libvpx: Pull from upstream

source/libvpx/vpx_dsp/inv_txfm.c

 /*
  *  Copyright (c) 2015 The WebM project authors. All Rights Reserved.
  *
  *  Use of this source code is governed by a BSD-style license
  *  that can be found in the LICENSE file in the root of the source
  *  tree. An additional intellectual property rights grant can be found
  *  in the file PATENTS.  All contributing project authors may
  *  be found in the AUTHORS file in the root of the source tree.
  */
 
 #include <math.h>
 #include <string.h>
 
 #include "vpx_dsp/inv_txfm.h"
 
-void vp9_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int stride) {
+void vpx_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int stride) {
 /* 4-point reversible, orthonormal inverse Walsh-Hadamard in 3.5 adds,
    0.5 shifts per pixel. */
   int i;
   tran_low_t output[16];
   tran_high_t a1, b1, c1, d1, e1;
   const tran_low_t *ip = input;
   tran_low_t *op = output;
 
   for (i = 0; i < 4; i++) {
     a1 = ip[0] >> UNIT_QUANT_SHIFT;
@@ skipping 31 matching lines @@
     dest[stride * 0] = clip_pixel_add(dest[stride * 0], a1);
     dest[stride * 1] = clip_pixel_add(dest[stride * 1], b1);
     dest[stride * 2] = clip_pixel_add(dest[stride * 2], c1);
     dest[stride * 3] = clip_pixel_add(dest[stride * 3], d1);
 
     ip++;
     dest++;
   }
 }
 
-void vp9_iwht4x4_1_add_c(const tran_low_t *in, uint8_t *dest, int dest_stride) {
+void vpx_iwht4x4_1_add_c(const tran_low_t *in, uint8_t *dest, int dest_stride) {
   int i;
   tran_high_t a1, e1;
   tran_low_t tmp[4];
   const tran_low_t *ip = in;
   tran_low_t *op = tmp;
 
   a1 = ip[0] >> UNIT_QUANT_SHIFT;
   e1 = a1 >> 1;
   a1 -= e1;
   op[0] = WRAPLOW(a1, 8);
@@ skipping 25 matching lines @@
   step[2] = WRAPLOW(dct_const_round_shift(temp1), 8);
   step[3] = WRAPLOW(dct_const_round_shift(temp2), 8);
 
   // stage 2
   output[0] = WRAPLOW(step[0] + step[3], 8);
   output[1] = WRAPLOW(step[1] + step[2], 8);
   output[2] = WRAPLOW(step[1] - step[2], 8);
   output[3] = WRAPLOW(step[0] - step[3], 8);
 }
 
-void vp9_idct4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int stride) {
+void vpx_idct4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int stride) {
   tran_low_t out[4 * 4];
   tran_low_t *outptr = out;
   int i, j;
   tran_low_t temp_in[4], temp_out[4];
 
   // Rows
   for (i = 0; i < 4; ++i) {
     idct4_c(input, outptr);
     input += 4;
     outptr += 4;
   }
 
   // Columns
   for (i = 0; i < 4; ++i) {
     for (j = 0; j < 4; ++j)
       temp_in[j] = out[j * 4 + i];
     idct4_c(temp_in, temp_out);
     for (j = 0; j < 4; ++j) {
       dest[j * stride + i] = clip_pixel_add(dest[j * stride + i],
                                             ROUND_POWER_OF_TWO(temp_out[j], 4));
     }
   }
 }
 
-void vp9_idct4x4_1_add_c(const tran_low_t *input, uint8_t *dest,
+void vpx_idct4x4_1_add_c(const tran_low_t *input, uint8_t *dest,
                          int dest_stride) {
   int i;
   tran_high_t a1;
   tran_low_t out = WRAPLOW(dct_const_round_shift(input[0] * cospi_16_64), 8);
   out = WRAPLOW(dct_const_round_shift(out * cospi_16_64), 8);
   a1 = ROUND_POWER_OF_TWO(out, 4);
 
   for (i = 0; i < 4; i++) {
     dest[0] = clip_pixel_add(dest[0], a1);
     dest[1] = clip_pixel_add(dest[1], a1);
@@ skipping 41 matching lines @@
   output[0] = WRAPLOW(step1[0] + step1[7], 8);
   output[1] = WRAPLOW(step1[1] + step1[6], 8);
   output[2] = WRAPLOW(step1[2] + step1[5], 8);
   output[3] = WRAPLOW(step1[3] + step1[4], 8);
   output[4] = WRAPLOW(step1[3] - step1[4], 8);
   output[5] = WRAPLOW(step1[2] - step1[5], 8);
   output[6] = WRAPLOW(step1[1] - step1[6], 8);
   output[7] = WRAPLOW(step1[0] - step1[7], 8);
 }
 
