t halfHV[SIZE*SIZE*sizeof(pixel)];\
-    FUNC(put_h264_qpel ## SIZE ## _h_lowpass)(halfH, src + stride, SIZE*sizeof(pixel), stride);\
-    FUNC(put_h264_qpel ## SIZE ## _hv_lowpass)(halfHV, tmp, src, SIZE*sizeof(pixel), SIZE*sizeof(pixel), stride);\
-    FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfH, halfHV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\
-}\
-\
-static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc12)(uint8_t *dst, uint8_t *src, int stride){\
-    uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\
-    uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\
-    int16_t tmp[SIZE*(SIZE+5)*sizeof(pixel)];\
-    uint8_t halfV[SIZE*SIZE*sizeof(pixel)];\
-    uint8_t halfHV[SIZE*SIZE*sizeof(pixel)];\
-    FUNC(copy_block ## SIZE )(full, src - stride*2, SIZE*sizeof(pixel),  stride, SIZE + 5);\
-    FUNC(put_h264_qpel ## SIZE ## _v_lowpass)(halfV, full_mid, SIZE*sizeof(pixel), SIZE*sizeof(pixel));\
-    FUNC(put_h264_qpel ## SIZE ## _hv_lowpass)(halfHV, tmp, src, SIZE*sizeof(pixel), SIZE*sizeof(pixel), stride);\
-    FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfV, halfHV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\
-}\
-\
-static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc32)(uint8_t *dst, uint8_t *src, int stride){\
-    uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\
-    uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\
-    int16_t tmp[SIZE*(SIZE+5)*sizeof(pixel)];\
-    uint8_t halfV[SIZE*SIZE*sizeof(pixel)];\
-    uint8_t halfHV[SIZE*SIZE*sizeof(pixel)];\
-    FUNC(copy_block ## SIZE )(full, src - stride*2 + sizeof(pixel), SIZE*sizeof(pixel),  stride, SIZE + 5);\
-    FUNC(put_h264_qpel ## SIZE ## _v_lowpass)(halfV, full_mid, SIZE*sizeof(pixel), SIZE*sizeof(pixel));\
-    FUNC(put_h264_qpel ## SIZE ## _hv_lowpass)(halfHV, tmp, src, SIZE*sizeof(pixel), SIZE*sizeof(pixel), stride);\
-    FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfV, halfHV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\
-}\
-
-#define op_avg(a, b)  a = (((a)+CLIP(((b) + 16)>>5)+1)>>1)
-//#define op_avg2(a, b) a = (((a)*w1+cm[((b) + 16)>>5]*w2 + o + 64)>>7)
-#define op_put(a, b)  a = CLIP(((b) + 16)>>5)
-#define op2_avg(a, b)  a = (((a)+CLIP(((b) + 512)>>10)+1)>>1)
-#define op2_put(a, b)  a = CLIP(((b) + 512)>>10)
-
-H264_LOWPASS(put_       , op_put, op2_put)
-H264_LOWPASS(avg_       , op_avg, op2_avg)
-H264_MC(put_, 2)
-H264_MC(put_, 4)
-H264_MC(put_, 8)
-H264_MC(put_, 16)
-H264_MC(avg_, 4)
-H264_MC(avg_, 8)
-H264_MC(avg_, 16)
-
-#undef op_avg
-#undef op_put
-#undef op2_avg
-#undef op2_put
-#endif
-
-#if BIT_DEPTH == 8
-#   define put_h264_qpel8_mc00_8_c  ff_put_pixels8x8_8_c
-#   define avg_h264_qpel8_mc00_8_c  ff_avg_pixels8x8_8_c
-#   define put_h264_qpel16_mc00_8_c ff_put_pixels16x16_8_c
-#   define avg_h264_qpel16_mc00_8_c ff_avg_pixels16x16_8_c
-#elif BIT_DEPTH == 9
-#   define put_h264_qpel8_mc00_9_c  ff_put_pixels8x8_9_c
-#   define avg_h264_qpel8_mc00_9_c  ff_avg_pixels8x8_9_c
-#   define put_h264_qpel16_mc00_9_c ff_put_pixels16x16_9_c
-#   define avg_h264_qpel16_mc00_9_c ff_avg_pixels16x16_9_c
-#elif BIT_DEPTH == 10
-#   define put_h264_qpel8_mc00_10_c  ff_put_pixels8x8_10_c
-#   define avg_h264_qpel8_mc00_10_c  ff_avg_pixels8x8_10_c
-#   define put_h264_qpel16_mc00_10_c ff_put_pixels16x16_10_c
-#   define avg_h264_qpel16_mc00_10_c ff_avg_pixels16x16_10_c
-#endif
-
-void FUNCC(ff_put_pixels8x8)(uint8_t *dst, uint8_t *src, int stride) {
-    FUNCC(put_pixels8)(dst, src, stride, 8);
-}
-void FUNCC(ff_avg_pixels8x8)(uint8_t *dst, uint8_t *src, int stride) {
-    FUNCC(avg_pixels8)(dst, src, stride, 8);
-}
-void FUNCC(ff_put_pixels16x16)(uint8_t *dst, uint8_t *src, int stride) {
-    FUNCC(put_pixels16)(dst, src, stride, 16);
-}
-void FUNCC(ff_avg_pixels16x16)(uint8_t *dst, uint8_t *src, int stride) {
-    FUNCC(avg_pixels16)(dst, src, stride, 16);
-}
-
-static void FUNCC(clear_block)(DCTELEM *block)
-{
-    memset(block, 0, sizeof(dctcoef)*64);
-}
-
-/**
- * memset(blocks, 0, sizeof(DCTELEM)*6*64)
- */
-static void FUNCC(clear_blocks)(DCTELEM *blocks)
-{
-    memset(blocks, 0, sizeof(dctcoef)*6*64);
-}
diff --git a/libavcodec/dsputil_template.c b/libavcodec/dsputil_template.c
index 8ca6d3e414..1d73a7d858 100644
--- a/libavcodec/dsputil_template.c
+++ b/libavcodec/dsputil_template.c
@@ -5,20 +5,20 @@
  *
  * gmc & q-pel & 32/64 bit based MC by Michael Niedermayer <michaelni@gmx.at>
  *
- * This file is part of Libav.
+ * This file is part of FFmpeg.
  *
- * Libav is free software; you can redistribute it and/or
+ * FFmpeg 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.
  *
- * Libav is distributed in the hope that it will be useful,
+ * FFmpeg 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 Libav; if not, write to the Free Software
+ * License along with FFmpeg; if not, write to the Free Software
  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  */
 
@@ -79,10 +79,10 @@ static inline void FUNC(copy_block16)(uint8_t *dst, const uint8_t *src, int dstS
 
 /* draw the edges of width 'w' of an image of size width, height */
 //FIXME check that this is ok for mpeg4 interlaced
-static void FUNCC(draw_edges)(uint8_t *_buf, int _wrap, int width, int height, int w, int sides)
+static void FUNCC(draw_edges)(uint8_t *p_buf, int p_wrap, int width, int height, int w, int sides)
 {
-    pixel *buf = (pixel*)_buf;
-    int wrap = _wrap / sizeof(pixel);
+    pixel *buf = (pixel*)p_buf;
+    int wrap = p_wrap / sizeof(pixel);
     pixel *ptr, *last_line;
     int i;
 
@@ -192,12 +192,12 @@ void FUNC(ff_emulated_edge_mc)(uint8_t *buf, const uint8_t *src, int linesize, i
     }
 }
 
-static void FUNCC(add_pixels8)(uint8_t *restrict _pixels, DCTELEM *_block, int line_size)
+static void FUNCC(add_pixels8)(uint8_t *restrict p_pixels, DCTELEM *p_block, int line_size)
 {
     int i;
-    pixel *restrict pixels = (pixel *restrict)_pixels;
-    dctcoef *block = (dctcoef*)_block;
-    line_size /= sizeof(pixel);
+    pixel *restrict pixels = (pixel *restrict)p_pixels;
+    dctcoef *block = (dctcoef*)p_block;
+    line_size >>= sizeof(pixel)-1;
 
     for(i=0;i<8;i++) {
         pixels[0] += block[0];
@@ -213,12 +213,12 @@ static void FUNCC(add_pixels8)(uint8_t *restrict _pixels, DCTELEM *_block, int l
     }
 }
 
-static void FUNCC(add_pixels4)(uint8_t *restrict _pixels, DCTELEM *_block, int line_size)
+static void FUNCC(add_pixels4)(uint8_t *restrict p_pixels, DCTELEM *p_block, int line_size)
 {
     int i;
-    pixel *restrict pixels = (pixel *restrict)_pixels;
-    dctcoef *block = (dctcoef*)_block;
-    line_size /= sizeof(pixel);
+    pixel *restrict pixels = (pixel *restrict)p_pixels;
+    dctcoef *block = (dctcoef*)p_block;
+    line_size >>= sizeof(pixel)-1;
 
     for(i=0;i<4;i++) {
         pixels[0] += block[0];
@@ -582,12 +582,12 @@ 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 *_block, const uint8_t *_pixels, int line_size, int h)\
+static inline void FUNCC(OPNAME ## _pixels2_xy2)(uint8_t *p_block, const uint8_t *p_pixels, int line_size, int h)\
 {\
         int i, a0, b0, a1, b1;\
-        pixel *block = (pixel*)_block;\
-        const pixel *pixels = (const pixel*)_pixels;\
-        line_size /= sizeof(pixel);\
+        pixel *block = (pixel*)p_block;\
+        const pixel *pixels = (const pixel*)p_pixels;\
+        line_size >>= sizeof(pixel)-1;\
         a0= pixels[0];\
         b0= pixels[1] + 2;\
         a0 += b0;\
@@ -769,15 +769,15 @@ static void FUNCC(put_no_rnd_pixels8_l2)(uint8_t *dst, const uint8_t *a, const u
 }
 
 #define H264_CHROMA_MC(OPNAME, OP)\
-static void FUNCC(OPNAME ## h264_chroma_mc2)(uint8_t *_dst/*align 8*/, uint8_t *_src/*align 1*/, int stride, int h, int x, int y){\
-    pixel *dst = (pixel*)_dst;\
-    pixel *src = (pixel*)_src;\
+static void FUNCC(OPNAME ## h264_chroma_mc2)(uint8_t *p_dst/*align 8*/, uint8_t *p_src/*align 1*/, int stride, int h, int x, int y){\
+    pixel *dst = (pixel*)p_dst;\
+    pixel *src = (pixel*)p_src;\
     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;\
-    stride /= sizeof(pixel);\
+    stride >>= sizeof(pixel)-1;\
     \
     assert(x<8 && y<8 && x>=0 && y>=0);\
 \
@@ -800,15 +800,15 @@ static void FUNCC(OPNAME ## h264_chroma_mc2)(uint8_t *_dst/*align 8*/, uint8_t *
     }\
 }\
 \
-static void FUNCC(OPNAME ## h264_chroma_mc4)(uint8_t *_dst/*align 8*/, uint8_t *_src/*align 1*/, int stride, int h, int x, int y){\
-    pixel *dst = (pixel*)_dst;\
-    pixel *src = (pixel*)_src;\
+static void FUNCC(OPNAME ## h264_chroma_mc4)(uint8_t *p_dst/*align 8*/, uint8_t *p_src/*align 1*/, int stride, int h, int x, int y){\
+    pixel *dst = (pixel*)p_dst;\
+    pixel *src = (pixel*)p_src;\
     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;\
-    stride /= sizeof(pixel);\
+    stride >>= sizeof(pixel)-1;\
     \
     assert(x<8 && y<8 && x>=0 && y>=0);\
 \
@@ -835,15 +835,15 @@ static void FUNCC(OPNAME ## h264_chroma_mc4)(uint8_t *_dst/*align 8*/, uint8_t *
     }\
 }\
 \
-static void FUNCC(OPNAME ## h264_chroma_mc8)(uint8_t *_dst/*align 8*/, uint8_t *_src/*align 1*/, int stride, int h, int x, int y){\
-    pixel *dst = (pixel*)_dst;\
-    pixel *src = (pixel*)_src;\
+static void FUNCC(OPNAME ## h264_chroma_mc8)(uint8_t *p_dst/*align 8*/, uint8_t *p_src/*align 1*/, int stride, int h, int x, int y){\
+    pixel *dst = (pixel*)p_dst;\
+    pixel *src = (pixel*)p_src;\
     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;\
-    stride /= sizeof(pixel);\
+    stride >>= sizeof(pixel)-1;\
     \
     assert(x<8 && y<8 && x>=0 && y>=0);\
 \
@@ -887,14 +887,14 @@ H264_CHROMA_MC(avg_       , op_avg)
 #undef op_put
 
