+#define MK_IDCT_DC_ADD4_C(name) \
+static void name ## _idct_dc_add4uv_c(uint8_t *dst, int16_t block[4][16], \
+ ptrdiff_t stride) \
+{ \
+ name ## _idct_dc_add_c(dst + stride * 0 + 0, block[0], stride); \
+ name ## _idct_dc_add_c(dst + stride * 0 + 4, block[1], stride); \
+ name ## _idct_dc_add_c(dst + stride * 4 + 0, block[2], stride); \
+ name ## _idct_dc_add_c(dst + stride * 4 + 4, block[3], stride); \
+} \
+ \
+static void name ## _idct_dc_add4y_c(uint8_t *dst, int16_t block[4][16], \
+ ptrdiff_t stride) \
+{ \
+ name ## _idct_dc_add_c(dst + 0, block[0], stride); \
+ name ## _idct_dc_add_c(dst + 4, block[1], stride); \
+ name ## _idct_dc_add_c(dst + 8, block[2], stride); \
+ name ## _idct_dc_add_c(dst + 12, block[3], stride); \
+}
+
+#if CONFIG_VP7_DECODER
+static void vp7_luma_dc_wht_c(int16_t block[4][4][16], int16_t dc[16])
+{
+ int i, a1, b1, c1, d1;
+ int16_t tmp[16];
+
+ for (i = 0; i < 4; i++) {
+ a1 = (dc[i * 4 + 0] + dc[i * 4 + 2]) * 23170;
+ b1 = (dc[i * 4 + 0] - dc[i * 4 + 2]) * 23170;
+ c1 = dc[i * 4 + 1] * 12540 - dc[i * 4 + 3] * 30274;
+ d1 = dc[i * 4 + 1] * 30274 + dc[i * 4 + 3] * 12540;
+ tmp[i * 4 + 0] = (a1 + d1) >> 14;
+ tmp[i * 4 + 3] = (a1 - d1) >> 14;
+ tmp[i * 4 + 1] = (b1 + c1) >> 14;
+ tmp[i * 4 + 2] = (b1 - c1) >> 14;
+ }
+
+ for (i = 0; i < 4; i++) {
+ a1 = (tmp[i + 0] + tmp[i + 8]) * 23170;
+ b1 = (tmp[i + 0] - tmp[i + 8]) * 23170;
+ c1 = tmp[i + 4] * 12540 - tmp[i + 12] * 30274;
+ d1 = tmp[i + 4] * 30274 + tmp[i + 12] * 12540;
+ dc[i * 4 + 0] = 0;
+ dc[i * 4 + 1] = 0;
+ dc[i * 4 + 2] = 0;
+ dc[i * 4 + 3] = 0;
+ block[0][i][0] = (a1 + d1 + 0x20000) >> 18;
+ block[3][i][0] = (a1 - d1 + 0x20000) >> 18;
+ block[1][i][0] = (b1 + c1 + 0x20000) >> 18;
+ block[2][i][0] = (b1 - c1 + 0x20000) >> 18;
+ }
+}
+
+static void vp7_luma_dc_wht_dc_c(int16_t block[4][4][16], int16_t dc[16])
+{
+ int i, val = (23170 * (23170 * dc[0] >> 14) + 0x20000) >> 18;
+ dc[0] = 0;
+
+ for (i = 0; i < 4; i++) {
+ block[i][0][0] = val;
+ block[i][1][0] = val;
+ block[i][2][0] = val;
+ block[i][3][0] = val;
+ }
+}
+
+static void vp7_idct_add_c(uint8_t *dst, int16_t block[16], ptrdiff_t stride)
+{
+ int i, a1, b1, c1, d1;
+ int16_t tmp[16];
+
+ for (i = 0; i < 4; i++) {
+ a1 = (block[i * 4 + 0] + block[i * 4 + 2]) * 23170;
+ b1 = (block[i * 4 + 0] - block[i * 4 + 2]) * 23170;
+ c1 = block[i * 4 + 1] * 12540 - block[i * 4 + 3] * 30274;
+ d1 = block[i * 4 + 1] * 30274 + block[i * 4 + 3] * 12540;
+ block[i * 4 + 0] = 0;
+ block[i * 4 + 1] = 0;
+ block[i * 4 + 2] = 0;
+ block[i * 4 + 3] = 0;
+ tmp[i * 4 + 0] = (a1 + d1) >> 14;
+ tmp[i * 4 + 3] = (a1 - d1) >> 14;
+ tmp[i * 4 + 1] = (b1 + c1) >> 14;
+ tmp[i * 4 + 2] = (b1 - c1) >> 14;
+ }
+
+ for (i = 0; i < 4; i++) {
+ a1 = (tmp[i + 0] + tmp[i + 8]) * 23170;
+ b1 = (tmp[i + 0] - tmp[i + 8]) * 23170;
+ c1 = tmp[i + 4] * 12540 - tmp[i + 12] * 30274;
+ d1 = tmp[i + 4] * 30274 + tmp[i + 12] * 12540;
+ dst[0 * stride + i] = av_clip_uint8(dst[0 * stride + i] +
+ ((a1 + d1 + 0x20000) >> 18));
+ dst[3 * stride + i] = av_clip_uint8(dst[3 * stride + i] +
+ ((a1 - d1 + 0x20000) >> 18));
+ dst[1 * stride + i] = av_clip_uint8(dst[1 * stride + i] +
+ ((b1 + c1 + 0x20000) >> 18));
+ dst[2 * stride + i] = av_clip_uint8(dst[2 * stride + i] +
+ ((b1 - c1 + 0x20000) >> 18));
+ }
+}
+
+static void vp7_idct_dc_add_c(uint8_t *dst, int16_t block[16], ptrdiff_t stride)
+{
+ int i, dc = (23170 * (23170 * block[0] >> 14) + 0x20000) >> 18;
+ block[0] = 0;
+
+ for (i = 0; i < 4; i++) {
+ dst[0] = av_clip_uint8(dst[0] + dc);
+ dst[1] = av_clip_uint8(dst[1] + dc);
+ dst[2] = av_clip_uint8(dst[2] + dc);
+ dst[3] = av_clip_uint8(dst[3] + dc);
+ dst += stride;
+ }
+}
+
+MK_IDCT_DC_ADD4_C(vp7)
+#endif /* CONFIG_VP7_DECODER */
+