-void vp9_idct8x8_64_add_c(const tran_low_t *input, uint8_t *dest, int stride) {
+void vpx_idct8x8_64_add_c(const tran_low_t *input, uint8_t *dest, int stride) {
   tran_low_t out[8 * 8];
   tran_low_t *outptr = out;
   int i, j;
   tran_low_t temp_in[8], temp_out[8];
 
   // First transform rows
   for (i = 0; i < 8; ++i) {
     idct8_c(input, outptr);
     input += 8;
     outptr += 8;
   }
 
   // Then transform columns
   for (i = 0; i < 8; ++i) {
     for (j = 0; j < 8; ++j)
       temp_in[j] = out[j * 8 + i];
     idct8_c(temp_in, temp_out);
     for (j = 0; j < 8; ++j) {
       dest[j * stride + i] = clip_pixel_add(dest[j * stride + i],
                                             ROUND_POWER_OF_TWO(temp_out[j], 5));
     }
   }
 }
 
-void vp9_idct8x8_1_add_c(const tran_low_t *input, uint8_t *dest, int stride) {
+void vpx_idct8x8_1_add_c(const tran_low_t *input, uint8_t *dest, int stride) {
   int i, j;
   tran_high_t a1;
   tran_low_t out = WRAPLOW(dct_const_round_shift(input[0] * cospi_16_64), 8);
   out = WRAPLOW(dct_const_round_shift(out * cospi_16_64), 8);
   a1 = ROUND_POWER_OF_TWO(out, 5);
   for (j = 0; j < 8; ++j) {
     for (i = 0; i < 8; ++i)
       dest[i] = clip_pixel_add(dest[i], a1);
     dest += stride;
   }
@@ skipping 106 matching lines @@
   output[0] = WRAPLOW(x0, 8);
   output[1] = WRAPLOW(-x4, 8);
   output[2] = WRAPLOW(x6, 8);
   output[3] = WRAPLOW(-x2, 8);
   output[4] = WRAPLOW(x3, 8);
   output[5] = WRAPLOW(-x7, 8);
   output[6] = WRAPLOW(x5, 8);
   output[7] = WRAPLOW(-x1, 8);
 }
 
-void vp9_idct8x8_12_add_c(const tran_low_t *input, uint8_t *dest, int stride) {
+void vpx_idct8x8_12_add_c(const tran_low_t *input, uint8_t *dest, int stride) {
   tran_low_t out[8 * 8] = { 0 };
   tran_low_t *outptr = out;
   int i, j;
   tran_low_t temp_in[8], temp_out[8];
 
   // First transform rows
   // only first 4 row has non-zero coefs
   for (i = 0; i < 4; ++i) {
     idct8_c(input, outptr);
     input += 8;
@@ skipping 170 matching lines @@
   output[8] = WRAPLOW(step2[7] - step2[8], 8);
   output[9] = WRAPLOW(step2[6] - step2[9], 8);
   output[10] = WRAPLOW(step2[5] - step2[10], 8);
   output[11] = WRAPLOW(step2[4] - step2[11], 8);
   output[12] = WRAPLOW(step2[3] - step2[12], 8);
   output[13] = WRAPLOW(step2[2] - step2[13], 8);
   output[14] = WRAPLOW(step2[1] - step2[14], 8);
   output[15] = WRAPLOW(step2[0] - step2[15], 8);
 }
 
-void vp9_idct16x16_256_add_c(const tran_low_t *input, uint8_t *dest,
+void vpx_idct16x16_256_add_c(const tran_low_t *input, uint8_t *dest,
                              int stride) {
   tran_low_t out[16 * 16];
   tran_low_t *outptr = out;
   int i, j;
   tran_low_t temp_in[16], temp_out[16];
 