 #define H264_LOWPASS(OPNAME, OP, OP2) \
-static av_unused void FUNC(OPNAME ## h264_qpel2_h_lowpass)(uint8_t *_dst, uint8_t *_src, int dstStride, int srcStride){\
+static av_unused void FUNC(OPNAME ## h264_qpel2_h_lowpass)(uint8_t *p_dst, uint8_t *p_src, int dstStride, int srcStride){\
     const int h=2;\
     INIT_CLIP\
     int i;\
-    pixel *dst = (pixel*)_dst;\
-    pixel *src = (pixel*)_src;\
-    dstStride /= sizeof(pixel);\
-    srcStride /= sizeof(pixel);\
+    pixel *dst = (pixel*)p_dst;\
+    pixel *src = (pixel*)p_src;\
+    dstStride >>= sizeof(pixel)-1;\
+    srcStride >>= sizeof(pixel)-1;\
     for(i=0; i<h; i++)\
     {\
         OP(dst[0], (src[0]+src[1])*20 - (src[-1]+src[2])*5 + (src[-2]+src[3]));\
@@ -904,14 +904,14 @@ static av_unused void FUNC(OPNAME ## h264_qpel2_h_lowpass)(uint8_t *_dst, uint8_
     }\
 }\
 \
-static av_unused void FUNC(OPNAME ## h264_qpel2_v_lowpass)(uint8_t *_dst, uint8_t *_src, int dstStride, int srcStride){\
+static av_unused void FUNC(OPNAME ## h264_qpel2_v_lowpass)(uint8_t *p_dst, uint8_t *p_src, int dstStride, int srcStride){\
     const int w=2;\
     INIT_CLIP\
     int i;\
-    pixel *dst = (pixel*)_dst;\
-    pixel *src = (pixel*)_src;\
-    dstStride /= sizeof(pixel);\
-    srcStride /= sizeof(pixel);\
+    pixel *dst = (pixel*)p_dst;\
+    pixel *src = (pixel*)p_src;\
+    dstStride >>= sizeof(pixel)-1;\
+    srcStride >>= sizeof(pixel)-1;\
     for(i=0; i<w; i++)\
     {\
         const int srcB= src[-2*srcStride];\
@@ -928,16 +928,16 @@ static av_unused void FUNC(OPNAME ## h264_qpel2_v_lowpass)(uint8_t *_dst, uint8_
     }\
 }\
 \
-static av_unused void FUNC(OPNAME ## h264_qpel2_hv_lowpass)(uint8_t *_dst, int16_t *tmp, uint8_t *_src, int dstStride, int tmpStride, int srcStride){\
+static av_unused void FUNC(OPNAME ## h264_qpel2_hv_lowpass)(uint8_t *p_dst, int16_t *tmp, uint8_t *p_src, int dstStride, int tmpStride, int srcStride){\
     const int h=2;\
     const int w=2;\
     const int pad = (BIT_DEPTH > 9) ? (-10 * ((1<<BIT_DEPTH)-1)) : 0;\
     INIT_CLIP\
     int i;\
-    pixel *dst = (pixel*)_dst;\
-    pixel *src = (pixel*)_src;\
-    dstStride /= sizeof(pixel);\
-    srcStride /= sizeof(pixel);\
+    pixel *dst = (pixel*)p_dst;\
+    pixel *src = (pixel*)p_src;\
+    dstStride >>= sizeof(pixel)-1;\
+    srcStride >>= sizeof(pixel)-1;\
     src -= 2*srcStride;\
     for(i=0; i<h+5; i++)\
     {\
@@ -962,14 +962,14 @@ static av_unused void FUNC(OPNAME ## h264_qpel2_hv_lowpass)(uint8_t *_dst, int16
         tmp++;\
     }\
 }\
-static void FUNC(OPNAME ## h264_qpel4_h_lowpass)(uint8_t *_dst, uint8_t *_src, int dstStride, int srcStride){\
+static void FUNC(OPNAME ## h264_qpel4_h_lowpass)(uint8_t *p_dst, uint8_t *p_src, int dstStride, int srcStride){\
     const int h=4;\
     INIT_CLIP\
     int i;\
-    pixel *dst = (pixel*)_dst;\
-    pixel *src = (pixel*)_src;\
-    dstStride /= sizeof(pixel);\
-    srcStride /= sizeof(pixel);\
+    pixel *dst = (pixel*)p_dst;\
+    pixel *src = (pixel*)p_src;\
+    dstStride >>= sizeof(pixel)-1;\
+    srcStride >>= sizeof(pixel)-1;\
     for(i=0; i<h; i++)\
     {\
         OP(dst[0], (src[0]+src[1])*20 - (src[-1]+src[2])*5 + (src[-2]+src[3]));\
@@ -981,14 +981,14 @@ static void FUNC(OPNAME ## h264_qpel4_h_lowpass)(uint8_t *_dst, uint8_t *_src, i
     }\
 }\
 \
-static void FUNC(OPNAME ## h264_qpel4_v_lowpass)(uint8_t *_dst, uint8_t *_src, int dstStride, int srcStride){\
+static void FUNC(OPNAME ## h264_qpel4_v_lowpass)(uint8_t *p_dst, uint8_t *p_src, int dstStride, int srcStride){\
     const int w=4;\
     INIT_CLIP\
     int i;\
-    pixel *dst = (pixel*)_dst;\
-    pixel *src = (pixel*)_src;\
-    dstStride /= sizeof(pixel);\
-    srcStride /= sizeof(pixel);\
+    pixel *dst = (pixel*)p_dst;\
+    pixel *src = (pixel*)p_src;\
+    dstStride >>= sizeof(pixel)-1;\
+    srcStride >>= sizeof(pixel)-1;\
     for(i=0; i<w; i++)\
     {\
         const int srcB= src[-2*srcStride];\
@@ -1009,16 +1009,16 @@ static void FUNC(OPNAME ## h264_qpel4_v_lowpass)(uint8_t *_dst, uint8_t *_src, i
     }\
 }\
 \
-static void FUNC(OPNAME ## h264_qpel4_hv_lowpass)(uint8_t *_dst, int16_t *tmp, uint8_t *_src, int dstStride, int tmpStride, int srcStride){\
+static void FUNC(OPNAME ## h264_qpel4_hv_lowpass)(uint8_t *p_dst, int16_t *tmp, uint8_t *p_src, int dstStride, int tmpStride, int srcStride){\
     const int h=4;\
     const int w=4;\
     const int pad = (BIT_DEPTH > 9) ? (-10 * ((1<<BIT_DEPTH)-1)) : 0;\
     INIT_CLIP\
     int i;\
-    pixel *dst = (pixel*)_dst;\
-    pixel *src = (pixel*)_src;\
-    dstStride /= sizeof(pixel);\
-    srcStride /= sizeof(pixel);\
+    pixel *dst = (pixel*)p_dst;\
+    pixel *src = (pixel*)p_src;\
+    dstStride >>= sizeof(pixel)-1;\
+    srcStride >>= sizeof(pixel)-1;\
     src -= 2*srcStride;\
     for(i=0; i<h+5; i++)\
     {\
@@ -1050,14 +1050,14 @@ static void FUNC(OPNAME ## h264_qpel4_hv_lowpass)(uint8_t *_dst, int16_t *tmp, u
     }\
 }\
 \
-static void FUNC(OPNAME ## h264_qpel8_h_lowpass)(uint8_t *_dst, uint8_t *_src, int dstStride, int srcStride){\
+static void FUNC(OPNAME ## h264_qpel8_h_lowpass)(uint8_t *p_dst, uint8_t *p_src, int dstStride, int srcStride){\
     const int h=8;\
     INIT_CLIP\
     int i;\
-    pixel *dst = (pixel*)_dst;\
-    pixel *src = (pixel*)_src;\
-    dstStride /= sizeof(pixel);\
-    srcStride /= sizeof(pixel);\
+    pixel *dst = (pixel*)p_dst;\
+    pixel *src = (pixel*)p_src;\
+    dstStride >>= sizeof(pixel)-1;\
+    srcStride >>= sizeof(pixel)-1;\
     for(i=0; i<h; i++)\
     {\
         OP(dst[0], (src[0]+src[1])*20 - (src[-1]+src[2])*5 + (src[-2]+src[3 ]));\
@@ -1073,14 +1073,14 @@ static void FUNC(OPNAME ## h264_qpel8_h_lowpass)(uint8_t *_dst, uint8_t *_src, i
     }\
 }\
 \
-static void FUNC(OPNAME ## h264_qpel8_v_lowpass)(uint8_t *_dst, uint8_t *_src, int dstStride, int srcStride){\
+static void FUNC(OPNAME ## h264_qpel8_v_lowpass)(uint8_t *p_dst, uint8_t *p_src, int dstStride, int srcStride){\
     const int w=8;\
     INIT_CLIP\
     int i;\
-    pixel *dst = (pixel*)_dst;\
-    pixel *src = (pixel*)_src;\
-    dstStride /= sizeof(pixel);\
-    srcStride /= sizeof(pixel);\
+    pixel *dst = (pixel*)p_dst;\
+    pixel *src = (pixel*)p_src;\
+    dstStride >>= sizeof(pixel)-1;\
+    srcStride >>= sizeof(pixel)-1;\
     for(i=0; i<w; i++)\
     {\
         const int srcB= src[-2*srcStride];\
@@ -1109,16 +1109,16 @@ static void FUNC(OPNAME ## h264_qpel8_v_lowpass)(uint8_t *_dst, uint8_t *_src, i
     }\
 }\
 \
-static void FUNC(OPNAME ## h264_qpel8_hv_lowpass)(uint8_t *_dst, int16_t *tmp, uint8_t *_src, int dstStride, int tmpStride, int srcStride){\
+static void FUNC(OPNAME ## h264_qpel8_hv_lowpass)(uint8_t *p_dst, int16_t *tmp, uint8_t *p_src, int dstStride, int tmpStride, int srcStride){\
     const int h=8;\
     const int w=8;\
     const int pad = (BIT_DEPTH > 9) ? (-10 * ((1<<BIT_DEPTH)-1)) : 0;\
     INIT_CLIP\
     int i;\
-    pixel *dst = (pixel*)_dst;\
-    pixel *src = (pixel*)_src;\
-    dstStride /= sizeof(pixel);\
-    srcStride /= sizeof(pixel);\
+    pixel *dst = (pixel*)p_dst;\
+    pixel *src = (pixel*)p_src;\
+    dstStride >>= sizeof(pixel)-1;\
+    srcStride >>= sizeof(pixel)-1;\
     src -= 2*srcStride;\
     for(i=0; i<h+5; i++)\
     {\
@@ -1389,3 +1389,4 @@ static void FUNCC(clear_blocks)(DCTELEM *blocks)
 {
     memset(blocks, 0, sizeof(dctcoef)*6*64);
 }
+
diff --git a/libavcodec/h264_high_depth.h b/libavcodec/h264_high_depth.h
deleted file mode 100644
index da12d94b7c..0000000000
--- a/libavcodec/h264_high_depth.h
+++ /dev/null
@@ -1,86 +0,0 @@
-#include "dsputil.h"
-
-#ifndef BIT_DEPTH
-#define BIT_DEPTH 8
-#endif
-
-#ifdef AVCODEC_H264_HIGH_DEPTH_H
-#   undef pixel
-#   undef pixel2
-#   undef pixel4
-#   undef dctcoef
-#   undef INIT_CLIP
-#   undef no_rnd_avg_pixel4
-#   undef rnd_avg_pixel4
-#   undef AV_RN2P
-#   undef AV_RN4P
-#   undef AV_WN2P
-#   undef AV_WN4P
-#   undef AV_WN4PA
-#   undef CLIP
-#   undef FUNC
-#   undef FUNCC
-#   undef av_clip_pixel
-#   undef PIXEL_SPLAT_X4
-#else
-#   define AVCODEC_H264_HIGH_DEPTH_H
-#   define CLIP_PIXEL(depth)\
-    static inline uint16_t av_clip_pixel_ ## depth (int p)\
-    {\
-        const int pixel_max = (1 << depth)-1;\
-        return (p & ~pixel_max) ? (-p)>>31 & pixel_max : p;\
-    }
-
-CLIP_PIXEL( 9)
-CLIP_PIXEL(10)
-#endif
-
-#if BIT_DEPTH > 8
-#   define pixel  uint16_t
-#   define pixel2 uint32_t
-#   define pixel4 uint64_t
-#   define dctcoef int32_t
-
-#   define INIT_CLIP
-#   define no_rnd_avg_pixel4 no_rnd_avg64
-#   define    rnd_avg_pixel4    rnd_avg64
-#   define AV_RN2P  AV_RN32
-#   define AV_RN4P  AV_RN64
-#   define AV_WN2P  AV_WN32
-#   define AV_WN4P  AV_WN64
-#   define AV_WN4PA AV_WN64A
-#   define PIXEL_SPLAT_X4(x) ((x)*0x0001000100010001ULL)
-#else
-#   define pixel  uint8_t
-#   define pixel2 uint16_t
-#   define pixel4 uint32_t
-#   define dctcoef int16_t
-
-#   define INIT_CLIP uint8_t *cm = ff_cropTbl + MAX_NEG_CROP;
-#   define no_rnd_avg_pixel4 no_rnd_avg32
-#   define    rnd_avg_pixel4    rnd_avg32
-#   define AV_RN2P  AV_RN16
-#   define AV_RN4P  AV_RN32
-#   define AV_WN2P  AV_WN16
-#   define AV_WN4P  AV_WN32
-#   define AV_WN4PA AV_WN32A
-#   define PIXEL_SPLAT_X4(x) ((x)*0x01010101U)
-#endif
-
-#if BIT_DEPTH == 8
-#   define av_clip_pixel(a) av_clip_uint8(a)
-#   define CLIP(a) cm[a]
-#   define FUNC(a)  a ## _8
-#   define FUNCC(a) a ## _8_c
-#elif BIT_DEPTH == 9
-#   define av_clip_pixel(a) av_clip_pixel_9(a)
-#   define CLIP(a)          av_clip_pixel_9(a)
-#   define FUNC(a)  a ## _9
-#   define FUNCC(a) a ## _9_c
-#elif BIT_DEPTH == 10
-#   define av_clip_pixel(a) av_clip_pixel_10(a)
-#   define CLIP(a)          av_clip_pixel_10(a)
-#   define FUNC(a)  a ## _10
-#   define FUNCC(a) a ## _10_c
-#endif
-
diff --git a/libavcodec/h264dsp_internal.h b/libavcodec/h264dsp_internal.h
deleted file mode 100644
index f9a717b8c4..0000000000
--- a/libavcodec/h264dsp_internal.h
+++ /dev/null
@@ -1,288 +0,0 @@
-/*
- * H.26L/H.264/AVC/JVT/14496-10/... encoder/decoder
- * Copyright (c) 2003-2011 Michael Niedermayer <michaelni@gmx.at>
- *
- * This file is part of FFmpeg.
- *
- * FFmpeg 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,
- * 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
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
- */
-
-/**
- * @file
- * H.264 / AVC / MPEG4 part10 DSP functions.
- * @author Michael Niedermayer <michaelni@gmx.at>
- */
-
-#include "h264_high_depth.h"
-
-#define op_scale1(x)  block[x] = av_clip_pixel( (block[x]*weight + offset) >> log2_denom )
-#define op_scale2(x)  dst[x] = av_clip_pixel( (src[x]*weights + dst[x]*weightd + offset) >> (log2_denom+1))
-#define H264_WEIGHT(W,H) \
-static void FUNCC(weight_h264_pixels ## W ## x ## H)(uint8_t *p_block, int stride, int log2_denom, int weight, int offset){ \
-    int y; \
-    pixel *block = (pixel*)p_block; \
-    stride >>= sizeof(pixel)-1; \
-    offset <<= (log2_denom + (BIT_DEPTH-8)); \
-    if(log2_denom) offset += 1<<(log2_denom-1); \
-    for(y=0; y<H; y++, block += stride){ \
-        op_scale1(0); \
-        op_scale1(1); \
-        if(W==2) continue; \
-        op_scale1(2); \
-        op_scale1(3); \
-        if(W==4) continue; \
-        op_scale1(4); \
-        op_scale1(5); \
-        op_scale1(6); \
-        op_scale1(7); \
-        if(W==8) continue; \
-        op_scale1(8); \
-        op_scale1(9); \
-        op_scale1(10); \
-        op_scale1(11); \
-        op_scale1(12); \
-        op_scale1(13); \
-        op_scale1(14); \
-        op_scale1(15); \
-    } \
-} \
-static void FUNCC(biweight_h264_pixels ## W ## x ## H)(uint8_t *p_dst, uint8_t *p_src, int stride, int log2_denom, int weightd, int weights, int offset){ \
-    int y; \
-    pixel *dst = (pixel*)p_dst; \
-    pixel *src = (pixel*)p_src; \
-    stride >>= sizeof(pixel)-1; \
-    offset = ((offset + 1) | 1) << log2_denom; \
-    for(y=0; y<H; y++, dst += stride, src += stride){ \
-        op_scale2(0); \
-        op_scale2(1); \
-        if(W==2) continue; \
-        op_scale2(2); \
-        op_scale2(3); \
-        if(W==4) continue; \
-        op_scale2(4); \
-        op_scale2(5); \
-        op_scale2(6); \
-        op_scale2(7); \
-        if(W==8) continue; \
-        op_scale2(8); \
-        op_scale2(9); \
-        op_scale2(10); \
-        op_scale2(11); \
-        op_scale2(12); \
-        op_scale2(13); \
-        op_scale2(14); \
-        op_scale2(15); \
-    } \
-}
-
-H264_WEIGHT(16,16)
-H264_WEIGHT(16,8)
-H264_WEIGHT(8,16)
-H264_WEIGHT(8,8)
-H264_WEIGHT(8,4)
-H264_WEIGHT(4,8)
-H264_WEIGHT(4,4)
-H264_WEIGHT(4,2)
-H264_WEIGHT(2,4)
-H264_WEIGHT(2,2)
-
-#undef op_scale1
-#undef op_scale2
-#undef H264_WEIGHT
-
-static av_always_inline av_flatten void FUNCC(h264_loop_filter_luma)(uint8_t *p_pix, int xstride, int ystride, int alpha, int beta, int8_t *tc0)
-{
-    pixel *pix = (pixel*)p_pix;
-    int i, d;
-    xstride >>= sizeof(pixel)-1;
-    ystride >>= sizeof(pixel)-1;
-    for( i = 0; i < 4; i++ ) {
-        if( tc0[i] < 0 ) {
-            pix += 4*ystride;
-            continue;
-        }
-        for( d = 0; d < 4; d++ ) {
-            const int p0 = pix[-1*xstride];
-            const int p1 = pix[-2*xstride];
-            const int p2 = pix[-3*xstride];
-            const int q0 = pix[0];
-            const int q1 = pix[1*xstride];
-            const int q2 = pix[2*xstride];
-
-            if( FFABS( p0 - q0 ) < alpha &&
-                FFABS( p1 - p0 ) < beta &&
-                FFABS( q1 - q0 ) < beta ) {
-
-                int tc = tc0[i];
-                int i_delta;
-
-                if( FFABS( p2 - p0 ) < beta ) {
-                    if(tc0[i])
-                    pix[-2*xstride] = p1 + av_clip( (( p2 + ( ( p0 + q0 + 1 ) >> 1 ) ) >> 1) - p1, -tc0[i], tc0[i] );
-                    tc++;
-                }
-                if( FFABS( q2 - q0 ) < beta ) {
-                    if(tc0[i])
-                    pix[   xstride] = q1 + av_clip( (( q2 + ( ( p0 + q0 + 1 ) >> 1 ) ) >> 1) - q1, -tc0[i], tc0[i] );
-                    tc++;
-                }
-
-                i_delta = av_clip( (((q0 - p0 ) << 2) + (p1 - q1) + 4) >> 3, -tc, tc );
-                pix[-xstride] = av_clip_pixel( p0 + i_delta );    /* p0' */
-                pix[0]        = av_clip_pixel( q0 - i_delta );    /* q0' */
-            }
-            pix += ystride;
-        }
-    }
-}
-static void FUNCC(h264_v_loop_filter_luma)(uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0)
-{
-    FUNCC(h264_loop_filter_luma)(pix, stride, sizeof(pixel), alpha, beta, tc0);
-}
-static void FUNCC(h264_h_loop_filter_luma)(uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0)
-{
-    FUNCC(h264_loop_filter_luma)(pix, sizeof(pixel), stride, alpha, beta, tc0);
-}
-
-static av_always_inline av_flatten void FUNCC(h264_loop_filter_luma_intra)(uint8_t *p_pix, int xstride, int ystride, int alpha, int beta)
-{
-    pixel *pix = (pixel*)p_pix;
-    int d;
-    xstride >>= sizeof(pixel)-1;
-    ystride >>= sizeof(pixel)-1;
-    for( d = 0; d < 16; d++ ) {
-        const int p2 = pix[-3*xstride];
-        const int p1 = pix[-2*xstride];
-        const int p0 = pix[-1*xstride];
-
-        const int q0 = pix[ 0*xstride];
-        const int q1 = pix[ 1*xstride];
-        const int q2 = pix[ 2*xstride];
-
-        if( FFABS( p0 - q0 ) < alpha &&
-            FFABS( p1 - p0 ) < beta &&
-            FFABS( q1 - q0 ) < beta ) {
-
-            if(FFABS( p0 - q0 ) < (( alpha >> 2 ) + 2 )){
-                if( FFABS( p2 - p0 ) < beta)
-                {
-                    const int p3 = pix[-4*xstride];
-                    /* p0', p1', p2' */
-                    pix[-1*xstride] = ( p2 + 2*p1 + 2*p0 + 2*q0 + q1 + 4 ) >> 3;
-                    pix[-2*xstride] = ( p2 + p1 + p0 + q0 + 2 ) >> 2;
-                    pix[-3*xstride] = ( 2*p3 + 3*p2 + p1 + p0 + q0 + 4 ) >> 3;
-                } else {
-                    /* p0' */
-                    pix[-1*xstride] = ( 2*p1 + p0 + q1 + 2 ) >> 2;
-                }
-                if( FFABS( q2 - q0 ) < beta)
-                {
-                    const int q3 = pix[3*xstride];
-                    /* q0', q1', q2' */
-                    pix[0*xstride] = ( p1 + 2*p0 + 2*q0 + 2*q1 + q2 + 4 ) >> 3;
-                    pix[1*xstride] = ( p0 + q0 + q1 + q2 + 2 ) >> 2;
-                    pix[2*xstride] = ( 2*q3 + 3*q2 + q1 + q0 + p0 + 4 ) >> 3;
-                } else {
-                    /* q0' */
-                    pix[0*xstride] = ( 2*q1 + q0 + p1 + 2 ) >> 2;
-                }
-            }else{
-                /* p0', q0' */
-                pix[-1*xstride] = ( 2*p1 + p0 + q1 + 2 ) >> 2;
-                pix[ 0*xstride] = ( 2*q1 + q0 + p1 + 2 ) >> 2;
-            }
-        }
-        pix += ystride;
-    }
-}
-static void FUNCC(h264_v_loop_filter_luma_intra)(uint8_t *pix, int stride, int alpha, int beta)
-{
-    FUNCC(h264_loop_filter_luma_intra)(pix, stride, sizeof(pixel), alpha, beta);
-}
-static void FUNCC(h264_h_loop_filter_luma_intra)(uint8_t *pix, int stride, int alpha, int beta)
-{
-    FUNCC(h264_loop_filter_luma_intra)(pix, sizeof(pixel), stride, alpha, beta);
-}
-
-static av_always_inline av_flatten void FUNCC(h264_loop_filter_chroma)(uint8_t *p_pix, int xstride, int ystride, int alpha, int beta, int8_t *tc0)
-{
-    pixel *pix = (pixel*)p_pix;
-    int i, d;
-    xstride >>= sizeof(pixel)-1;
-    ystride >>= sizeof(pixel)-1;
-    for( i = 0; i < 4; i++ ) {
-        const int tc = tc0[i];
-        if( tc <= 0 ) {
-            pix += 2*ystride;
-            continue;
-        }
-        for( d = 0; d < 2; d++ ) {
-            const int p0 = pix[-1*xstride];
-            const int p1 = pix[-2*xstride];
-            const int q0 = pix[0];
-            const int q1 = pix[1*xstride];
-
-            if( FFABS( p0 - q0 ) < alpha &&
-                FFABS( p1 - p0 ) < beta &&
-                FFABS( q1 - q0 ) < beta ) {
-
-                int delta = av_clip( (((q0 - p0 ) << 2) + (p1 - q1) + 4) >> 3, -tc, tc );
-
-                pix[-xstride] = av_clip_pixel( p0 + delta );    /* p0' */
-                pix[0]        = av_clip_pixel( q0 - delta );    /* q0' */
-            }
-            pix += ystride;
-        }
-    }
-}
-static void FUNCC(h264_v_loop_filter_chroma)(uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0)
-{
-    FUNCC(h264_loop_filter_chroma)(pix, stride, sizeof(pixel), alpha, beta, tc0);
-}
-static void FUNCC(h264_h_loop_filter_chroma)(uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0)
-{
-    FUNCC(h264_loop_filter_chroma)(pix, sizeof(pixel), stride, alpha, beta, tc0);
-}
-
-static av_always_inline av_flatten void FUNCC(h264_loop_filter_chroma_intra)(uint8_t *p_pix, int xstride, int ystride, int alpha, int beta)
-{
-    pixel *pix = (pixel*)p_pix;
-    int d;
-    xstride >>= sizeof(pixel)-1;
-    ystride >>= sizeof(pixel)-1;
-    for( d = 0; d < 8; d++ ) {
-        const int p0 = pix[-1*xstride];
-        const int p1 = pix[-2*xstride];
-        const int q0 = pix[0];
-        const int q1 = pix[1*xstride];
-
-        if( FFABS( p0 - q0 ) < alpha &&
-            FFABS( p1 - p0 ) < beta &&
-            FFABS( q1 - q0 ) < beta ) {
-
-            pix[-xstride] = ( 2*p1 + p0 + q1 + 2 ) >> 2;   /* p0' */
-            pix[0]        = ( 2*q1 + q0 + p1 + 2 ) >> 2;   /* q0' */
-        }
-        pix += ystride;
-    }
-}
-static void FUNCC(h264_v_loop_filter_chroma_intra)(uint8_t *pix, int stride, int alpha, int beta)
-{
-    FUNCC(h264_loop_filter_chroma_intra)(pix, stride, sizeof(pixel), alpha, beta);
-}
-static void FUNCC(h264_h_loop_filter_chroma_intra)(uint8_t *pix, int stride, int alpha, int beta)
-{
-    FUNCC(h264_loop_filter_chroma_intra)(pix, sizeof(pixel), stride, alpha, beta);
-}
diff --git a/libavcodec/h264dsp_template.c b/libavcodec/h264dsp_template.c
index 91162ea900..6816e81ff2 100644
--- a/libavcodec/h264dsp_template.c
+++ b/libavcodec/h264dsp_template.c
@@ -1,21 +1,21 @@
 /*
  * H.26L/H.264/AVC/JVT/14496-10/... encoder/decoder
- * Copyright (c) 2003-2010 Michael Niedermayer <michaelni@gmx.at>
+ * Copyright (c) 2003-2011 Michael Niedermayer <michaelni@gmx.at>
  *
- * This file is part of Libav.
+ * This file is part of FFmpeg.
  *
- * Libav is free software; you can redistribute it and/or
+ * FFmpeg 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.
  *
- * Libav is distributed in the hope that it will be useful,
+ * FFmpeg 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 Libav; if not, write to the Free Software
+ * License along with FFmpeg; if not, write to the Free Software
  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  */
 
