tests/checkasm/vp6dsp.c

   1 /*
   2  * This file is part of FFmpeg.
   3  *
   4  * FFmpeg is free software; you can redistribute it and/or modify
   5  * it under the terms of the GNU General Public License as published by
   6  * the Free Software Foundation; either version 2 of the License, or
   7  * (at your option) any later version.
   8  *
   9  * FFmpeg is distributed in the hope that it will be useful,
  10  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  11  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12  * GNU General Public License for more details.
  13  *
  14  * You should have received a copy of the GNU General Public License along
  15  * with FFmpeg; if not, write to the Free Software Foundation, Inc.,
  16  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
  17  */
  18
  19 #include <assert.h>
  20 #include <stddef.h>
  21 #include <string.h>
  22
  23 #include "checkasm.h"
  24 #include "libavutil/intreadwrite.h"
  25 #include "libavutil/macros.h"
  26 #include "libavutil/mem_internal.h"
  27 #include "libavcodec/vp6data.h"
  28 #include "libavcodec/vp56dsp.h"
  29
  30 #define randomize_buffer(buf)                                   \
  31     do {                                                        \
  32         for (size_t k = 0; k < (sizeof(buf) & ~3); k += 4)      \
  33             AV_WN32A(buf + k, rnd());                           \
  34         for (size_t k = sizeof(buf) & ~3; k < sizeof(buf); ++k) \
  35             buf[k] = rnd();                                     \
  36     } while (0)
  37
  38
  39 void checkasm_check_vp6dsp(void)
  40 {
  41     enum {
  42         BLOCK_SIZE_1D  = 8,
  43         SRC_ROWS_ABOVE = 1,
  44         SRC_ROWS_BELOW = 2,
  45         SRC_COLS_LEFT  = 1,
  46         SRC_COLS_RIGHT = 2,
  47         SRC_ROWS       = SRC_ROWS_ABOVE + BLOCK_SIZE_1D + SRC_ROWS_BELOW,
  48         SRC_ROW_SIZE   = SRC_COLS_LEFT  + BLOCK_SIZE_1D + SRC_COLS_RIGHT,
  49         MAX_STRIDE     = 64,    ///< arbitrary
  50         SRC_BUF_SIZE   = (SRC_ROWS - 1) * MAX_STRIDE + SRC_ROW_SIZE + 7 /* to vary misalignment */,
  51         DST_BUF_SIZE   = (BLOCK_SIZE_1D - 1) * MAX_STRIDE + BLOCK_SIZE_1D,
  52     };
  53     VP6DSPContext vp6dsp;
  54
  55     ff_vp6dsp_init(&vp6dsp);
  56
  57     declare_func(void, uint8_t *dst, const uint8_t *src, ptrdiff_t stride,
  58                        const int16_t *h_weights, const int16_t *v_weights);
  59
  60     if (check_func(vp6dsp.vp6_filter_diag4, "filter_diag4")) {
  61         DECLARE_ALIGNED(8, uint8_t, dstbuf_ref)[DST_BUF_SIZE];
  62         DECLARE_ALIGNED(8, uint8_t, dstbuf_new)[DST_BUF_SIZE];
  63         DECLARE_ALIGNED(8, uint8_t, srcbuf)[SRC_BUF_SIZE];
  64
  65         randomize_buffer(dstbuf_ref);
  66         randomize_buffer(srcbuf);
  67         memcpy(dstbuf_new, dstbuf_ref, sizeof(dstbuf_new));
  68
  69         ptrdiff_t  stride = (rnd() % (MAX_STRIDE / 16) + 1) * 16;
  70         const uint8_t *src = srcbuf + SRC_COLS_LEFT + rnd() % 8U;
  71         uint8_t *dst_new = dstbuf_new, *dst_ref = dstbuf_ref;
  72
  73         if (rnd() & 1) {
  74             dst_new += (BLOCK_SIZE_1D - 1) * stride;
  75             dst_ref += (BLOCK_SIZE_1D - 1) * stride;
  76             src     += (SRC_ROWS - 1) * stride;
  77             stride  *= -1;
  78         }
  79         src += SRC_ROWS_ABOVE * stride;
  80
  81         unsigned select = rnd() % FF_ARRAY_ELEMS(vp6_block_copy_filter);
  82         unsigned x8 = 1 + rnd() % (FF_ARRAY_ELEMS(vp6_block_copy_filter[0]) - 1);
  83         unsigned y8 = 1 + rnd() % (FF_ARRAY_ELEMS(vp6_block_copy_filter[0]) - 1);
  84         const int16_t *h_weights = vp6_block_copy_filter[select][x8];
  85         const int16_t *v_weights = vp6_block_copy_filter[select][y8];
  86
  87         call_ref(dst_ref, src, stride, h_weights, v_weights);
  88         call_new(dst_new, src, stride, h_weights, v_weights);
  89         if (memcmp(dstbuf_new, dstbuf_ref, sizeof(dstbuf_new)))
  90             fail();
  91         bench_new(dst_new, src, stride, h_weights, v_weights);
  92     }
  93 }