   // First transform rows
   for (i = 0; i < 16; ++i) {
     idct16_c(input, outptr);
     input += 16;
@@ skipping 177 matching lines @@
   output[8] = WRAPLOW(x3, 8);
   output[9] = WRAPLOW(x11, 8);
   output[10] = WRAPLOW(x15, 8);
   output[11] = WRAPLOW(x7, 8);
   output[12] = WRAPLOW(x5, 8);
   output[13] = WRAPLOW(-x13, 8);
   output[14] = WRAPLOW(x9, 8);
   output[15] = WRAPLOW(-x1, 8);
 }
 
-void vp9_idct16x16_10_add_c(const tran_low_t *input, uint8_t *dest,
+void vpx_idct16x16_10_add_c(const tran_low_t *input, uint8_t *dest,
                             int stride) {
   tran_low_t out[16 * 16] = { 0 };
   tran_low_t *outptr = out;
   int i, j;
   tran_low_t temp_in[16], temp_out[16];
 
   // First transform rows. Since all non-zero dct coefficients are in
   // upper-left 4x4 area, we only need to calculate first 4 rows here.
   for (i = 0; i < 4; ++i) {
     idct16_c(input, outptr);
     input += 16;
     outptr += 16;
   }
 
   // Then transform columns
   for (i = 0; i < 16; ++i) {
     for (j = 0; j < 16; ++j)
       temp_in[j] = out[j*16 + i];
     idct16_c(temp_in, temp_out);
     for (j = 0; j < 16; ++j) {
       dest[j * stride + i] = clip_pixel_add(dest[j * stride + i],
                                             ROUND_POWER_OF_TWO(temp_out[j], 6));
     }
   }
 }
 
-void vp9_idct16x16_1_add_c(const tran_low_t *input, uint8_t *dest, int stride) {
+void vpx_idct16x16_1_add_c(const tran_low_t *input, uint8_t *dest, int stride) {
   int i, j;
   tran_high_t a1;
   tran_low_t out = WRAPLOW(dct_const_round_shift(input[0] * cospi_16_64), 8);
   out = WRAPLOW(dct_const_round_shift(out * cospi_16_64), 8);
   a1 = ROUND_POWER_OF_TWO(out, 6);
   for (j = 0; j < 16; ++j) {
     for (i = 0; i < 16; ++i)
       dest[i] = clip_pixel_add(dest[i], a1);
     dest += stride;
   }
@@ skipping 359 matching lines @@
   output[24] = WRAPLOW(step1[7] - step1[24], 8);
   output[25] = WRAPLOW(step1[6] - step1[25], 8);
   output[26] = WRAPLOW(step1[5] - step1[26], 8);
   output[27] = WRAPLOW(step1[4] - step1[27], 8);
   output[28] = WRAPLOW(step1[3] - step1[28], 8);
   output[29] = WRAPLOW(step1[2] - step1[29], 8);
   output[30] = WRAPLOW(step1[1] - step1[30], 8);
   output[31] = WRAPLOW(step1[0] - step1[31], 8);
 }
 
-void vp9_idct32x32_1024_add_c(const tran_low_t *input, uint8_t *dest,
+void vpx_idct32x32_1024_add_c(const tran_low_t *input, uint8_t *dest,
                               int stride) {
   tran_low_t out[32 * 32];
   tran_low_t *outptr = out;
   int i, j;
   tran_low_t temp_in[32], temp_out[32];
 
   // Rows
   for (i = 0; i < 32; ++i) {
     int16_t zero_coeff[16];
     for (j = 0; j < 16; ++j)
@@ skipping 18 matching lines @@
     for (j = 0; j < 32; ++j)
       temp_in[j] = out[j * 32 + i];
     idct32_c(temp_in, temp_out);
     for (j = 0; j < 32; ++j) {
       dest[j * stride + i] = clip_pixel_add(dest[j * stride + i],
                                             ROUND_POWER_OF_TWO(temp_out[j], 6));
     }
   }
 }
 
-void vp9_idct32x32_34_add_c(const tran_low_t *input, uint8_t *dest,
+void vpx_idct32x32_34_add_c(const tran_low_t *input, uint8_t *dest,
                             int stride) {
   tran_low_t out[32 * 32] = {0};
   tran_low_t *outptr = out;
   int i, j;
   tran_low_t temp_in[32], temp_out[32];
 