@@ -30,10 +30,10 @@
 #define op_scale1(x)  block[x] = av_clip_pixel( (block[x]*weight + offset) >> log2_denom )
 #define op_scale2(x)  dst[x] = av_clip_pixel( (src[x]*weights + dst[x]*weightd + offset) >> (log2_denom+1))
 #define H264_WEIGHT(W,H) \
-static void FUNCC(weight_h264_pixels ## W ## x ## H)(uint8_t *_block, int stride, int log2_denom, int weight, int offset){ \
+static void FUNCC(weight_h264_pixels ## W ## x ## H)(uint8_t *p_block, int stride, int log2_denom, int weight, int offset){ \
     int y; \
-    pixel *block = (pixel*)_block; \
-    stride /= sizeof(pixel); \
+    pixel *block = (pixel*)p_block; \
+    stride >>= sizeof(pixel)-1; \
     offset <<= (log2_denom + (BIT_DEPTH-8)); \
     if(log2_denom) offset += 1<<(log2_denom-1); \
     for(y=0; y<H; y++, block += stride){ \
@@ -58,11 +58,11 @@ static void FUNCC(weight_h264_pixels ## W ## x ## H)(uint8_t *_block, int stride
         op_scale1(15); \
     } \
 } \
-static void FUNCC(biweight_h264_pixels ## W ## x ## H)(uint8_t *_dst, uint8_t *_src, int stride, int log2_denom, int weightd, int weights, int offset){ \
+static void FUNCC(biweight_h264_pixels ## W ## x ## H)(uint8_t *p_dst, uint8_t *p_src, int stride, int log2_denom, int weightd, int weights, int offset){ \
     int y; \
-    pixel *dst = (pixel*)_dst; \
-    pixel *src = (pixel*)_src; \
-    stride /= sizeof(pixel); \
+    pixel *dst = (pixel*)p_dst; \
+    pixel *src = (pixel*)p_src; \
+    stride >>= sizeof(pixel)-1; \
     offset = ((offset + 1) | 1) << log2_denom; \
     for(y=0; y<H; y++, dst += stride, src += stride){ \
         op_scale2(0); \
@@ -102,12 +102,12 @@ H264_WEIGHT(2,2)
 #undef op_scale2
 #undef H264_WEIGHT
 
-static av_always_inline av_flatten void FUNCC(h264_loop_filter_luma)(uint8_t *_pix, int xstride, int ystride, int inner_iters, int alpha, int beta, int8_t *tc0)
+static av_always_inline av_flatten void FUNCC(h264_loop_filter_luma)(uint8_t *p_pix, int xstride, int ystride, int inner_iters, int alpha, int beta, int8_t *tc0)
 {
-    pixel *pix = (pixel*)_pix;
+    pixel *pix = (pixel*)p_pix;
     int i, d;
-    xstride /= sizeof(pixel);
-    ystride /= sizeof(pixel);
+    xstride >>= sizeof(pixel)-1;
+    ystride >>= sizeof(pixel)-1;
     alpha <<= BIT_DEPTH - 8;
     beta  <<= BIT_DEPTH - 8;
     for( i = 0; i < 4; i++ ) {
@@ -163,12 +163,12 @@ static void FUNCC(h264_h_loop_filter_luma_mbaff)(uint8_t *pix, int stride, int a
     FUNCC(h264_loop_filter_luma)(pix, sizeof(pixel), stride, 2, alpha, beta, tc0);
 }
 
-static av_always_inline av_flatten void FUNCC(h264_loop_filter_luma_intra)(uint8_t *_pix, int xstride, int ystride, int inner_iters, int alpha, int beta)
+static av_always_inline av_flatten void FUNCC(h264_loop_filter_luma_intra)(uint8_t *p_pix, int xstride, int ystride, int inner_iters, int alpha, int beta)
 {
-    pixel *pix = (pixel*)_pix;
+    pixel *pix = (pixel*)p_pix;
     int d;
-    xstride /= sizeof(pixel);
-    ystride /= sizeof(pixel);
+    xstride >>= sizeof(pixel)-1;
+    ystride >>= sizeof(pixel)-1;
     alpha <<= BIT_DEPTH - 8;
     beta  <<= BIT_DEPTH - 8;
     for( d = 0; d < 4 * inner_iters; d++ ) {
@@ -229,14 +229,14 @@ static void FUNCC(h264_h_loop_filter_luma_mbaff_intra)(uint8_t *pix, int stride,
     FUNCC(h264_loop_filter_luma_intra)(pix, sizeof(pixel), stride, 2, alpha, beta);
 }
 
-static av_always_inline av_flatten void FUNCC(h264_loop_filter_chroma)(uint8_t *_pix, int xstride, int ystride, int inner_iters, int alpha, int beta, int8_t *tc0)
+static av_always_inline av_flatten void FUNCC(h264_loop_filter_chroma)(uint8_t *p_pix, int xstride, int ystride, int inner_iters, int alpha, int beta, int8_t *tc0)
 {
-    pixel *pix = (pixel*)_pix;
+    pixel *pix = (pixel*)p_pix;
     int i, d;
-    xstride /= sizeof(pixel);
-    ystride /= sizeof(pixel);
     alpha <<= BIT_DEPTH - 8;
     beta  <<= BIT_DEPTH - 8;
+    xstride >>= sizeof(pixel)-1;
+    ystride >>= sizeof(pixel)-1;
     for( i = 0; i < 4; i++ ) {
         const int tc = ((tc0[i] - 1) << (BIT_DEPTH - 8)) + 1;
         if( tc <= 0 ) {
@@ -275,12 +275,12 @@ static void FUNCC(h264_h_loop_filter_chroma_mbaff)(uint8_t *pix, int stride, int
     FUNCC(h264_loop_filter_chroma)(pix, sizeof(pixel), stride, 1, alpha, beta, tc0);
 }
 
-static av_always_inline av_flatten void FUNCC(h264_loop_filter_chroma_intra)(uint8_t *_pix, int xstride, int ystride, int inner_iters, int alpha, int beta)
+static av_always_inline av_flatten void FUNCC(h264_loop_filter_chroma_intra)(uint8_t *p_pix, int xstride, int ystride, int inner_iters, int alpha, int beta)
 {
-    pixel *pix = (pixel*)_pix;
+    pixel *pix = (pixel*)p_pix;
     int d;
-    xstride /= sizeof(pixel);
-    ystride /= sizeof(pixel);
+    xstride >>= sizeof(pixel)-1;
+    ystride >>= sizeof(pixel)-1;
     alpha <<= BIT_DEPTH - 8;
     beta  <<= BIT_DEPTH - 8;
     for( d = 0; d < 4 * inner_iters; d++ ) {
diff --git a/libavcodec/h264idct_internal.h b/libavcodec/h264idct_internal.h
deleted file mode 100644
index d19051df90..0000000000
--- a/libavcodec/h264idct_internal.h
+++ /dev/null
@@ -1,291 +0,0 @@
-/*
- * H.264 IDCT
- * Copyright (c) 2004-2011 Michael Niedermayer <michaelni@gmx.at>
- *
- * This file is part of FFmpeg.
- *
- * FFmpeg 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,
- * 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
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
- */
-
-/**
- * @file
- * H.264 IDCT.
- * @author Michael Niedermayer <michaelni@gmx.at>
- */
-
-#include "h264_high_depth.h"
-
-#ifndef AVCODEC_H264IDCT_INTERNAL_H
-#define AVCODEC_H264IDCT_INTERNAL_H
-//FIXME this table is a duplicate from h264data.h, and will be removed once the tables from, h264 have been split
-static const uint8_t scan8[16 + 2*4]={
- 4+1*8, 5+1*8, 4+2*8, 5+2*8,
- 6+1*8, 7+1*8, 6+2*8, 7+2*8,
- 4+3*8, 5+3*8, 4+4*8, 5+4*8,
- 6+3*8, 7+3*8, 6+4*8, 7+4*8,
- 1+1*8, 2+1*8,
- 1+2*8, 2+2*8,
- 1+4*8, 2+4*8,
- 1+5*8, 2+5*8,
-};
-#endif
-
-static av_always_inline void FUNCC(idct_internal)(uint8_t *p_dst, DCTELEM *p_block, int stride, int block_stride, int shift, int add){
-    int i;
-    INIT_CLIP
-    pixel *dst = (pixel*)p_dst;
-    dctcoef *block = (dctcoef*)p_block;
-    stride >>= sizeof(pixel)-1;
-
-    block[0] += 1<<(shift-1);
-
-    for(i=0; i<4; i++){
-        const int z0=  block[i + block_stride*0]     +  block[i + block_stride*2];
-        const int z1=  block[i + block_stride*0]     -  block[i + block_stride*2];
-        const int z2= (block[i + block_stride*1]>>1) -  block[i + block_stride*3];
-        const int z3=  block[i + block_stride*1]     + (block[i + block_stride*3]>>1);
-
-        block[i + block_stride*0]= z0 + z3;
-        block[i + block_stride*1]= z1 + z2;
-        block[i + block_stride*2]= z1 - z2;
-        block[i + block_stride*3]= z0 - z3;
-    }
-
-    for(i=0; i<4; i++){
-        const int z0=  block[0 + block_stride*i]     +  block[2 + block_stride*i];
-        const int z1=  block[0 + block_stride*i]     -  block[2 + block_stride*i];
-        const int z2= (block[1 + block_stride*i]>>1) -  block[3 + block_stride*i];
-        const int z3=  block[1 + block_stride*i]     + (block[3 + block_stride*i]>>1);
-
-        dst[i + 0*stride]= CLIP(add*dst[i + 0*stride] + ((z0 + z3) >> shift));
-        dst[i + 1*stride]= CLIP(add*dst[i + 1*stride] + ((z1 + z2) >> shift));
-        dst[i + 2*stride]= CLIP(add*dst[i + 2*stride] + ((z1 - z2) >> shift));
-        dst[i + 3*stride]= CLIP(add*dst[i + 3*stride] + ((z0 - z3) >> shift));
-    }
-}
-
-void FUNCC(ff_h264_idct_add)(uint8_t *dst, DCTELEM *block, int stride){
-    FUNCC(idct_internal)(dst, block, stride, 4, 6, 1);
-}
-
-void FUNCC(ff_h264_lowres_idct_add)(uint8_t *dst, int stride, DCTELEM *block){
-    FUNCC(idct_internal)(dst, block, stride, 8, 3, 1);
-}
-
-void FUNCC(ff_h264_lowres_idct_put)(uint8_t *dst, int stride, DCTELEM *block){
-    FUNCC(idct_internal)(dst, block, stride, 8, 3, 0);
-}
-
-void FUNCC(ff_h264_idct8_add)(uint8_t *p_dst, DCTELEM *p_block, int stride){
-    int i;
-    INIT_CLIP
-    pixel *dst = (pixel*)p_dst;
-    dctcoef *block = (dctcoef*)p_block;
-    stride >>= sizeof(pixel)-1;
-
-    block[0] += 32;
-
-    for( i = 0; i < 8; i++ )
-    {
-        const int a0 =  block[i+0*8] + block[i+4*8];
-        const int a2 =  block[i+0*8] - block[i+4*8];
-        const int a4 = (block[i+2*8]>>1) - block[i+6*8];
-        const int a6 = (block[i+6*8]>>1) + block[i+2*8];
-
-        const int b0 = a0 + a6;
-        const int b2 = a2 + a4;
-        const int b4 = a2 - a4;
-        const int b6 = a0 - a6;
-
-        const int a1 = -block[i+3*8] + block[i+5*8] - block[i+7*8] - (block[i+7*8]>>1);
-        const int a3 =  block[i+1*8] + block[i+7*8] - block[i+3*8] - (block[i+3*8]>>1);
-        const int a5 = -block[i+1*8] + block[i+7*8] + block[i+5*8] + (block[i+5*8]>>1);
-        const int a7 =  block[i+3*8] + block[i+5*8] + block[i+1*8] + (block[i+1*8]>>1);
-
-        const int b1 = (a7>>2) + a1;
-        const int b3 =  a3 + (a5>>2);
-        const int b5 = (a3>>2) - a5;
-        const int b7 =  a7 - (a1>>2);
-
-        block[i+0*8] = b0 + b7;
-        block[i+7*8] = b0 - b7;
-        block[i+1*8] = b2 + b5;
-        block[i+6*8] = b2 - b5;
-        block[i+2*8] = b4 + b3;
-        block[i+5*8] = b4 - b3;
-        block[i+3*8] = b6 + b1;
-        block[i+4*8] = b6 - b1;
-    }
-    for( i = 0; i < 8; i++ )
-    {
-        const int a0 =  block[0+i*8] + block[4+i*8];
-        const int a2 =  block[0+i*8] - block[4+i*8];
-        const int a4 = (block[2+i*8]>>1) - block[6+i*8];
-        const int a6 = (block[6+i*8]>>1) + block[2+i*8];
-
-        const int b0 = a0 + a6;
-        const int b2 = a2 + a4;
-        const int b4 = a2 - a4;
-        const int b6 = a0 - a6;
-
-        const int a1 = -block[3+i*8] + block[5+i*8] - block[7+i*8] - (block[7+i*8]>>1);
-        const int a3 =  block[1+i*8] + block[7+i*8] - block[3+i*8] - (block[3+i*8]>>1);
-        const int a5 = -block[1+i*8] + block[7+i*8] + block[5+i*8] + (block[5+i*8]>>1);
-        const int a7 =  block[3+i*8] + block[5+i*8] + block[1+i*8] + (block[1+i*8]>>1);
-
-        const int b1 = (a7>>2) + a1;
-        const int b3 =  a3 + (a5>>2);
-        const int b5 = (a3>>2) - a5;
-        const int b7 =  a7 - (a1>>2);
-
-        dst[i + 0*stride] = CLIP( dst[i + 0*stride] + ((b0 + b7) >> 6) );
-        dst[i + 1*stride] = CLIP( dst[i + 1*stride] + ((b2 + b5) >> 6) );
-        dst[i + 2*stride] = CLIP( dst[i + 2*stride] + ((b4 + b3) >> 6) );
-        dst[i + 3*stride] = CLIP( dst[i + 3*stride] + ((b6 + b1) >> 6) );
-        dst[i + 4*stride] = CLIP( dst[i + 4*stride] + ((b6 - b1) >> 6) );
-        dst[i + 5*stride] = CLIP( dst[i + 5*stride] + ((b4 - b3) >> 6) );
-        dst[i + 6*stride] = CLIP( dst[i + 6*stride] + ((b2 - b5) >> 6) );
-        dst[i + 7*stride] = CLIP( dst[i + 7*stride] + ((b0 - b7) >> 6) );
-    }
-}
-
-// assumes all AC coefs are 0
-void FUNCC(ff_h264_idct_dc_add)(uint8_t *p_dst, DCTELEM *block, int stride){
-    int i, j;
-    int dc = (((dctcoef*)block)[0] + 32) >> 6;
-    INIT_CLIP
-    pixel *dst = (pixel*)p_dst;
-    stride >>= sizeof(pixel)-1;
-    for( j = 0; j < 4; j++ )
-    {
-        for( i = 0; i < 4; i++ )
-            dst[i] = CLIP( dst[i] + dc );
-        dst += stride;
-    }
-}
-
-void FUNCC(ff_h264_idct8_dc_add)(uint8_t *p_dst, DCTELEM *block, int stride){
-    int i, j;
-    int dc = (((dctcoef*)block)[0] + 32) >> 6;
-    INIT_CLIP
-    pixel *dst = (pixel*)p_dst;
-    stride >>= sizeof(pixel)-1;
-    for( j = 0; j < 8; j++ )
-    {
-        for( i = 0; i < 8; i++ )
-            dst[i] = CLIP( dst[i] + dc );
-        dst += stride;
-    }
-}
-
-void FUNCC(ff_h264_idct_add16)(uint8_t *dst, const int *block_offset, DCTELEM *block, int stride, const uint8_t nnzc[6*8]){
-    int i;
-    for(i=0; i<16; i++){
-        int nnz = nnzc[ scan8[i] ];
-        if(nnz){
-            if(nnz==1 && ((dctcoef*)block)[i*16]) FUNCC(ff_h264_idct_dc_add)(dst + block_offset[i], block + i*16*sizeof(pixel), stride);
-            else                                  FUNCC(idct_internal      )(dst + block_offset[i], block + i*16*sizeof(pixel), stride, 4, 6, 1);
-        }
-    }
-}
-
-void FUNCC(ff_h264_idct_add16intra)(uint8_t *dst, const int *block_offset, DCTELEM *block, int stride, const uint8_t nnzc[6*8]){
-    int i;
-    for(i=0; i<16; i++){
-        if(nnzc[ scan8[i] ])             FUNCC(idct_internal      )(dst + block_offset[i], block + i*16*sizeof(pixel), stride, 4, 6, 1);
-        else if(((dctcoef*)block)[i*16]) FUNCC(ff_h264_idct_dc_add)(dst + block_offset[i], block + i*16*sizeof(pixel), stride);
-    }
-}
-
-void FUNCC(ff_h264_idct8_add4)(uint8_t *dst, const int *block_offset, DCTELEM *block, int stride, const uint8_t nnzc[6*8]){
-    int i;
-    for(i=0; i<16; i+=4){
-        int nnz = nnzc[ scan8[i] ];
-        if(nnz){
-            if(nnz==1 && ((dctcoef*)block)[i*16]) FUNCC(ff_h264_idct8_dc_add)(dst + block_offset[i], block + i*16*sizeof(pixel), stride);
-            else                                  FUNCC(ff_h264_idct8_add   )(dst + block_offset[i], block + i*16*sizeof(pixel), stride);
-        }
-    }
-}
-
-void FUNCC(ff_h264_idct_add8)(uint8_t **dest, const int *block_offset, DCTELEM *block, int stride, const uint8_t nnzc[6*8]){
-    int i;
-    for(i=16; i<16+8; i++){
-        if(nnzc[ scan8[i] ])
-            FUNCC(ff_h264_idct_add   )(dest[(i&4)>>2] + block_offset[i], block + i*16*sizeof(pixel), stride);
-        else if(((dctcoef*)block)[i*16])
-            FUNCC(ff_h264_idct_dc_add)(dest[(i&4)>>2] + block_offset[i], block + i*16*sizeof(pixel), stride);
-    }
-}
-/**
- * IDCT transforms the 16 dc values and dequantizes them.
- * @param qp quantization parameter
- */
-void FUNCC(ff_h264_luma_dc_dequant_idct)(DCTELEM *p_output, DCTELEM *p_input, int qmul){
-#define stride 16
-    int i;
-    int temp[16];
-    static const uint8_t x_offset[4]={0, 2*stride, 8*stride, 10*stride};
-    dctcoef *input = (dctcoef*)p_input;
-    dctcoef *output = (dctcoef*)p_output;
-
-    for(i=0; i<4; i++){
-        const int z0= input[4*i+0] + input[4*i+1];
-        const int z1= input[4*i+0] - input[4*i+1];
-        const int z2= input[4*i+2] - input[4*i+3];
-        const int z3= input[4*i+2] + input[4*i+3];
-
-        temp[4*i+0]= z0+z3;
-        temp[4*i+1]= z0-z3;
-        temp[4*i+2]= z1-z2;
-        temp[4*i+3]= z1+z2;
-    }
-
-    for(i=0; i<4; i++){
-        const int offset= x_offset[i];
-        const int z0= temp[4*0+i] + temp[4*2+i];
-        const int z1= temp[4*0+i] - temp[4*2+i];
-        const int z2= temp[4*1+i] - temp[4*3+i];
-        const int z3= temp[4*1+i] + temp[4*3+i];
-
-        output[stride* 0+offset]= ((((z0 + z3)*qmul + 128 ) >> 8));
-        output[stride* 1+offset]= ((((z1 + z2)*qmul + 128 ) >> 8));
-        output[stride* 4+offset]= ((((z1 - z2)*qmul + 128 ) >> 8));
-        output[stride* 5+offset]= ((((z0 - z3)*qmul + 128 ) >> 8));
-    }
-#undef stride
-}
-
-void FUNCC(ff_h264_chroma_dc_dequant_idct)(DCTELEM *p_block, int qmul){
-    const int stride= 16*2;
-    const int xStride= 16;
-    int a,b,c,d,e;
-    dctcoef *block = (dctcoef*)p_block;
-
-    a= block[stride*0 + xStride*0];
-    b= block[stride*0 + xStride*1];
-    c= block[stride*1 + xStride*0];
-    d= block[stride*1 + xStride*1];
-
-    e= a-b;
-    a= a+b;
-    b= c-d;
-    c= c+d;
-
-    block[stride*0 + xStride*0]= ((a+c)*qmul) >> 7;
-    block[stride*0 + xStride*1]= ((e+b)*qmul) >> 7;
-    block[stride*1 + xStride*0]= ((a-c)*qmul) >> 7;
-    block[stride*1 + xStride*1]= ((e-b)*qmul) >> 7;
-}
diff --git a/libavcodec/h264idct_template.c b/libavcodec/h264idct_template.c
index 39c9a1c9eb..d19051df90 100644
--- a/libavcodec/h264idct_template.c
+++ b/libavcodec/h264idct_template.c
@@ -2,20 +2,20 @@
  * H.264 IDCT
  * Copyright (c) 2004-2011 Michael Niedermayer <michaelni@gmx.at>
  *
- * This file is part of Libav.
+ * This file is part of FFmpeg.
  *
- * Libav is free software; you can redistribute it and/or
+ * FFmpeg 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.
  *
- * Libav is distributed in the hope that it will be useful,
+ * FFmpeg 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 Libav; if not, write to the Free Software
+ * License along with FFmpeg; if not, write to the Free Software
  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  */
 
@@ -25,7 +25,7 @@
  * @author Michael Niedermayer <michaelni@gmx.at>
  */
 
-#include "high_bit_depth.h"
+#include "h264_high_depth.h"
 
 #ifndef AVCODEC_H264IDCT_INTERNAL_H
 #define AVCODEC_H264IDCT_INTERNAL_H
@@ -42,12 +42,12 @@ static const uint8_t scan8[16 + 2*4]={
 };
 #endif
 
-static av_always_inline void FUNCC(idct_internal)(uint8_t *_dst, DCTELEM *_block, int stride, int block_stride, int shift, int add){
+static av_always_inline void FUNCC(idct_internal)(uint8_t *p_dst, DCTELEM *p_block, int stride, int block_stride, int shift, int add){
     int i;
     INIT_CLIP
-    pixel *dst = (pixel*)_dst;
-    dctcoef *block = (dctcoef*)_block;
-    stride /= sizeof(pixel);
+    pixel *dst = (pixel*)p_dst;
+    dctcoef *block = (dctcoef*)p_block;
+    stride >>= sizeof(pixel)-1;
 
     block[0] += 1<<(shift-1);
 
@@ -88,12 +88,12 @@ void FUNCC(ff_h264_lowres_idct_put)(uint8_t *dst, int stride, DCTELEM *block){
     FUNCC(idct_internal)(dst, block, stride, 8, 3, 0);
 }
 
-void FUNCC(ff_h264_idct8_add)(uint8_t *_dst, DCTELEM *_block, int stride){
+void FUNCC(ff_h264_idct8_add)(uint8_t *p_dst, DCTELEM *p_block, int stride){
     int i;
     INIT_CLIP
-    pixel *dst = (pixel*)_dst;
-    dctcoef *block = (dctcoef*)_block;
-    stride /= sizeof(pixel);
+    pixel *dst = (pixel*)p_dst;
+    dctcoef *block = (dctcoef*)p_block;
+    stride >>= sizeof(pixel)-1;
 
     block[0] += 32;
 
@@ -162,12 +162,12 @@ void FUNCC(ff_h264_idct8_add)(uint8_t *_dst, DCTELEM *_block, int stride){
 }
 
 // assumes all AC coefs are 0
-void FUNCC(ff_h264_idct_dc_add)(uint8_t *_dst, DCTELEM *block, int stride){
+void FUNCC(ff_h264_idct_dc_add)(uint8_t *p_dst, DCTELEM *block, int stride){
     int i, j;
     int dc = (((dctcoef*)block)[0] + 32) >> 6;
     INIT_CLIP
-    pixel *dst = (pixel*)_dst;
-    stride /= sizeof(pixel);
+    pixel *dst = (pixel*)p_dst;
+    stride >>= sizeof(pixel)-1;
     for( j = 0; j < 4; j++ )
     {
         for( i = 0; i < 4; i++ )
@@ -176,12 +176,12 @@ void FUNCC(ff_h264_idct_dc_add)(uint8_t *_dst, DCTELEM *block, int stride){
     }
 }
 
-void FUNCC(ff_h264_idct8_dc_add)(uint8_t *_dst, DCTELEM *block, int stride){
+void FUNCC(ff_h264_idct8_dc_add)(uint8_t *p_dst, DCTELEM *block, int stride){
     int i, j;
     int dc = (((dctcoef*)block)[0] + 32) >> 6;
     INIT_CLIP
-    pixel *dst = (pixel*)_dst;
-    stride /= sizeof(pixel);
+    pixel *dst = (pixel*)p_dst;
+    stride >>= sizeof(pixel)-1;
     for( j = 0; j < 8; j++ )
     {
         for( i = 0; i < 8; i++ )
@@ -233,13 +233,13 @@ void FUNCC(ff_h264_idct_add8)(uint8_t **dest, const int *block_offset, DCTELEM *
  * IDCT transforms the 16 dc values and dequantizes them.
  * @param qp quantization parameter
  */
-void FUNCC(ff_h264_luma_dc_dequant_idct)(DCTELEM *_output, DCTELEM *_input, int qmul){
+void FUNCC(ff_h264_luma_dc_dequant_idct)(DCTELEM *p_output, DCTELEM *p_input, int qmul){
 #define stride 16
     int i;
     int temp[16];
     static const uint8_t x_offset[4]={0, 2*stride, 8*stride, 10*stride};
-    dctcoef *input = (dctcoef*)_input;
-    dctcoef *output = (dctcoef*)_output;
+    dctcoef *input = (dctcoef*)p_input;
+    dctcoef *output = (dctcoef*)p_output;
 
     for(i=0; i<4; i++){
         const int z0= input[4*i+0] + input[4*i+1];
@@ -268,11 +268,11 @@ void FUNCC(ff_h264_luma_dc_dequant_idct)(DCTELEM *_output, DCTELEM *_input, int
 #undef stride
 }
 
-void FUNCC(ff_h264_chroma_dc_dequant_idct)(DCTELEM *_block, int qmul){
+void FUNCC(ff_h264_chroma_dc_dequant_idct)(DCTELEM *p_block, int qmul){
     const int stride= 16*2;
     const int xStride= 16;
     int a,b,c,d,e;
-    dctcoef *block = (dctcoef*)_block;
+    dctcoef *block = (dctcoef*)p_block;
 