   // Rows
   // only upper-left 8x8 has non-zero coeff
   for (i = 0; i < 8; ++i) {
     idct32_c(input, outptr);
     input += 32;
     outptr += 32;
   }
 
   // Columns
   for (i = 0; i < 32; ++i) {
     for (j = 0; j < 32; ++j)
       temp_in[j] = out[j * 32 + i];
     idct32_c(temp_in, temp_out);
     for (j = 0; j < 32; ++j) {
       dest[j * stride + i] = clip_pixel_add(dest[j * stride + i],
                                             ROUND_POWER_OF_TWO(temp_out[j], 6));
     }
   }
 }
 
-void vp9_idct32x32_1_add_c(const tran_low_t *input, uint8_t *dest, int stride) {
+void vpx_idct32x32_1_add_c(const tran_low_t *input, uint8_t *dest, int stride) {
   int i, j;
   tran_high_t a1;
 
   tran_low_t out = WRAPLOW(dct_const_round_shift(input[0] * cospi_16_64), 8);
   out = WRAPLOW(dct_const_round_shift(out * cospi_16_64), 8);
   a1 = ROUND_POWER_OF_TWO(out, 6);
 
   for (j = 0; j < 32; ++j) {
     for (i = 0; i < 32; ++i)
       dest[i] = clip_pixel_add(dest[i], a1);
     dest += stride;
   }
 }
 
 #if CONFIG_VP9_HIGHBITDEPTH
-void vp9_highbd_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest8,
+void vpx_highbd_iwht4x4_16_add_c(const tran_low_t *input, uint8_t *dest8,
                                  int stride, int bd) {
   /* 4-point reversible, orthonormal inverse Walsh-Hadamard in 3.5 adds,
      0.5 shifts per pixel. */
   int i;
   tran_low_t output[16];
   tran_high_t a1, b1, c1, d1, e1;
   const tran_low_t *ip = input;
   tran_low_t *op = output;
   uint16_t *dest = CONVERT_TO_SHORTPTR(dest8);
 
@@ skipping 33 matching lines @@
     dest[stride * 0] = highbd_clip_pixel_add(dest[stride * 0], a1, bd);
     dest[stride * 1] = highbd_clip_pixel_add(dest[stride * 1], b1, bd);
     dest[stride * 2] = highbd_clip_pixel_add(dest[stride * 2], c1, bd);
     dest[stride * 3] = highbd_clip_pixel_add(dest[stride * 3], d1, bd);
 
     ip++;
     dest++;
   }
 }
 
-void vp9_highbd_iwht4x4_1_add_c(const tran_low_t *in, uint8_t *dest8,
+void vpx_highbd_iwht4x4_1_add_c(const tran_low_t *in, uint8_t *dest8,
                                 int dest_stride, int bd) {
   int i;
   tran_high_t a1, e1;
   tran_low_t tmp[4];
   const tran_low_t *ip = in;
   tran_low_t *op = tmp;
   uint16_t *dest = CONVERT_TO_SHORTPTR(dest8);
   (void) bd;
 
   a1 = ip[0] >> UNIT_QUANT_SHIFT;
@@ skipping 12 matching lines @@
         dest[dest_stride * 1], e1, bd);
     dest[dest_stride * 2] = highbd_clip_pixel_add(
         dest[dest_stride * 2], e1, bd);
     dest[dest_stride * 3] = highbd_clip_pixel_add(
         dest[dest_stride * 3], e1, bd);
     ip++;
     dest++;
   }
 }
 
-void vp9_highbd_idct4_c(const tran_low_t *input, tran_low_t *output, int bd) {
+void vpx_highbd_idct4_c(const tran_low_t *input, tran_low_t *output, int bd) {
   tran_low_t step[4];
   tran_high_t temp1, temp2;
   (void) bd;
   // stage 1
   temp1 = (input[0] + input[2]) * cospi_16_64;
   temp2 = (input[0] - input[2]) * cospi_16_64;
   step[0] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);
   step[1] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);
   temp1 = input[1] * cospi_24_64 - input[3] * cospi_8_64;
   temp2 = input[1] * cospi_8_64 + input[3] * cospi_24_64;
   step[2] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);
   step[3] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);
 