     a= block[stride*0 + xStride*0];
     b= block[stride*0 + xStride*1];
diff --git a/libavcodec/h264pred_internal.h b/libavcodec/h264pred_internal.h
deleted file mode 100644
index a51be6c382..0000000000
--- a/libavcodec/h264pred_internal.h
+++ /dev/null
@@ -1,1339 +0,0 @@
-/*
- * H.26L/H.264/AVC/JVT/14496-10/... encoder/decoder
- * Copyright (c) 2003-2011 Michael Niedermayer <michaelni@gmx.at>
- *
- * This file is part of FFmpeg.
- *
- * FFmpeg 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,
- * 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
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
- */
-
-/**
- * @file
- * H.264 / AVC / MPEG4 part10 prediction functions.
- * @author Michael Niedermayer <michaelni@gmx.at>
- */
-
-#include "mathops.h"
-#include "h264_high_depth.h"
-
-static void FUNCC(pred4x4_vertical)(uint8_t *p_src, const uint8_t *topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    const pixel4 a= ((pixel4*)(src-stride))[0];
-    ((pixel4*)(src+0*stride))[0]= a;
-    ((pixel4*)(src+1*stride))[0]= a;
-    ((pixel4*)(src+2*stride))[0]= a;
-    ((pixel4*)(src+3*stride))[0]= a;
-}
-
-static void FUNCC(pred4x4_horizontal)(uint8_t *p_src, const uint8_t *topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    ((pixel4*)(src+0*stride))[0]= PIXEL_SPLAT_X4(src[-1+0*stride]);
-    ((pixel4*)(src+1*stride))[0]= PIXEL_SPLAT_X4(src[-1+1*stride]);
-    ((pixel4*)(src+2*stride))[0]= PIXEL_SPLAT_X4(src[-1+2*stride]);
-    ((pixel4*)(src+3*stride))[0]= PIXEL_SPLAT_X4(src[-1+3*stride]);
-}
-
-static void FUNCC(pred4x4_dc)(uint8_t *p_src, const uint8_t *topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    const int dc= (  src[-stride] + src[1-stride] + src[2-stride] + src[3-stride]
-                   + src[-1+0*stride] + src[-1+1*stride] + src[-1+2*stride] + src[-1+3*stride] + 4) >>3;
-
-    ((pixel4*)(src+0*stride))[0]=
-    ((pixel4*)(src+1*stride))[0]=
-    ((pixel4*)(src+2*stride))[0]=
-    ((pixel4*)(src+3*stride))[0]= PIXEL_SPLAT_X4(dc);
-}
-
-static void FUNCC(pred4x4_left_dc)(uint8_t *p_src, const uint8_t *topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    const int dc= (  src[-1+0*stride] + src[-1+1*stride] + src[-1+2*stride] + src[-1+3*stride] + 2) >>2;
-
-    ((pixel4*)(src+0*stride))[0]=
-    ((pixel4*)(src+1*stride))[0]=
-    ((pixel4*)(src+2*stride))[0]=
-    ((pixel4*)(src+3*stride))[0]= PIXEL_SPLAT_X4(dc);
-}
-
-static void FUNCC(pred4x4_top_dc)(uint8_t *p_src, const uint8_t *topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    const int dc= (  src[-stride] + src[1-stride] + src[2-stride] + src[3-stride] + 2) >>2;
-
-    ((pixel4*)(src+0*stride))[0]=
-    ((pixel4*)(src+1*stride))[0]=
-    ((pixel4*)(src+2*stride))[0]=
-    ((pixel4*)(src+3*stride))[0]= PIXEL_SPLAT_X4(dc);
-}
-
-static void FUNCC(pred4x4_128_dc)(uint8_t *p_src, const uint8_t *topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    ((pixel4*)(src+0*stride))[0]=
-    ((pixel4*)(src+1*stride))[0]=
-    ((pixel4*)(src+2*stride))[0]=
-    ((pixel4*)(src+3*stride))[0]= PIXEL_SPLAT_X4(1<<(BIT_DEPTH-1));
-}
-
-static void FUNCC(pred4x4_127_dc)(uint8_t *p_src, const uint8_t *topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    ((pixel4*)(src+0*stride))[0]=
-    ((pixel4*)(src+1*stride))[0]=
-    ((pixel4*)(src+2*stride))[0]=
-    ((pixel4*)(src+3*stride))[0]= PIXEL_SPLAT_X4((1<<(BIT_DEPTH-1))-1);
-}
-
-static void FUNCC(pred4x4_129_dc)(uint8_t *p_src, const uint8_t *topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    ((pixel4*)(src+0*stride))[0]=
-    ((pixel4*)(src+1*stride))[0]=
-    ((pixel4*)(src+2*stride))[0]=
-    ((pixel4*)(src+3*stride))[0]= PIXEL_SPLAT_X4((1<<(BIT_DEPTH-1))+1);
-}
-
-
-#define LOAD_TOP_RIGHT_EDGE\
-    const int av_unused t4= topright[0];\
-    const int av_unused t5= topright[1];\
-    const int av_unused t6= topright[2];\
-    const int av_unused t7= topright[3];\
-
-#define LOAD_DOWN_LEFT_EDGE\
-    const int av_unused l4= src[-1+4*stride];\
-    const int av_unused l5= src[-1+5*stride];\
-    const int av_unused l6= src[-1+6*stride];\
-    const int av_unused l7= src[-1+7*stride];\
-
-#define LOAD_LEFT_EDGE\
-    const int av_unused l0= src[-1+0*stride];\
-    const int av_unused l1= src[-1+1*stride];\
-    const int av_unused l2= src[-1+2*stride];\
-    const int av_unused l3= src[-1+3*stride];\
-
-#define LOAD_TOP_EDGE\
-    const int av_unused t0= src[ 0-1*stride];\
-    const int av_unused t1= src[ 1-1*stride];\
-    const int av_unused t2= src[ 2-1*stride];\
-    const int av_unused t3= src[ 3-1*stride];\
-
-static void FUNCC(pred4x4_vertical_vp8)(uint8_t *p_src, const uint8_t *p_topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    const pixel *topright = (const pixel*)p_topright;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    const int lt= src[-1-1*stride];
-    LOAD_TOP_EDGE
-    LOAD_TOP_RIGHT_EDGE
-    pixel4 v = PACK_4U8((lt + 2*t0 + t1 + 2) >> 2,
-                            (t0 + 2*t1 + t2 + 2) >> 2,
-                            (t1 + 2*t2 + t3 + 2) >> 2,
-                            (t2 + 2*t3 + t4 + 2) >> 2);
-
-    AV_WN4PA(src+0*stride, v);
-    AV_WN4PA(src+1*stride, v);
-    AV_WN4PA(src+2*stride, v);
-    AV_WN4PA(src+3*stride, v);
-}
-
-static void FUNCC(pred4x4_horizontal_vp8)(uint8_t *p_src, const uint8_t *topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    const int lt= src[-1-1*stride];
-    LOAD_LEFT_EDGE
-
-    AV_WN4PA(src+0*stride, PIXEL_SPLAT_X4((lt + 2*l0 + l1 + 2) >> 2));
-    AV_WN4PA(src+1*stride, PIXEL_SPLAT_X4((l0 + 2*l1 + l2 + 2) >> 2));
-    AV_WN4PA(src+2*stride, PIXEL_SPLAT_X4((l1 + 2*l2 + l3 + 2) >> 2));
-    AV_WN4PA(src+3*stride, PIXEL_SPLAT_X4((l2 + 2*l3 + l3 + 2) >> 2));
-}
-
-static void FUNCC(pred4x4_down_right)(uint8_t *p_src, const uint8_t *topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    const int lt= src[-1-1*stride];
-    LOAD_TOP_EDGE
-    LOAD_LEFT_EDGE
-
-    src[0+3*stride]=(l3 + 2*l2 + l1 + 2)>>2;
-    src[0+2*stride]=
-    src[1+3*stride]=(l2 + 2*l1 + l0 + 2)>>2;
-    src[0+1*stride]=
-    src[1+2*stride]=
-    src[2+3*stride]=(l1 + 2*l0 + lt + 2)>>2;
-    src[0+0*stride]=
-    src[1+1*stride]=
-    src[2+2*stride]=
-    src[3+3*stride]=(l0 + 2*lt + t0 + 2)>>2;
-    src[1+0*stride]=
-    src[2+1*stride]=
-    src[3+2*stride]=(lt + 2*t0 + t1 + 2)>>2;
-    src[2+0*stride]=
-    src[3+1*stride]=(t0 + 2*t1 + t2 + 2)>>2;
-    src[3+0*stride]=(t1 + 2*t2 + t3 + 2)>>2;
-}
-
-static void FUNCC(pred4x4_down_left)(uint8_t *p_src, const uint8_t *p_topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    const pixel *topright = (const pixel*)p_topright;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    LOAD_TOP_EDGE
-    LOAD_TOP_RIGHT_EDGE
-//    LOAD_LEFT_EDGE
-
-    src[0+0*stride]=(t0 + t2 + 2*t1 + 2)>>2;
-    src[1+0*stride]=
-    src[0+1*stride]=(t1 + t3 + 2*t2 + 2)>>2;
-    src[2+0*stride]=
-    src[1+1*stride]=
-    src[0+2*stride]=(t2 + t4 + 2*t3 + 2)>>2;
-    src[3+0*stride]=
-    src[2+1*stride]=
-    src[1+2*stride]=
-    src[0+3*stride]=(t3 + t5 + 2*t4 + 2)>>2;
-    src[3+1*stride]=
-    src[2+2*stride]=
-    src[1+3*stride]=(t4 + t6 + 2*t5 + 2)>>2;
-    src[3+2*stride]=
-    src[2+3*stride]=(t5 + t7 + 2*t6 + 2)>>2;
-    src[3+3*stride]=(t6 + 3*t7 + 2)>>2;
-}
-
-static void FUNCC(pred4x4_down_left_svq3)(uint8_t *p_src, const uint8_t *topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    LOAD_TOP_EDGE
-    LOAD_LEFT_EDGE
-    const av_unused int unu0= t0;
-    const av_unused int unu1= l0;
-
-    src[0+0*stride]=(l1 + t1)>>1;
-    src[1+0*stride]=
-    src[0+1*stride]=(l2 + t2)>>1;
-    src[2+0*stride]=
-    src[1+1*stride]=
-    src[0+2*stride]=
-    src[3+0*stride]=
-    src[2+1*stride]=
-    src[1+2*stride]=
-    src[0+3*stride]=
-    src[3+1*stride]=
-    src[2+2*stride]=
-    src[1+3*stride]=
-    src[3+2*stride]=
-    src[2+3*stride]=
-    src[3+3*stride]=(l3 + t3)>>1;
-}
-
-static void FUNCC(pred4x4_down_left_rv40)(uint8_t *p_src, const uint8_t *p_topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    const pixel *topright = (const pixel*)p_topright;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    LOAD_TOP_EDGE
-    LOAD_TOP_RIGHT_EDGE
-    LOAD_LEFT_EDGE
-    LOAD_DOWN_LEFT_EDGE
-
-    src[0+0*stride]=(t0 + t2 + 2*t1 + 2 + l0 + l2 + 2*l1 + 2)>>3;
-    src[1+0*stride]=
-    src[0+1*stride]=(t1 + t3 + 2*t2 + 2 + l1 + l3 + 2*l2 + 2)>>3;
-    src[2+0*stride]=
-    src[1+1*stride]=
-    src[0+2*stride]=(t2 + t4 + 2*t3 + 2 + l2 + l4 + 2*l3 + 2)>>3;
-    src[3+0*stride]=
-    src[2+1*stride]=
-    src[1+2*stride]=
-    src[0+3*stride]=(t3 + t5 + 2*t4 + 2 + l3 + l5 + 2*l4 + 2)>>3;
-    src[3+1*stride]=
-    src[2+2*stride]=
-    src[1+3*stride]=(t4 + t6 + 2*t5 + 2 + l4 + l6 + 2*l5 + 2)>>3;
-    src[3+2*stride]=
-    src[2+3*stride]=(t5 + t7 + 2*t6 + 2 + l5 + l7 + 2*l6 + 2)>>3;
-    src[3+3*stride]=(t6 + t7 + 1 + l6 + l7 + 1)>>2;
-}
-
-static void FUNCC(pred4x4_down_left_rv40_nodown)(uint8_t *p_src, const uint8_t *p_topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    const pixel *topright = (const pixel*)p_topright;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    LOAD_TOP_EDGE
-    LOAD_TOP_RIGHT_EDGE
-    LOAD_LEFT_EDGE
-
-    src[0+0*stride]=(t0 + t2 + 2*t1 + 2 + l0 + l2 + 2*l1 + 2)>>3;
-    src[1+0*stride]=
-    src[0+1*stride]=(t1 + t3 + 2*t2 + 2 + l1 + l3 + 2*l2 + 2)>>3;
-    src[2+0*stride]=
-    src[1+1*stride]=
-    src[0+2*stride]=(t2 + t4 + 2*t3 + 2 + l2 + 3*l3 + 2)>>3;
-    src[3+0*stride]=
-    src[2+1*stride]=
-    src[1+2*stride]=
-    src[0+3*stride]=(t3 + t5 + 2*t4 + 2 + l3*4 + 2)>>3;
-    src[3+1*stride]=
-    src[2+2*stride]=
-    src[1+3*stride]=(t4 + t6 + 2*t5 + 2 + l3*4 + 2)>>3;
-    src[3+2*stride]=
-    src[2+3*stride]=(t5 + t7 + 2*t6 + 2 + l3*4 + 2)>>3;
-    src[3+3*stride]=(t6 + t7 + 1 + 2*l3 + 1)>>2;
-}
-
-static void FUNCC(pred4x4_vertical_right)(uint8_t *p_src, const uint8_t *topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    const int lt= src[-1-1*stride];
-    LOAD_TOP_EDGE
-    LOAD_LEFT_EDGE
-
-    src[0+0*stride]=
-    src[1+2*stride]=(lt + t0 + 1)>>1;
-    src[1+0*stride]=
-    src[2+2*stride]=(t0 + t1 + 1)>>1;
-    src[2+0*stride]=
-    src[3+2*stride]=(t1 + t2 + 1)>>1;
-    src[3+0*stride]=(t2 + t3 + 1)>>1;
-    src[0+1*stride]=
-    src[1+3*stride]=(l0 + 2*lt + t0 + 2)>>2;
-    src[1+1*stride]=
-    src[2+3*stride]=(lt + 2*t0 + t1 + 2)>>2;
-    src[2+1*stride]=
-    src[3+3*stride]=(t0 + 2*t1 + t2 + 2)>>2;
-    src[3+1*stride]=(t1 + 2*t2 + t3 + 2)>>2;
-    src[0+2*stride]=(lt + 2*l0 + l1 + 2)>>2;
-    src[0+3*stride]=(l0 + 2*l1 + l2 + 2)>>2;
-}
-
-static void FUNCC(pred4x4_vertical_left)(uint8_t *p_src, const uint8_t *p_topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    const pixel *topright = (const pixel*)p_topright;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    LOAD_TOP_EDGE
-    LOAD_TOP_RIGHT_EDGE
-
-    src[0+0*stride]=(t0 + t1 + 1)>>1;
-    src[1+0*stride]=
-    src[0+2*stride]=(t1 + t2 + 1)>>1;
-    src[2+0*stride]=
-    src[1+2*stride]=(t2 + t3 + 1)>>1;
-    src[3+0*stride]=
-    src[2+2*stride]=(t3 + t4+ 1)>>1;
-    src[3+2*stride]=(t4 + t5+ 1)>>1;
-    src[0+1*stride]=(t0 + 2*t1 + t2 + 2)>>2;
-    src[1+1*stride]=
-    src[0+3*stride]=(t1 + 2*t2 + t3 + 2)>>2;
-    src[2+1*stride]=
-    src[1+3*stride]=(t2 + 2*t3 + t4 + 2)>>2;
-    src[3+1*stride]=
-    src[2+3*stride]=(t3 + 2*t4 + t5 + 2)>>2;
-    src[3+3*stride]=(t4 + 2*t5 + t6 + 2)>>2;
-}
-
-static void FUNCC(pred4x4_vertical_left_rv40_internal)(uint8_t *p_src, const uint8_t *p_topright, int p_stride,
-                                      const int l0, const int l1, const int l2, const int l3, const int l4){
-    pixel *src = (pixel*)p_src;
-    const pixel *topright = (const pixel*)p_topright;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    LOAD_TOP_EDGE
-    LOAD_TOP_RIGHT_EDGE
-
-    src[0+0*stride]=(2*t0 + 2*t1 + l1 + 2*l2 + l3 + 4)>>3;
-    src[1+0*stride]=
-    src[0+2*stride]=(t1 + t2 + 1)>>1;
-    src[2+0*stride]=
-    src[1+2*stride]=(t2 + t3 + 1)>>1;
-    src[3+0*stride]=
-    src[2+2*stride]=(t3 + t4+ 1)>>1;
-    src[3+2*stride]=(t4 + t5+ 1)>>1;
-    src[0+1*stride]=(t0 + 2*t1 + t2 + l2 + 2*l3 + l4 + 4)>>3;
-    src[1+1*stride]=
-    src[0+3*stride]=(t1 + 2*t2 + t3 + 2)>>2;
-    src[2+1*stride]=
-    src[1+3*stride]=(t2 + 2*t3 + t4 + 2)>>2;
-    src[3+1*stride]=
-    src[2+3*stride]=(t3 + 2*t4 + t5 + 2)>>2;
-    src[3+3*stride]=(t4 + 2*t5 + t6 + 2)>>2;
-}
-
-static void FUNCC(pred4x4_vertical_left_rv40)(uint8_t *p_src, const uint8_t *topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    LOAD_LEFT_EDGE
-    LOAD_DOWN_LEFT_EDGE
-
-    FUNCC(pred4x4_vertical_left_rv40_internal)(p_src, topright, p_stride, l0, l1, l2, l3, l4);
-}
-
-static void FUNCC(pred4x4_vertical_left_rv40_nodown)(uint8_t *p_src, const uint8_t *topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    LOAD_LEFT_EDGE
-
-    FUNCC(pred4x4_vertical_left_rv40_internal)(p_src, topright, p_stride, l0, l1, l2, l3, l3);
-}
-
-static void FUNCC(pred4x4_vertical_left_vp8)(uint8_t *p_src, const uint8_t *p_topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    const pixel *topright = (const pixel*)p_topright;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    LOAD_TOP_EDGE
-    LOAD_TOP_RIGHT_EDGE
-
-    src[0+0*stride]=(t0 + t1 + 1)>>1;
-    src[1+0*stride]=
-    src[0+2*stride]=(t1 + t2 + 1)>>1;
-    src[2+0*stride]=
-    src[1+2*stride]=(t2 + t3 + 1)>>1;
-    src[3+0*stride]=
-    src[2+2*stride]=(t3 + t4 + 1)>>1;
-    src[0+1*stride]=(t0 + 2*t1 + t2 + 2)>>2;
-    src[1+1*stride]=
-    src[0+3*stride]=(t1 + 2*t2 + t3 + 2)>>2;
-    src[2+1*stride]=
-    src[1+3*stride]=(t2 + 2*t3 + t4 + 2)>>2;
-    src[3+1*stride]=
-    src[2+3*stride]=(t3 + 2*t4 + t5 + 2)>>2;
-    src[3+2*stride]=(t4 + 2*t5 + t6 + 2)>>2;
-    src[3+3*stride]=(t5 + 2*t6 + t7 + 2)>>2;
-}
-
-static void FUNCC(pred4x4_horizontal_up)(uint8_t *p_src, const uint8_t *topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    LOAD_LEFT_EDGE
-
-    src[0+0*stride]=(l0 + l1 + 1)>>1;
-    src[1+0*stride]=(l0 + 2*l1 + l2 + 2)>>2;
-    src[2+0*stride]=
-    src[0+1*stride]=(l1 + l2 + 1)>>1;
-    src[3+0*stride]=
-    src[1+1*stride]=(l1 + 2*l2 + l3 + 2)>>2;
-    src[2+1*stride]=
-    src[0+2*stride]=(l2 + l3 + 1)>>1;
-    src[3+1*stride]=
-    src[1+2*stride]=(l2 + 2*l3 + l3 + 2)>>2;
-    src[3+2*stride]=
-    src[1+3*stride]=
-    src[0+3*stride]=
-    src[2+2*stride]=
-    src[2+3*stride]=
-    src[3+3*stride]=l3;
-}
-
-static void FUNCC(pred4x4_horizontal_up_rv40)(uint8_t *p_src, const uint8_t *p_topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    const pixel *topright = (const pixel*)p_topright;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    LOAD_LEFT_EDGE
-    LOAD_DOWN_LEFT_EDGE
-    LOAD_TOP_EDGE
-    LOAD_TOP_RIGHT_EDGE
-
-    src[0+0*stride]=(t1 + 2*t2 + t3 + 2*l0 + 2*l1 + 4)>>3;
-    src[1+0*stride]=(t2 + 2*t3 + t4 + l0 + 2*l1 + l2 + 4)>>3;
-    src[2+0*stride]=
-    src[0+1*stride]=(t3 + 2*t4 + t5 + 2*l1 + 2*l2 + 4)>>3;
-    src[3+0*stride]=
-    src[1+1*stride]=(t4 + 2*t5 + t6 + l1 + 2*l2 + l3 + 4)>>3;
-    src[2+1*stride]=
-    src[0+2*stride]=(t5 + 2*t6 + t7 + 2*l2 + 2*l3 + 4)>>3;
-    src[3+1*stride]=
-    src[1+2*stride]=(t6 + 3*t7 + l2 + 3*l3 + 4)>>3;
-    src[3+2*stride]=
-    src[1+3*stride]=(l3 + 2*l4 + l5 + 2)>>2;
-    src[0+3*stride]=
-    src[2+2*stride]=(t6 + t7 + l3 + l4 + 2)>>2;
-    src[2+3*stride]=(l4 + l5 + 1)>>1;
-    src[3+3*stride]=(l4 + 2*l5 + l6 + 2)>>2;
-}
-
-static void FUNCC(pred4x4_horizontal_up_rv40_nodown)(uint8_t *p_src, const uint8_t *p_topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    const pixel *topright = (const pixel*)p_topright;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    LOAD_LEFT_EDGE
-    LOAD_TOP_EDGE
-    LOAD_TOP_RIGHT_EDGE
-
-    src[0+0*stride]=(t1 + 2*t2 + t3 + 2*l0 + 2*l1 + 4)>>3;
-    src[1+0*stride]=(t2 + 2*t3 + t4 + l0 + 2*l1 + l2 + 4)>>3;
-    src[2+0*stride]=
-    src[0+1*stride]=(t3 + 2*t4 + t5 + 2*l1 + 2*l2 + 4)>>3;
-    src[3+0*stride]=
-    src[1+1*stride]=(t4 + 2*t5 + t6 + l1 + 2*l2 + l3 + 4)>>3;
-    src[2+1*stride]=
-    src[0+2*stride]=(t5 + 2*t6 + t7 + 2*l2 + 2*l3 + 4)>>3;
-    src[3+1*stride]=
-    src[1+2*stride]=(t6 + 3*t7 + l2 + 3*l3 + 4)>>3;
-    src[3+2*stride]=
-    src[1+3*stride]=l3;
-    src[0+3*stride]=
-    src[2+2*stride]=(t6 + t7 + 2*l3 + 2)>>2;
-    src[2+3*stride]=
-    src[3+3*stride]=l3;
-}
-
-static void FUNCC(pred4x4_horizontal_down)(uint8_t *p_src, const uint8_t *topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    const int lt= src[-1-1*stride];
-    LOAD_TOP_EDGE
-    LOAD_LEFT_EDGE
-
-    src[0+0*stride]=
-    src[2+1*stride]=(lt + l0 + 1)>>1;
-    src[1+0*stride]=
-    src[3+1*stride]=(l0 + 2*lt + t0 + 2)>>2;
-    src[2+0*stride]=(lt + 2*t0 + t1 + 2)>>2;
-    src[3+0*stride]=(t0 + 2*t1 + t2 + 2)>>2;
-    src[0+1*stride]=
-    src[2+2*stride]=(l0 + l1 + 1)>>1;
-    src[1+1*stride]=
-    src[3+2*stride]=(lt + 2*l0 + l1 + 2)>>2;
-    src[0+2*stride]=
-    src[2+3*stride]=(l1 + l2+ 1)>>1;
-    src[1+2*stride]=
-    src[3+3*stride]=(l0 + 2*l1 + l2 + 2)>>2;
-    src[0+3*stride]=(l2 + l3 + 1)>>1;
-    src[1+3*stride]=(l1 + 2*l2 + l3 + 2)>>2;
-}
-
-static void FUNCC(pred4x4_tm_vp8)(uint8_t *p_src, const uint8_t *topright, int p_stride){
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    uint8_t *cm = ff_cropTbl + MAX_NEG_CROP - src[-1-stride];
-    pixel *top = src-stride;
-    int y;
-
-    for (y = 0; y < 4; y++) {
-        uint8_t *cm_in = cm + src[-1];
-        src[0] = cm_in[top[0]];
-        src[1] = cm_in[top[1]];
-        src[2] = cm_in[top[2]];
-        src[3] = cm_in[top[3]];
-        src += stride;
-    }
-}
-
-static void FUNCC(pred16x16_vertical)(uint8_t *p_src, int p_stride){
-    int i;
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    const pixel4 a = ((pixel4*)(src-stride))[0];
-    const pixel4 b = ((pixel4*)(src-stride))[1];
-    const pixel4 c = ((pixel4*)(src-stride))[2];
-    const pixel4 d = ((pixel4*)(src-stride))[3];
-
-    for(i=0; i<16; i++){
-        ((pixel4*)(src+i*stride))[0] = a;
-        ((pixel4*)(src+i*stride))[1] = b;
-        ((pixel4*)(src+i*stride))[2] = c;
-        ((pixel4*)(src+i*stride))[3] = d;
-    }
-}
-
-static void FUNCC(pred16x16_horizontal)(uint8_t *p_src, int stride){
-    int i;
-    pixel *src = (pixel*)p_src;
-    stride >>= sizeof(pixel)-1;
-
-    for(i=0; i<16; i++){
-        ((pixel4*)(src+i*stride))[0] =
-        ((pixel4*)(src+i*stride))[1] =
-        ((pixel4*)(src+i*stride))[2] =
-        ((pixel4*)(src+i*stride))[3] = PIXEL_SPLAT_X4(src[-1+i*stride]);
-    }
-}
-
-#define PREDICT_16x16_DC(v)\
-    for(i=0; i<16; i++){\
-        AV_WN4P(src+ 0, v);\
-        AV_WN4P(src+ 4, v);\
-        AV_WN4P(src+ 8, v);\
-        AV_WN4P(src+12, v);\
-        src += stride;\
-    }
-
-static void FUNCC(pred16x16_dc)(uint8_t *p_src, int stride){
-    int i, dc=0;
-    pixel *src = (pixel*)p_src;
-    pixel4 dcsplat;
-    stride >>= sizeof(pixel)-1;
-
-    for(i=0;i<16; i++){
-        dc+= src[-1+i*stride];
-    }
-
-    for(i=0;i<16; i++){
-        dc+= src[i-stride];
-    }
-
-    dcsplat = PIXEL_SPLAT_X4((dc+16)>>5);
-    PREDICT_16x16_DC(dcsplat);
-}
-
-static void FUNCC(pred16x16_left_dc)(uint8_t *p_src, int stride){
-    int i, dc=0;
-    pixel *src = (pixel*)p_src;
-    pixel4 dcsplat;
-    stride >>= sizeof(pixel)-1;
-
-    for(i=0;i<16; i++){
-        dc+= src[-1+i*stride];
-    }
-
-    dcsplat = PIXEL_SPLAT_X4((dc+8)>>4);
-    PREDICT_16x16_DC(dcsplat);
-}
-
-static void FUNCC(pred16x16_top_dc)(uint8_t *p_src, int stride){
-    int i, dc=0;
-    pixel *src = (pixel*)p_src;
-    pixel4 dcsplat;
-    stride >>= sizeof(pixel)-1;
-
-    for(i=0;i<16; i++){
-        dc+= src[i-stride];
-    }
-
-    dcsplat = PIXEL_SPLAT_X4((dc+8)>>4);
-    PREDICT_16x16_DC(dcsplat);
-}
-
-#define PRED16x16_X(n, v) \
-static void FUNCC(pred16x16_##n##_dc)(uint8_t *p_src, int stride){\
-    int i;\
-    pixel *src = (pixel*)p_src;\
-    stride >>= sizeof(pixel)-1;\
-    PREDICT_16x16_DC(PIXEL_SPLAT_X4(v));\
-}
-
-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 *p_src, int p_stride, const int svq3, const int rv40){
-  int i, j, k;
-  int a;
-  INIT_CLIP
-  pixel *src = (pixel*)p_src;
-  int stride = p_stride>>(sizeof(pixel)-1);
-  const pixel * const src0 = src +7-stride;
-  const pixel *       src1 = src +8*stride-1;
-  const pixel *       src2 = src1-2*stride;    // == src+6*stride-1;
-  int H = src0[1] - src0[-1];
-  int V = src1[0] - src2[ 0];
-  for(k=2; k<=8; ++k) {
-    src1 += stride; src2 -= stride;
-    H += k*(src0[k] - src0[-k]);
-    V += k*(src1[0] - src2[ 0]);
-  }
-  if(svq3){
-    H = ( 5*(H/4) ) / 16;
-    V = ( 5*(V/4) ) / 16;
-
-    /* required for 100% accuracy */
-    i = H; H = V; V = i;
-  }else if(rv40){
-    H = ( H + (H>>2) ) >> 4;
-    V = ( V + (V>>2) ) >> 4;
-  }else{
-    H = ( 5*H+32 ) >> 6;
-    V = ( 5*V+32 ) >> 6;
-  }
-
-  a = 16*(src1[0] + src2[16] + 1) - 7*(V+H);
-  for(j=16; j>0; --j) {
-    int b = a;
-    a += V;
-    for(i=-16; i<0; i+=4) {
-      src[16+i] = CLIP((b    ) >> 5);
-      src[17+i] = CLIP((b+  H) >> 5);
-      src[18+i] = CLIP((b+2*H) >> 5);
-      src[19+i] = CLIP((b+3*H) >> 5);
-      b += 4*H;
-    }
-    src += stride;
-  }
-}
-
-static void FUNCC(pred16x16_plane)(uint8_t *src, int stride){
-    FUNCC(pred16x16_plane_compat)(src, stride, 0, 0);
-}
-
-static void FUNCC(pred16x16_plane_svq3)(uint8_t *src, int stride){
-    FUNCC(pred16x16_plane_compat)(src, stride, 1, 0);
-}
-
-static void FUNCC(pred16x16_plane_rv40)(uint8_t *src, int stride){
-    FUNCC(pred16x16_plane_compat)(src, stride, 0, 1);
-}
-
-static void FUNCC(pred16x16_tm_vp8)(uint8_t *src, int stride){
-    uint8_t *cm = ff_cropTbl + MAX_NEG_CROP - src[-1-stride];
-    uint8_t *top = src-stride;
-    int y;
-
-    for (y = 0; y < 16; y++) {
-        uint8_t *cm_in = cm + src[-1];
-        src[0]  = cm_in[top[0]];
-        src[1]  = cm_in[top[1]];
-        src[2]  = cm_in[top[2]];
-        src[3]  = cm_in[top[3]];
-        src[4]  = cm_in[top[4]];
-        src[5]  = cm_in[top[5]];
-        src[6]  = cm_in[top[6]];
-        src[7]  = cm_in[top[7]];
-        src[8]  = cm_in[top[8]];
-        src[9]  = cm_in[top[9]];
-        src[10] = cm_in[top[10]];
-        src[11] = cm_in[top[11]];
-        src[12] = cm_in[top[12]];
-        src[13] = cm_in[top[13]];
-        src[14] = cm_in[top[14]];
-        src[15] = cm_in[top[15]];
-        src += stride;
-    }
-}
-
-static void FUNCC(pred8x8_vertical)(uint8_t *p_src, int p_stride){
-    int i;
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    const pixel4 a= ((pixel4*)(src-stride))[0];
-    const pixel4 b= ((pixel4*)(src-stride))[1];
-
-    for(i=0; i<8; i++){
-        ((pixel4*)(src+i*stride))[0]= a;
-        ((pixel4*)(src+i*stride))[1]= b;
-    }
-}
-
-static void FUNCC(pred8x8_horizontal)(uint8_t *p_src, int stride){
-    int i;
-    pixel *src = (pixel*)p_src;
-    stride >>= sizeof(pixel)-1;
-
-    for(i=0; i<8; i++){
-        ((pixel4*)(src+i*stride))[0]=
-        ((pixel4*)(src+i*stride))[1]= PIXEL_SPLAT_X4(src[-1+i*stride]);
-    }
-}
-
-#define PRED8x8_X(n, v)\
-static void FUNCC(pred8x8_##n##_dc)(uint8_t *p_src, int stride){\
-    int i;\
-    pixel *src = (pixel*)p_src;\
-    stride >>= sizeof(pixel)-1;\
-    for(i=0; i<8; i++){\
-        ((pixel4*)(src+i*stride))[0]=\
-        ((pixel4*)(src+i*stride))[1]= PIXEL_SPLAT_X4(v);\
-    }\
-}
-
-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(pred8x8_left_dc)(uint8_t *p_src, int stride){
-    int i;
-    int dc0, dc2;
-    pixel4 dc0splat, dc2splat;
-    pixel *src = (pixel*)p_src;
-    stride >>= sizeof(pixel)-1;
-
-    dc0=dc2=0;
-    for(i=0;i<4; i++){
-        dc0+= src[-1+i*stride];
-        dc2+= src[-1+(i+4)*stride];
-    }
-    dc0splat = PIXEL_SPLAT_X4((dc0 + 2)>>2);
-    dc2splat = PIXEL_SPLAT_X4((dc2 + 2)>>2);
-
-    for(i=0; i<4; i++){
-        ((pixel4*)(src+i*stride))[0]=
-        ((pixel4*)(src+i*stride))[1]= dc0splat;
-    }
-    for(i=4; i<8; i++){
-        ((pixel4*)(src+i*stride))[0]=
-        ((pixel4*)(src+i*stride))[1]= dc2splat;
-    }
-}
-
-static void FUNCC(pred8x8_left_dc_rv40)(uint8_t *p_src, int stride){
-    int i;
-    int dc0;
-    pixel4 dc0splat;
-    pixel *src = (pixel*)p_src;
-    stride >>= sizeof(pixel)-1;
-
-    dc0=0;
-    for(i=0;i<8; i++)
-        dc0+= src[-1+i*stride];
-    dc0splat = PIXEL_SPLAT_X4((dc0 + 4)>>3);
-
-    for(i=0; i<8; i++){
-        ((pixel4*)(src+i*stride))[0]=
-        ((pixel4*)(src+i*stride))[1]= dc0splat;
-    }
-}
-
-static void FUNCC(pred8x8_top_dc)(uint8_t *p_src, int stride){
-    int i;
-    int dc0, dc1;
-    pixel4 dc0splat, dc1splat;
-    pixel *src = (pixel*)p_src;
-    stride >>= sizeof(pixel)-1;
-
-    dc0=dc1=0;
-    for(i=0;i<4; i++){
-        dc0+= src[i-stride];
-        dc1+= src[4+i-stride];
-    }
-    dc0splat = PIXEL_SPLAT_X4((dc0 + 2)>>2);
-    dc1splat = PIXEL_SPLAT_X4((dc1 + 2)>>2);
-
-    for(i=0; i<4; i++){
-        ((pixel4*)(src+i*stride))[0]= dc0splat;
-        ((pixel4*)(src+i*stride))[1]= dc1splat;
-    }
-    for(i=4; i<8; i++){
-        ((pixel4*)(src+i*stride))[0]= dc0splat;
-        ((pixel4*)(src+i*stride))[1]= dc1splat;
-    }
-}
-
-static void FUNCC(pred8x8_top_dc_rv40)(uint8_t *p_src, int stride){
-    int i;
-    int dc0;
-    pixel4 dc0splat;
-    pixel *src = (pixel*)p_src;
-    stride >>= sizeof(pixel)-1;
-
-    dc0=0;
-    for(i=0;i<8; i++)
-        dc0+= src[i-stride];
-    dc0splat = PIXEL_SPLAT_X4((dc0 + 4)>>3);
-
-    for(i=0; i<8; i++){
-        ((pixel4*)(src+i*stride))[0]=
-        ((pixel4*)(src+i*stride))[1]= dc0splat;
-    }
-}
-
-
-static void FUNCC(pred8x8_dc)(uint8_t *p_src, int stride){
-    int i;
-    int dc0, dc1, dc2;
-    pixel4 dc0splat, dc1splat, dc2splat, dc3splat;
-    pixel *src = (pixel*)p_src;
-    stride >>= sizeof(pixel)-1;
-
-    dc0=dc1=dc2=0;
-    for(i=0;i<4; i++){
-        dc0+= src[-1+i*stride] + src[i-stride];
-        dc1+= src[4+i-stride];
-        dc2+= src[-1+(i+4)*stride];
-    }
-    dc0splat = PIXEL_SPLAT_X4((dc0 + 4)>>3);
-    dc1splat = PIXEL_SPLAT_X4((dc1 + 2)>>2);
-    dc2splat = PIXEL_SPLAT_X4((dc2 + 2)>>2);
-    dc3splat = PIXEL_SPLAT_X4((dc1 + dc2 + 4)>>3);
-
-    for(i=0; i<4; i++){
-        ((pixel4*)(src+i*stride))[0]= dc0splat;
-        ((pixel4*)(src+i*stride))[1]= dc1splat;
-    }
-    for(i=4; i<8; i++){
-        ((pixel4*)(src+i*stride))[0]= dc2splat;
-        ((pixel4*)(src+i*stride))[1]= dc3splat;
-    }
-}
-
-//the following 4 function should not be optimized!
-static void FUNC(pred8x8_mad_cow_dc_l0t)(uint8_t *src, int stride){
-    FUNCC(pred8x8_top_dc)(src, stride);
-    FUNCC(pred4x4_dc)(src, NULL, stride);
-}
-
-static void FUNC(pred8x8_mad_cow_dc_0lt)(uint8_t *src, int stride){
-    FUNCC(pred8x8_dc)(src, stride);
-    FUNCC(pred4x4_top_dc)(src, NULL, stride);
-}
-
-static void FUNC(pred8x8_mad_cow_dc_l00)(uint8_t *src, int 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(pred8x8_mad_cow_dc_0l0)(uint8_t *src, int 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 FUNCC(pred8x8_dc_rv40)(uint8_t *p_src, int stride){
-    int i;
-    int dc0=0;
-    pixel4 dc0splat;
-    pixel *src = (pixel*)p_src;
-    stride >>= sizeof(pixel)-1;
-
-    for(i=0;i<4; i++){
-        dc0+= src[-1+i*stride] + src[i-stride];
-        dc0+= src[4+i-stride];
-        dc0+= src[-1+(i+4)*stride];
-    }
-    dc0splat = PIXEL_SPLAT_X4((dc0 + 8)>>4);
-
-    for(i=0; i<4; i++){
-        ((pixel4*)(src+i*stride))[0]= dc0splat;
-        ((pixel4*)(src+i*stride))[1]= dc0splat;
-    }
-    for(i=4; i<8; i++){
-        ((pixel4*)(src+i*stride))[0]= dc0splat;
-        ((pixel4*)(src+i*stride))[1]= dc0splat;
-    }
-}
-
-static void FUNCC(pred8x8_plane)(uint8_t *p_src, int p_stride){
-  int j, k;
-  int a;
-  INIT_CLIP
-  pixel *src = (pixel*)p_src;
-  int stride = p_stride>>(sizeof(pixel)-1);
-  const pixel * const src0 = src +3-stride;
-  const pixel *       src1 = src +4*stride-1;
-  const pixel *       src2 = src1-2*stride;    // == src+2*stride-1;
-  int H = src0[1] - src0[-1];
-  int V = src1[0] - src2[ 0];
-  for(k=2; k<=4; ++k) {
-    src1 += stride; src2 -= stride;
-    H += k*(src0[k] - src0[-k]);
-    V += k*(src1[0] - src2[ 0]);
-  }
-  H = ( 17*H+16 ) >> 5;
-  V = ( 17*V+16 ) >> 5;
-
-  a = 16*(src1[0] + src2[8]+1) - 3*(V+H);
-  for(j=8; j>0; --j) {
-    int b = a;
-    a += V;
-    src[0] = CLIP((b    ) >> 5);
-    src[1] = CLIP((b+  H) >> 5);
-    src[2] = CLIP((b+2*H) >> 5);
-    src[3] = CLIP((b+3*H) >> 5);
-    src[4] = CLIP((b+4*H) >> 5);
-    src[5] = CLIP((b+5*H) >> 5);
-    src[6] = CLIP((b+6*H) >> 5);
-    src[7] = CLIP((b+7*H) >> 5);
-    src += stride;
-  }
-}
-
-static void FUNCC(pred8x8_tm_vp8)(uint8_t *p_src, int p_stride){
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    uint8_t *cm = ff_cropTbl + MAX_NEG_CROP - src[-1-stride];
-    pixel *top = src-stride;
-    int y;
-
-    for (y = 0; y < 8; y++) {
-        uint8_t *cm_in = cm + src[-1];
-        src[0] = cm_in[top[0]];
-        src[1] = cm_in[top[1]];
-        src[2] = cm_in[top[2]];
-        src[3] = cm_in[top[3]];
-        src[4] = cm_in[top[4]];
-        src[5] = cm_in[top[5]];
-        src[6] = cm_in[top[6]];
-        src[7] = cm_in[top[7]];
-        src += stride;
-    }
-}
-
-#define SRC(x,y) src[(x)+(y)*stride]
-#define PL(y) \
-    const int l##y = (SRC(-1,y-1) + 2*SRC(-1,y) + SRC(-1,y+1) + 2) >> 2;
-#define PREDICT_8x8_LOAD_LEFT \
-    const int l0 = ((has_topleft ? SRC(-1,-1) : SRC(-1,0)) \
-                     + 2*SRC(-1,0) + SRC(-1,1) + 2) >> 2; \
-    PL(1) PL(2) PL(3) PL(4) PL(5) PL(6) \
-    const int l7 av_unused = (SRC(-1,6) + 3*SRC(-1,7) + 2) >> 2
-
-#define PT(x) \
-    const int t##x = (SRC(x-1,-1) + 2*SRC(x,-1) + SRC(x+1,-1) + 2) >> 2;
-#define PREDICT_8x8_LOAD_TOP \
-    const int t0 = ((has_topleft ? SRC(-1,-1) : SRC(0,-1)) \
-                     + 2*SRC(0,-1) + SRC(1,-1) + 2) >> 2; \
-    PT(1) PT(2) PT(3) PT(4) PT(5) PT(6) \
-    const int t7 av_unused = ((has_topright ? SRC(8,-1) : SRC(7,-1)) \
-                     + 2*SRC(7,-1) + SRC(6,-1) + 2) >> 2
-
-#define PTR(x) \
-    t##x = (SRC(x-1,-1) + 2*SRC(x,-1) + SRC(x+1,-1) + 2) >> 2;
-#define PREDICT_8x8_LOAD_TOPRIGHT \
-    int t8, t9, t10, t11, t12, t13, t14, t15; \
-    if(has_topright) { \
-        PTR(8) PTR(9) PTR(10) PTR(11) PTR(12) PTR(13) PTR(14) \
-        t15 = (SRC(14,-1) + 3*SRC(15,-1) + 2) >> 2; \
-    } else t8=t9=t10=t11=t12=t13=t14=t15= SRC(7,-1);
-
-#define PREDICT_8x8_LOAD_TOPLEFT \
-    const int lt = (SRC(-1,0) + 2*SRC(-1,-1) + SRC(0,-1) + 2) >> 2
-
-#define PREDICT_8x8_DC(v) \
-    int y; \
-    for( y = 0; y < 8; y++ ) { \
-        ((pixel4*)src)[0] = \
-        ((pixel4*)src)[1] = v; \
-        src += stride; \
-    }
-
-static void FUNCC(pred8x8l_128_dc)(uint8_t *p_src, int has_topleft, int has_topright, int p_stride)
-{
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-
-    PREDICT_8x8_DC(PIXEL_SPLAT_X4(1<<(BIT_DEPTH-1)));
-}
-static void FUNCC(pred8x8l_left_dc)(uint8_t *p_src, int has_topleft, int has_topright, int p_stride)
-{
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-
-    PREDICT_8x8_LOAD_LEFT;
-    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 *p_src, int has_topleft, int has_topright, int p_stride)
-{
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-
-    PREDICT_8x8_LOAD_TOP;
-    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 *p_src, int has_topleft, int has_topright, int p_stride)
-{
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-
-    PREDICT_8x8_LOAD_LEFT;
-    PREDICT_8x8_LOAD_TOP;
-    const pixel4 dc = PIXEL_SPLAT_X4((l0+l1+l2+l3+l4+l5+l6+l7
-                                     +t0+t1+t2+t3+t4+t5+t6+t7+8) >> 4);
-    PREDICT_8x8_DC(dc);
-}
-static void FUNCC(pred8x8l_horizontal)(uint8_t *p_src, int has_topleft, int has_topright, int p_stride)
-{
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-
-    PREDICT_8x8_LOAD_LEFT;
-#define ROW(y) ((pixel4*)(src+y*stride))[0] =\
-               ((pixel4*)(src+y*stride))[1] = PIXEL_SPLAT_X4(l##y)
-    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 *p_src, int has_topleft, int has_topright, int p_stride)
-{
-    int y;
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-
-    PREDICT_8x8_LOAD_TOP;
-    src[0] = t0;
-    src[1] = t1;
-    src[2] = t2;
-    src[3] = t3;
-    src[4] = t4;
-    src[5] = t5;
-    src[6] = t6;
-    src[7] = t7;
-    for( y = 1; y < 8; y++ ) {
-        ((pixel4*)(src+y*stride))[0] = ((pixel4*)src)[0];
-        ((pixel4*)(src+y*stride))[1] = ((pixel4*)src)[1];
-    }
-}
-static void FUNCC(pred8x8l_down_left)(uint8_t *p_src, int has_topleft, int has_topright, int p_stride)
-{
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    PREDICT_8x8_LOAD_TOP;
-    PREDICT_8x8_LOAD_TOPRIGHT;
-    SRC(0,0)= (t0 + 2*t1 + t2 + 2) >> 2;
-    SRC(0,1)=SRC(1,0)= (t1 + 2*t2 + t3 + 2) >> 2;
-    SRC(0,2)=SRC(1,1)=SRC(2,0)= (t2 + 2*t3 + t4 + 2) >> 2;
-    SRC(0,3)=SRC(1,2)=SRC(2,1)=SRC(3,0)= (t3 + 2*t4 + t5 + 2) >> 2;
-    SRC(0,4)=SRC(1,3)=SRC(2,2)=SRC(3,1)=SRC(4,0)= (t4 + 2*t5 + t6 + 2) >> 2;
-    SRC(0,5)=SRC(1,4)=SRC(2,3)=SRC(3,2)=SRC(4,1)=SRC(5,0)= (t5 + 2*t6 + t7 + 2) >> 2;
-    SRC(0,6)=SRC(1,5)=SRC(2,4)=SRC(3,3)=SRC(4,2)=SRC(5,1)=SRC(6,0)= (t6 + 2*t7 + t8 + 2) >> 2;
-    SRC(0,7)=SRC(1,6)=SRC(2,5)=SRC(3,4)=SRC(4,3)=SRC(5,2)=SRC(6,1)=SRC(7,0)= (t7 + 2*t8 + t9 + 2) >> 2;
-    SRC(1,7)=SRC(2,6)=SRC(3,5)=SRC(4,4)=SRC(5,3)=SRC(6,2)=SRC(7,1)= (t8 + 2*t9 + t10 + 2) >> 2;
-    SRC(2,7)=SRC(3,6)=SRC(4,5)=SRC(5,4)=SRC(6,3)=SRC(7,2)= (t9 + 2*t10 + t11 + 2) >> 2;
-    SRC(3,7)=SRC(4,6)=SRC(5,5)=SRC(6,4)=SRC(7,3)= (t10 + 2*t11 + t12 + 2) >> 2;
-    SRC(4,7)=SRC(5,6)=SRC(6,5)=SRC(7,4)= (t11 + 2*t12 + t13 + 2) >> 2;
-    SRC(5,7)=SRC(6,6)=SRC(7,5)= (t12 + 2*t13 + t14 + 2) >> 2;
-    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 *p_src, int has_topleft, int has_topright, int p_stride)
-{
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    PREDICT_8x8_LOAD_TOP;
-    PREDICT_8x8_LOAD_LEFT;
-    PREDICT_8x8_LOAD_TOPLEFT;
-    SRC(0,7)= (l7 + 2*l6 + l5 + 2) >> 2;
-    SRC(0,6)=SRC(1,7)= (l6 + 2*l5 + l4 + 2) >> 2;
-    SRC(0,5)=SRC(1,6)=SRC(2,7)= (l5 + 2*l4 + l3 + 2) >> 2;
-    SRC(0,4)=SRC(1,5)=SRC(2,6)=SRC(3,7)= (l4 + 2*l3 + l2 + 2) >> 2;
-    SRC(0,3)=SRC(1,4)=SRC(2,5)=SRC(3,6)=SRC(4,7)= (l3 + 2*l2 + l1 + 2) >> 2;
-    SRC(0,2)=SRC(1,3)=SRC(2,4)=SRC(3,5)=SRC(4,6)=SRC(5,7)= (l2 + 2*l1 + l0 + 2) >> 2;
-    SRC(0,1)=SRC(1,2)=SRC(2,3)=SRC(3,4)=SRC(4,5)=SRC(5,6)=SRC(6,7)= (l1 + 2*l0 + lt + 2) >> 2;
-    SRC(0,0)=SRC(1,1)=SRC(2,2)=SRC(3,3)=SRC(4,4)=SRC(5,5)=SRC(6,6)=SRC(7,7)= (l0 + 2*lt + t0 + 2) >> 2;
-    SRC(1,0)=SRC(2,1)=SRC(3,2)=SRC(4,3)=SRC(5,4)=SRC(6,5)=SRC(7,6)= (lt + 2*t0 + t1 + 2) >> 2;
-    SRC(2,0)=SRC(3,1)=SRC(4,2)=SRC(5,3)=SRC(6,4)=SRC(7,5)= (t0 + 2*t1 + t2 + 2) >> 2;
-    SRC(3,0)=SRC(4,1)=SRC(5,2)=SRC(6,3)=SRC(7,4)= (t1 + 2*t2 + t3 + 2) >> 2;
-    SRC(4,0)=SRC(5,1)=SRC(6,2)=SRC(7,3)= (t2 + 2*t3 + t4 + 2) >> 2;
-    SRC(5,0)=SRC(6,1)=SRC(7,2)= (t3 + 2*t4 + t5 + 2) >> 2;
-    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 *p_src, int has_topleft, int has_topright, int p_stride)
-{
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    PREDICT_8x8_LOAD_TOP;
-    PREDICT_8x8_LOAD_LEFT;
-    PREDICT_8x8_LOAD_TOPLEFT;
-    SRC(0,6)= (l5 + 2*l4 + l3 + 2) >> 2;
-    SRC(0,7)= (l6 + 2*l5 + l4 + 2) >> 2;
-    SRC(0,4)=SRC(1,6)= (l3 + 2*l2 + l1 + 2) >> 2;
-    SRC(0,5)=SRC(1,7)= (l4 + 2*l3 + l2 + 2) >> 2;
-    SRC(0,2)=SRC(1,4)=SRC(2,6)= (l1 + 2*l0 + lt + 2) >> 2;
-    SRC(0,3)=SRC(1,5)=SRC(2,7)= (l2 + 2*l1 + l0 + 2) >> 2;
-    SRC(0,1)=SRC(1,3)=SRC(2,5)=SRC(3,7)= (l0 + 2*lt + t0 + 2) >> 2;
-    SRC(0,0)=SRC(1,2)=SRC(2,4)=SRC(3,6)= (lt + t0 + 1) >> 1;
-    SRC(1,1)=SRC(2,3)=SRC(3,5)=SRC(4,7)= (lt + 2*t0 + t1 + 2) >> 2;
-    SRC(1,0)=SRC(2,2)=SRC(3,4)=SRC(4,6)= (t0 + t1 + 1) >> 1;
-    SRC(2,1)=SRC(3,3)=SRC(4,5)=SRC(5,7)= (t0 + 2*t1 + t2 + 2) >> 2;
-    SRC(2,0)=SRC(3,2)=SRC(4,4)=SRC(5,6)= (t1 + t2 + 1) >> 1;
-    SRC(3,1)=SRC(4,3)=SRC(5,5)=SRC(6,7)= (t1 + 2*t2 + t3 + 2) >> 2;
-    SRC(3,0)=SRC(4,2)=SRC(5,4)=SRC(6,6)= (t2 + t3 + 1) >> 1;
-    SRC(4,1)=SRC(5,3)=SRC(6,5)=SRC(7,7)= (t2 + 2*t3 + t4 + 2) >> 2;
-    SRC(4,0)=SRC(5,2)=SRC(6,4)=SRC(7,6)= (t3 + t4 + 1) >> 1;
-    SRC(5,1)=SRC(6,3)=SRC(7,5)= (t3 + 2*t4 + t5 + 2) >> 2;
-    SRC(5,0)=SRC(6,2)=SRC(7,4)= (t4 + t5 + 1) >> 1;
-    SRC(6,1)=SRC(7,3)= (t4 + 2*t5 + t6 + 2) >> 2;
-    SRC(6,0)=SRC(7,2)= (t5 + t6 + 1) >> 1;
-    SRC(7,1)= (t5 + 2*t6 + t7 + 2) >> 2;
-    SRC(7,0)= (t6 + t7 + 1) >> 1;
-}
-static void FUNCC(pred8x8l_horizontal_down)(uint8_t *p_src, int has_topleft, int has_topright, int p_stride)
-{
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    PREDICT_8x8_LOAD_TOP;
-    PREDICT_8x8_LOAD_LEFT;
-    PREDICT_8x8_LOAD_TOPLEFT;
-    SRC(0,7)= (l6 + l7 + 1) >> 1;
-    SRC(1,7)= (l5 + 2*l6 + l7 + 2) >> 2;
-    SRC(0,6)=SRC(2,7)= (l5 + l6 + 1) >> 1;
-    SRC(1,6)=SRC(3,7)= (l4 + 2*l5 + l6 + 2) >> 2;
-    SRC(0,5)=SRC(2,6)=SRC(4,7)= (l4 + l5 + 1) >> 1;
-    SRC(1,5)=SRC(3,6)=SRC(5,7)= (l3 + 2*l4 + l5 + 2) >> 2;
-    SRC(0,4)=SRC(2,5)=SRC(4,6)=SRC(6,7)= (l3 + l4 + 1) >> 1;
-    SRC(1,4)=SRC(3,5)=SRC(5,6)=SRC(7,7)= (l2 + 2*l3 + l4 + 2) >> 2;
-    SRC(0,3)=SRC(2,4)=SRC(4,5)=SRC(6,6)= (l2 + l3 + 1) >> 1;
-    SRC(1,3)=SRC(3,4)=SRC(5,5)=SRC(7,6)= (l1 + 2*l2 + l3 + 2) >> 2;
-    SRC(0,2)=SRC(2,3)=SRC(4,4)=SRC(6,5)= (l1 + l2 + 1) >> 1;
-    SRC(1,2)=SRC(3,3)=SRC(5,4)=SRC(7,5)= (l0 + 2*l1 + l2 + 2) >> 2;
-    SRC(0,1)=SRC(2,2)=SRC(4,3)=SRC(6,4)= (l0 + l1 + 1) >> 1;
-    SRC(1,1)=SRC(3,2)=SRC(5,3)=SRC(7,4)= (lt + 2*l0 + l1 + 2) >> 2;
-    SRC(0,0)=SRC(2,1)=SRC(4,2)=SRC(6,3)= (lt + l0 + 1) >> 1;
-    SRC(1,0)=SRC(3,1)=SRC(5,2)=SRC(7,3)= (l0 + 2*lt + t0 + 2) >> 2;
-    SRC(2,0)=SRC(4,1)=SRC(6,2)= (t1 + 2*t0 + lt + 2) >> 2;
-    SRC(3,0)=SRC(5,1)=SRC(7,2)= (t2 + 2*t1 + t0 + 2) >> 2;
-    SRC(4,0)=SRC(6,1)= (t3 + 2*t2 + t1 + 2) >> 2;
-    SRC(5,0)=SRC(7,1)= (t4 + 2*t3 + t2 + 2) >> 2;
-    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 *p_src, int has_topleft, int has_topright, int p_stride)
-{
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    PREDICT_8x8_LOAD_TOP;
-    PREDICT_8x8_LOAD_TOPRIGHT;
-    SRC(0,0)= (t0 + t1 + 1) >> 1;
-    SRC(0,1)= (t0 + 2*t1 + t2 + 2) >> 2;
-    SRC(0,2)=SRC(1,0)= (t1 + t2 + 1) >> 1;
-    SRC(0,3)=SRC(1,1)= (t1 + 2*t2 + t3 + 2) >> 2;
-    SRC(0,4)=SRC(1,2)=SRC(2,0)= (t2 + t3 + 1) >> 1;
-    SRC(0,5)=SRC(1,3)=SRC(2,1)= (t2 + 2*t3 + t4 + 2) >> 2;
-    SRC(0,6)=SRC(1,4)=SRC(2,2)=SRC(3,0)= (t3 + t4 + 1) >> 1;
-    SRC(0,7)=SRC(1,5)=SRC(2,3)=SRC(3,1)= (t3 + 2*t4 + t5 + 2) >> 2;
-    SRC(1,6)=SRC(2,4)=SRC(3,2)=SRC(4,0)= (t4 + t5 + 1) >> 1;
-    SRC(1,7)=SRC(2,5)=SRC(3,3)=SRC(4,1)= (t4 + 2*t5 + t6 + 2) >> 2;
-    SRC(2,6)=SRC(3,4)=SRC(4,2)=SRC(5,0)= (t5 + t6 + 1) >> 1;
-    SRC(2,7)=SRC(3,5)=SRC(4,3)=SRC(5,1)= (t5 + 2*t6 + t7 + 2) >> 2;
-    SRC(3,6)=SRC(4,4)=SRC(5,2)=SRC(6,0)= (t6 + t7 + 1) >> 1;
-    SRC(3,7)=SRC(4,5)=SRC(5,3)=SRC(6,1)= (t6 + 2*t7 + t8 + 2) >> 2;
-    SRC(4,6)=SRC(5,4)=SRC(6,2)=SRC(7,0)= (t7 + t8 + 1) >> 1;
-    SRC(4,7)=SRC(5,5)=SRC(6,3)=SRC(7,1)= (t7 + 2*t8 + t9 + 2) >> 2;
-    SRC(5,6)=SRC(6,4)=SRC(7,2)= (t8 + t9 + 1) >> 1;
-    SRC(5,7)=SRC(6,5)=SRC(7,3)= (t8 + 2*t9 + t10 + 2) >> 2;
-    SRC(6,6)=SRC(7,4)= (t9 + t10 + 1) >> 1;
-    SRC(6,7)=SRC(7,5)= (t9 + 2*t10 + t11 + 2) >> 2;
-    SRC(7,6)= (t10 + t11 + 1) >> 1;
-    SRC(7,7)= (t10 + 2*t11 + t12 + 2) >> 2;
-}
-static void FUNCC(pred8x8l_horizontal_up)(uint8_t *p_src, int has_topleft, int has_topright, int p_stride)
-{
-    pixel *src = (pixel*)p_src;
-    int stride = p_stride>>(sizeof(pixel)-1);
-    PREDICT_8x8_LOAD_LEFT;
-    SRC(0,0)= (l0 + l1 + 1) >> 1;
-    SRC(1,0)= (l0 + 2*l1 + l2 + 2) >> 2;
-    SRC(0,1)=SRC(2,0)= (l1 + l2 + 1) >> 1;
-    SRC(1,1)=SRC(3,0)= (l1 + 2*l2 + l3 + 2) >> 2;
-    SRC(0,2)=SRC(2,1)=SRC(4,0)= (l2 + l3 + 1) >> 1;
-    SRC(1,2)=SRC(3,1)=SRC(5,0)= (l2 + 2*l3 + l4 + 2) >> 2;
-    SRC(0,3)=SRC(2,2)=SRC(4,1)=SRC(6,0)= (l3 + l4 + 1) >> 1;
-    SRC(1,3)=SRC(3,2)=SRC(5,1)=SRC(7,0)= (l3 + 2*l4 + l5 + 2) >> 2;
-    SRC(0,4)=SRC(2,3)=SRC(4,2)=SRC(6,1)= (l4 + l5 + 1) >> 1;
-    SRC(1,4)=SRC(3,3)=SRC(5,2)=SRC(7,1)= (l4 + 2*l5 + l6 + 2) >> 2;
-    SRC(0,5)=SRC(2,4)=SRC(4,3)=SRC(6,2)= (l5 + l6 + 1) >> 1;
-    SRC(1,5)=SRC(3,4)=SRC(5,3)=SRC(7,2)= (l5 + 2*l6 + l7 + 2) >> 2;
-    SRC(0,6)=SRC(2,5)=SRC(4,4)=SRC(6,3)= (l6 + l7 + 1) >> 1;
-    SRC(1,6)=SRC(3,5)=SRC(5,4)=SRC(7,3)= (l6 + 3*l7 + 2) >> 2;
-    SRC(0,7)=SRC(1,7)=SRC(2,6)=SRC(2,7)=SRC(3,6)=
-    SRC(3,7)=SRC(4,5)=SRC(4,6)=SRC(4,7)=SRC(5,5)=
-    SRC(5,6)=SRC(5,7)=SRC(6,4)=SRC(6,5)=SRC(6,6)=
-    SRC(6,7)=SRC(7,4)=SRC(7,5)=SRC(7,6)=SRC(7,7)= l7;
-}
-#undef PREDICT_8x8_LOAD_LEFT
-#undef PREDICT_8x8_LOAD_TOP
-#undef PREDICT_8x8_LOAD_TOPLEFT
-#undef PREDICT_8x8_LOAD_TOPRIGHT
-#undef PREDICT_8x8_DC
-#undef PTR
-#undef PT
-#undef PL
-#undef SRC
-
-static void FUNCC(pred4x4_vertical_add)(uint8_t *p_pix, const DCTELEM *p_block, int stride){
-    int i;
-    pixel *pix = (pixel*)p_pix;
-    const dctcoef *block = (const dctcoef*)p_block;
-    stride >>= sizeof(pixel)-1;
-    pix -= stride;
-    for(i=0; i<4; i++){
-        pixel v = pix[0];
-        pix[1*stride]= v += block[0];
-        pix[2*stride]= v += block[4];
-        pix[3*stride]= v += block[8];
-        pix[4*stride]= v +  block[12];
-        pix++;
-        block++;
-    }
-}
-
-static void FUNCC(pred4x4_horizontal_add)(uint8_t *p_pix, const DCTELEM *p_block, int stride){
-    int i;
-    pixel *pix = (pixel*)p_pix;
-    const dctcoef *block = (const dctcoef*)p_block;
-    stride >>= sizeof(pixel)-1;
-    for(i=0; i<4; i++){
-        pixel v = pix[-1];
-        pix[0]= v += block[0];
-        pix[1]= v += block[1];
-        pix[2]= v += block[2];
-        pix[3]= v +  block[3];
-        pix+= stride;
-        block+= 4;
-    }
-}
-
-static void FUNCC(pred8x8l_vertical_add)(uint8_t *p_pix, const DCTELEM *p_block, int stride){
-    int i;
-    pixel *pix = (pixel*)p_pix;
-    const dctcoef *block = (const dctcoef*)p_block;
-    stride >>= sizeof(pixel)-1;
-    pix -= stride;
-    for(i=0; i<8; i++){
-        pixel v = pix[0];
-        pix[1*stride]= v += block[0];
-        pix[2*stride]= v += block[8];
-        pix[3*stride]= v += block[16];
-        pix[4*stride]= v += block[24];
-        pix[5*stride]= v += block[32];
-        pix[6*stride]= v += block[40];
-        pix[7*stride]= v += block[48];
-        pix[8*stride]= v +  block[56];
-        pix++;
-        block++;
-    }
-}
-
-static void FUNCC(pred8x8l_horizontal_add)(uint8_t *p_pix, const DCTELEM *p_block, int stride){
-    int i;
-    pixel *pix = (pixel*)p_pix;
-    const dctcoef *block = (const dctcoef*)p_block;
-    stride >>= sizeof(pixel)-1;
-    for(i=0; i<8; i++){
-        pixel v = pix[-1];
-        pix[0]= v += block[0];
-        pix[1]= v += block[1];
-        pix[2]= v += block[2];
-        pix[3]= v += block[3];
-        pix[4]= v += block[4];
-        pix[5]= v += block[5];
-        pix[6]= v += block[6];
-        pix[7]= v +  block[7];
-        pix+= stride;
-        block+= 8;
-    }
-}
-
-static void FUNCC(pred16x16_vertical_add)(uint8_t *pix, const int *block_offset, const DCTELEM *block, int 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){
-    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){
-    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(pred8x8_horizontal_add)(uint8_t *pix, const int *block_offset, const DCTELEM *block, int 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/h264pred_template.c b/libavcodec/h264pred_template.c
index c600133346..d056e20e80 100644
--- a/libavcodec/h264pred_template.c
+++ b/libavcodec/h264pred_template.c
@@ -2,20 +2,20 @@
  * H.26L/H.264/AVC/JVT/14496-10/... encoder/decoder
  * Copyright (c) 2003-2011 Michael Niedermayer <michaelni@gmx.at>
  *
- * This file is part of Libav.
+ * This file is part of FFmpeg.
  *
- * Libav is free software; you can redistribute it and/or
+ * FFmpeg 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.
  *
- * Libav is distributed in the hope that it will be useful,
+ * FFmpeg 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 Libav; if not, write to the Free Software
+ * License along with FFmpeg; if not, write to the Free Software
  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  */
 
@@ -30,7 +30,7 @@
 
 static void FUNCC(pred4x4_vertical)(uint8_t *_src, const uint8_t *topright, int _stride){
     pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    int stride = _stride>>(sizeof(pixel)-1);
     const pixel4 a= AV_RN4PA(src-stride);
 
     AV_WN4PA(src+0*stride, a);
@@ -41,7 +41,7 @@ static void FUNCC(pred4x4_vertical)(uint8_t *_src, const uint8_t *topright, int
 
 static void FUNCC(pred4x4_horizontal)(uint8_t *_src, const uint8_t *topright, int _stride){
     pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    int stride = _stride>>(sizeof(pixel)-1);
     AV_WN4PA(src+0*stride, PIXEL_SPLAT_X4(src[-1+0*stride]));
     AV_WN4PA(src+1*stride, PIXEL_SPLAT_X4(src[-1+1*stride]));
     AV_WN4PA(src+2*stride, PIXEL_SPLAT_X4(src[-1+2*stride]));
@@ -50,7 +50,7 @@ static void FUNCC(pred4x4_horizontal)(uint8_t *_src, const uint8_t *topright, in
 
 static void FUNCC(pred4x4_dc)(uint8_t *_src, const uint8_t *topright, int _stride){
     pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    int stride = _stride>>(sizeof(pixel)-1);
     const int dc= (  src[-stride] + src[1-stride] + src[2-stride] + src[3-stride]
                    + src[-1+0*stride] + src[-1+1*stride] + src[-1+2*stride] + src[-1+3*stride] + 4) >>3;
     const pixel4 a = PIXEL_SPLAT_X4(dc);
@@ -63,7 +63,7 @@ static void FUNCC(pred4x4_dc)(uint8_t *_src, const uint8_t *topright, int _strid
 
 static void FUNCC(pred4x4_left_dc)(uint8_t *_src, const uint8_t *topright, int _stride){
     pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    int stride = _stride>>(sizeof(pixel)-1);
     const int dc= (  src[-1+0*stride] + src[-1+1*stride] + src[-1+2*stride] + src[-1+3*stride] + 2) >>2;
     const pixel4 a = PIXEL_SPLAT_X4(dc);
 
@@ -75,7 +75,7 @@ static void FUNCC(pred4x4_left_dc)(uint8_t *_src, const uint8_t *topright, int _
 
 static void FUNCC(pred4x4_top_dc)(uint8_t *_src, const uint8_t *topright, int _stride){
     pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    int stride = _stride>>(sizeof(pixel)-1);
     const int dc= (  src[-stride] + src[1-stride] + src[2-stride] + src[3-stride] + 2) >>2;
     const pixel4 a = PIXEL_SPLAT_X4(dc);
 
@@ -87,7 +87,7 @@ static void FUNCC(pred4x4_top_dc)(uint8_t *_src, const uint8_t *topright, int _s
 
 static void FUNCC(pred4x4_128_dc)(uint8_t *_src, const uint8_t *topright, int _stride){
     pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    int stride = _stride>>(sizeof(pixel)-1);
     const pixel4 a = PIXEL_SPLAT_X4(1<<(BIT_DEPTH-1));
 
     AV_WN4PA(src+0*stride, a);
@@ -98,7 +98,7 @@ static void FUNCC(pred4x4_128_dc)(uint8_t *_src, const uint8_t *topright, int _s
 
 static void FUNCC(pred4x4_127_dc)(uint8_t *_src, const uint8_t *topright, int _stride){
     pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    int stride = _stride>>(sizeof(pixel)-1);
     const pixel4 a = PIXEL_SPLAT_X4((1<<(BIT_DEPTH-1))-1);
 
     AV_WN4PA(src+0*stride, a);
@@ -109,7 +109,7 @@ static void FUNCC(pred4x4_127_dc)(uint8_t *_src, const uint8_t *topright, int _s
 
 static void FUNCC(pred4x4_129_dc)(uint8_t *_src, const uint8_t *topright, int _stride){
     pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    int stride = _stride>>(sizeof(pixel)-1);
     const pixel4 a = PIXEL_SPLAT_X4((1<<(BIT_DEPTH-1))+1);
 
     AV_WN4PA(src+0*stride, a);
@@ -145,7 +145,7 @@ static void FUNCC(pred4x4_129_dc)(uint8_t *_src, const uint8_t *topright, int _s
 
 static void FUNCC(pred4x4_down_right)(uint8_t *_src, const uint8_t *topright, int _stride){
     pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    int stride = _stride>>(sizeof(pixel)-1);
     const int lt= src[-1-1*stride];
     LOAD_TOP_EDGE
     LOAD_LEFT_EDGE
@@ -171,7 +171,7 @@ static void FUNCC(pred4x4_down_right)(uint8_t *_src, const uint8_t *topright, in
 static void FUNCC(pred4x4_down_left)(uint8_t *_src, const uint8_t *_topright, int _stride){
     pixel *src = (pixel*)_src;
     const pixel *topright = (const pixel*)_topright;
-    int stride = _stride/sizeof(pixel);
+    int stride = _stride>>(sizeof(pixel)-1);
     LOAD_TOP_EDGE
     LOAD_TOP_RIGHT_EDGE
 //    LOAD_LEFT_EDGE
@@ -196,7 +196,7 @@ static void FUNCC(pred4x4_down_left)(uint8_t *_src, const uint8_t *_topright, in
 
 static void FUNCC(pred4x4_vertical_right)(uint8_t *_src, const uint8_t *topright, int _stride){
     pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    int stride = _stride>>(sizeof(pixel)-1);
     const int lt= src[-1-1*stride];
     LOAD_TOP_EDGE
     LOAD_LEFT_EDGE
@@ -222,7 +222,7 @@ static void FUNCC(pred4x4_vertical_right)(uint8_t *_src, const uint8_t *topright
 static void FUNCC(pred4x4_vertical_left)(uint8_t *_src, const uint8_t *_topright, int _stride){
     pixel *src = (pixel*)_src;
     const pixel *topright = (const pixel*)_topright;
-    int stride = _stride/sizeof(pixel);
+    int stride = _stride>>(sizeof(pixel)-1);
     LOAD_TOP_EDGE
     LOAD_TOP_RIGHT_EDGE
 
@@ -246,7 +246,7 @@ static void FUNCC(pred4x4_vertical_left)(uint8_t *_src, const uint8_t *_topright
 
 static void FUNCC(pred4x4_horizontal_up)(uint8_t *_src, const uint8_t *topright, int _stride){
     pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    int stride = _stride>>(sizeof(pixel)-1);
     LOAD_LEFT_EDGE
 
     src[0+0*stride]=(l0 + l1 + 1)>>1;
@@ -269,7 +269,7 @@ static void FUNCC(pred4x4_horizontal_up)(uint8_t *_src, const uint8_t *topright,
 
 static void FUNCC(pred4x4_horizontal_down)(uint8_t *_src, const uint8_t *topright, int _stride){
     pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    int stride = _stride>>(sizeof(pixel)-1);
     const int lt= src[-1-1*stride];
     LOAD_TOP_EDGE
     LOAD_LEFT_EDGE
@@ -295,7 +295,7 @@ static void FUNCC(pred4x4_horizontal_down)(uint8_t *_src, const uint8_t *toprigh
 static void FUNCC(pred16x16_vertical)(uint8_t *_src, int _stride){
     int i;
     pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    int stride = _stride>>(sizeof(pixel)-1);
     const pixel4 a = AV_RN4PA(((pixel4*)(src-stride))+0);
     const pixel4 b = AV_RN4PA(((pixel4*)(src-stride))+1);
     const pixel4 c = AV_RN4PA(((pixel4*)(src-stride))+2);
@@ -312,7 +312,7 @@ static void FUNCC(pred16x16_vertical)(uint8_t *_src, int _stride){
 static void FUNCC(pred16x16_horizontal)(uint8_t *_src, int stride){
     int i;
     pixel *src = (pixel*)_src;
-    stride /= sizeof(pixel);
+    stride >>= sizeof(pixel)-1;
 
     for(i=0; i<16; i++){
         const pixel4 a = PIXEL_SPLAT_X4(src[-1+i*stride]);
@@ -337,7 +337,7 @@ static void FUNCC(pred16x16_dc)(uint8_t *_src, int stride){
     int i, dc=0;
     pixel *src = (pixel*)_src;
     pixel4 dcsplat;
-    stride /= sizeof(pixel);
+    stride >>= sizeof(pixel)-1;
 
     for(i=0;i<16; i++){
         dc+= src[-1+i*stride];
@@ -355,7 +355,7 @@ static void FUNCC(pred16x16_left_dc)(uint8_t *_src, int stride){
     int i, dc=0;
     pixel *src = (pixel*)_src;
     pixel4 dcsplat;
-    stride /= sizeof(pixel);
+    stride >>= sizeof(pixel)-1;
 
     for(i=0;i<16; i++){
         dc+= src[-1+i*stride];
@@ -369,7 +369,7 @@ static void FUNCC(pred16x16_top_dc)(uint8_t *_src, int stride){
     int i, dc=0;
     pixel *src = (pixel*)_src;
     pixel4 dcsplat;
-    stride /= sizeof(pixel);
+    stride >>= sizeof(pixel)-1;
 
     for(i=0;i<16; i++){
         dc+= src[i-stride];
@@ -383,7 +383,7 @@ static void FUNCC(pred16x16_top_dc)(uint8_t *_src, int stride){
 static void FUNCC(pred16x16_##n##_dc)(uint8_t *_src, int stride){\
     int i;\
     pixel *src = (pixel*)_src;\
-    stride /= sizeof(pixel);\
+    stride >>= sizeof(pixel)-1;\
     PREDICT_16x16_DC(PIXEL_SPLAT_X4(v));\
 }
 
@@ -391,12 +391,12 @@ 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 *p_src, int p_stride, const int svq3, const int rv40){
   int i, j, k;
   int a;
   INIT_CLIP
-  pixel *src = (pixel*)_src;
-  int stride = _stride/sizeof(pixel);
+  pixel *src = (pixel*)p_src;
+  int stride = p_stride>>(sizeof(pixel)-1);
   const pixel * const src0 = src +7-stride;
   const pixel *       src1 = src +8*stride-1;
   const pixel *       src2 = src1-2*stride;    // == src+6*stride-1;
@@ -443,7 +443,7 @@ static void FUNCC(pred16x16_plane)(uint8_t *src, int stride){
 static void FUNCC(pred8x8_vertical)(uint8_t *_src, int _stride){
     int i;
     pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    int stride = _stride>>(sizeof(pixel)-1);
     const pixel4 a= AV_RN4PA(((pixel4*)(src-stride))+0);
     const pixel4 b= AV_RN4PA(((pixel4*)(src-stride))+1);
 
@@ -456,7 +456,7 @@ static void FUNCC(pred8x8_vertical)(uint8_t *_src, int _stride){
 static void FUNCC(pred8x8_horizontal)(uint8_t *_src, int stride){
     int i;
     pixel *src = (pixel*)_src;
-    stride /= sizeof(pixel);
+    stride >>= sizeof(pixel)-1;
 
     for(i=0; i<8; i++){
         const pixel4 a = PIXEL_SPLAT_X4(src[-1+i*stride]);
@@ -470,7 +470,7 @@ static void FUNCC(pred8x8_##n##_dc)(uint8_t *_src, int stride){\
     int i;\
     const pixel4 a = PIXEL_SPLAT_X4(v);\
     pixel *src = (pixel*)_src;\
-    stride /= sizeof(pixel);\
+    stride >>= sizeof(pixel)-1;\
     for(i=0; i<8; i++){\
         AV_WN4PA(((pixel4*)(src+i*stride))+0, a);\
         AV_WN4PA(((pixel4*)(src+i*stride))+1, a);\
@@ -486,7 +486,7 @@ static void FUNCC(pred8x8_left_dc)(uint8_t *_src, int stride){
     int dc0, dc2;
     pixel4 dc0splat, dc2splat;
     pixel *src = (pixel*)_src;
-    stride /= sizeof(pixel);
+    stride >>= sizeof(pixel)-1;
 
     dc0=dc2=0;
     for(i=0;i<4; i++){
@@ -511,7 +511,7 @@ static void FUNCC(pred8x8_top_dc)(uint8_t *_src, int stride){
     int dc0, dc1;
     pixel4 dc0splat, dc1splat;
     pixel *src = (pixel*)_src;
-    stride /= sizeof(pixel);
+    stride >>= sizeof(pixel)-1;
 
     dc0=dc1=0;
     for(i=0;i<4; i++){
@@ -536,7 +536,7 @@ static void FUNCC(pred8x8_dc)(uint8_t *_src, int stride){
     int dc0, dc1, dc2;
     pixel4 dc0splat, dc1splat, dc2splat, dc3splat;
     pixel *src = (pixel*)_src;
-    stride /= sizeof(pixel);
+    stride >>= sizeof(pixel)-1;
 
     dc0=dc1=dc2=0;
     for(i=0;i<4; i++){
@@ -587,7 +587,7 @@ static void FUNCC(pred8x8_plane)(uint8_t *_src, int _stride){
   int a;
   INIT_CLIP
   pixel *src = (pixel*)_src;
-  int stride = _stride/sizeof(pixel);
+  int stride = _stride>>(sizeof(pixel)-1);
   const pixel * const src0 = src +3-stride;
   const pixel *       src1 = src +4*stride-1;
   const pixel *       src2 = src1-2*stride;    // == src+2*stride-1;
@@ -658,32 +658,32 @@ static void FUNCC(pred8x8_plane)(uint8_t *_src, int _stride){
 static void FUNCC(pred8x8l_128_dc)(uint8_t *_src, int has_topleft, int has_topright, int _stride)
 {
     pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    int stride = _stride>>(sizeof(pixel)-1);
 
     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)
 {
     pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    int stride = _stride>>(sizeof(pixel)-1);
 
     PREDICT_8x8_LOAD_LEFT;
     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 *p_src, int has_topleft, int has_topright, int p_stride)
 {
-    pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    pixel *src = (pixel*)p_src;
+    int stride = p_stride>>(sizeof(pixel)-1);
 
     PREDICT_8x8_LOAD_TOP;
     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 *p_src, int has_topleft, int has_topright, int p_stride)
 {
-    pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    pixel *src = (pixel*)p_src;
+    int stride = p_stride>>(sizeof(pixel)-1);
 
     PREDICT_8x8_LOAD_LEFT;
     PREDICT_8x8_LOAD_TOP;
@@ -691,10 +691,10 @@ 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 *p_src, int has_topleft, int has_topright, int p_stride)
 {
-    pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    pixel *src = (pixel*)p_src;
+    int stride = p_stride>>(sizeof(pixel)-1);
 
     PREDICT_8x8_LOAD_LEFT;
 #define ROW(y) ((pixel4*)(src+y*stride))[0] =\
@@ -706,7 +706,7 @@ static void FUNCC(pred8x8l_vertical)(uint8_t *_src, int has_topleft, int has_top
 {
     int y;
     pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    int stride = _stride>>(sizeof(pixel)-1);
     pixel4 a, b;
 
     PREDICT_8x8_LOAD_TOP;
@@ -725,10 +725,10 @@ 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 *p_src, int has_topleft, int has_topright, int p_stride)
 {
-    pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    pixel *src = (pixel*)p_src;
+    int stride = p_stride>>(sizeof(pixel)-1);
     PREDICT_8x8_LOAD_TOP;
     PREDICT_8x8_LOAD_TOPRIGHT;
     SRC(0,0)= (t0 + 2*t1 + t2 + 2) >> 2;
@@ -747,10 +747,10 @@ 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 *p_src, int has_topleft, int has_topright, int p_stride)
 {
-    pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    pixel *src = (pixel*)p_src;
+    int stride = p_stride>>(sizeof(pixel)-1);
     PREDICT_8x8_LOAD_TOP;
     PREDICT_8x8_LOAD_LEFT;
     PREDICT_8x8_LOAD_TOPLEFT;
@@ -770,10 +770,10 @@ 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 *p_src, int has_topleft, int has_topright, int p_stride)
 {
-    pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    pixel *src = (pixel*)p_src;
+    int stride = p_stride>>(sizeof(pixel)-1);
     PREDICT_8x8_LOAD_TOP;
     PREDICT_8x8_LOAD_LEFT;
     PREDICT_8x8_LOAD_TOPLEFT;
@@ -800,10 +800,10 @@ 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 *p_src, int has_topleft, int has_topright, int p_stride)
 {
-    pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    pixel *src = (pixel*)p_src;
+    int stride = p_stride>>(sizeof(pixel)-1);
     PREDICT_8x8_LOAD_TOP;
     PREDICT_8x8_LOAD_LEFT;
     PREDICT_8x8_LOAD_TOPLEFT;
@@ -830,10 +830,10 @@ 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 *p_src, int has_topleft, int has_topright, int p_stride)
 {
-    pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    pixel *src = (pixel*)p_src;
+    int stride = p_stride>>(sizeof(pixel)-1);
     PREDICT_8x8_LOAD_TOP;
     PREDICT_8x8_LOAD_TOPRIGHT;
     SRC(0,0)= (t0 + t1 + 1) >> 1;
@@ -859,10 +859,10 @@ 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 *p_src, int has_topleft, int has_topright, int p_stride)
 {
-    pixel *src = (pixel*)_src;
-    int stride = _stride/sizeof(pixel);
+    pixel *src = (pixel*)p_src;
+    int stride = p_stride>>(sizeof(pixel)-1);
     PREDICT_8x8_LOAD_LEFT;
     SRC(0,0)= (l0 + l1 + 1) >> 1;
     SRC(1,0)= (l0 + 2*l1 + l2 + 2) >> 2;
@@ -893,11 +893,11 @@ 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 *p_pix, const DCTELEM *p_block, int stride){
     int i;
-    pixel *pix = (pixel*)_pix;
-    const dctcoef *block = (const dctcoef*)_block;
-    stride /= sizeof(pixel);
+    pixel *pix = (pixel*)p_pix;
+    const dctcoef *block = (const dctcoef*)p_block;
+    stride >>= sizeof(pixel)-1;
     pix -= stride;
     for(i=0; i<4; i++){
         pixel v = pix[0];
@@ -910,11 +910,11 @@ 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 *p_pix, const DCTELEM *p_block, int stride){
     int i;
-    pixel *pix = (pixel*)_pix;
-    const dctcoef *block = (const dctcoef*)_block;
-    stride /= sizeof(pixel);
+    pixel *pix = (pixel*)p_pix;
+    const dctcoef *block = (const dctcoef*)p_block;
+    stride >>= sizeof(pixel)-1;
     for(i=0; i<4; i++){
         pixel v = pix[-1];
         pix[0]= v += block[0];
@@ -926,11 +926,11 @@ 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 *p_pix, const DCTELEM *p_block, int stride){
     int i;
-    pixel *pix = (pixel*)_pix;
-    const dctcoef *block = (const dctcoef*)_block;
-    stride /= sizeof(pixel);
+    pixel *pix = (pixel*)p_pix;
+    const dctcoef *block = (const dctcoef*)p_block;
+    stride >>= sizeof(pixel)-1;
     pix -= stride;
     for(i=0; i<8; i++){
         pixel v = pix[0];
@@ -947,11 +947,11 @@ 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 *p_pix, const DCTELEM *p_block, int stride){
     int i;
-    pixel *pix = (pixel*)_pix;
-    const dctcoef *block = (const dctcoef*)_block;
-    stride /= sizeof(pixel);
+    pixel *pix = (pixel*)p_pix;
+    const dctcoef *block = (const dctcoef*)p_block;
+    stride >>= sizeof(pixel)-1;
     for(i=0; i<8; i++){
         pixel v = pix[-1];
         pix[0]= v += block[0];