   // stage 2
   output[0] = WRAPLOW(step[0] + step[3], bd);
   output[1] = WRAPLOW(step[1] + step[2], bd);
   output[2] = WRAPLOW(step[1] - step[2], bd);
   output[3] = WRAPLOW(step[0] - step[3], bd);
 }
 
-void vp9_highbd_idct4x4_16_add_c(const tran_low_t *input, uint8_t *dest8,
+void vpx_highbd_idct4x4_16_add_c(const tran_low_t *input, uint8_t *dest8,
                                  int stride, int bd) {
   tran_low_t out[4 * 4];
   tran_low_t *outptr = out;
   int i, j;
   tran_low_t temp_in[4], temp_out[4];
   uint16_t *dest = CONVERT_TO_SHORTPTR(dest8);
 
   // Rows
   for (i = 0; i < 4; ++i) {
-    vp9_highbd_idct4_c(input, outptr, bd);
+    vpx_highbd_idct4_c(input, outptr, bd);
     input += 4;
     outptr += 4;
   }
 
   // Columns
   for (i = 0; i < 4; ++i) {
     for (j = 0; j < 4; ++j)
       temp_in[j] = out[j * 4 + i];
-    vp9_highbd_idct4_c(temp_in, temp_out, bd);
+    vpx_highbd_idct4_c(temp_in, temp_out, bd);
     for (j = 0; j < 4; ++j) {
       dest[j * stride + i] = highbd_clip_pixel_add(
           dest[j * stride + i], ROUND_POWER_OF_TWO(temp_out[j], 4), bd);
     }
   }
 }
 
-void vp9_highbd_idct4x4_1_add_c(const tran_low_t *input, uint8_t *dest8,
+void vpx_highbd_idct4x4_1_add_c(const tran_low_t *input, uint8_t *dest8,
                                 int dest_stride, int bd) {
   int i;
   tran_high_t a1;
   tran_low_t out = WRAPLOW(
       highbd_dct_const_round_shift(input[0] * cospi_16_64, bd), bd);
   uint16_t *dest = CONVERT_TO_SHORTPTR(dest8);
 
   out = WRAPLOW(highbd_dct_const_round_shift(out * cospi_16_64, bd), bd);
   a1 = ROUND_POWER_OF_TWO(out, 4);
 
   for (i = 0; i < 4; i++) {
     dest[0] = highbd_clip_pixel_add(dest[0], a1, bd);
     dest[1] = highbd_clip_pixel_add(dest[1], a1, bd);
     dest[2] = highbd_clip_pixel_add(dest[2], a1, bd);
     dest[3] = highbd_clip_pixel_add(dest[3], a1, bd);
     dest += dest_stride;
   }
 }
 
-void vp9_highbd_idct8_c(const tran_low_t *input, tran_low_t *output, int bd) {
+void vpx_highbd_idct8_c(const tran_low_t *input, tran_low_t *output, int bd) {
   tran_low_t step1[8], step2[8];
   tran_high_t temp1, temp2;
   // stage 1
   step1[0] = input[0];
   step1[2] = input[4];
   step1[1] = input[2];
   step1[3] = input[6];
   temp1 = input[1] * cospi_28_64 - input[7] * cospi_4_64;
   temp2 = input[1] * cospi_4_64 + input[7] * cospi_28_64;
   step1[4] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);
   step1[7] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);
   temp1 = input[5] * cospi_12_64 - input[3] * cospi_20_64;
   temp2 = input[5] * cospi_20_64 + input[3] * cospi_12_64;
   step1[5] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);
   step1[6] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);
 
   // stage 2 & stage 3 - even half
-  vp9_highbd_idct4_c(step1, step1, bd);
+  vpx_highbd_idct4_c(step1, step1, bd);
 
   // stage 2 - odd half
   step2[4] = WRAPLOW(step1[4] + step1[5], bd);
   step2[5] = WRAPLOW(step1[4] - step1[5], bd);
   step2[6] = WRAPLOW(-step1[6] + step1[7], bd);
   step2[7] = WRAPLOW(step1[6] + step1[7], bd);
 
   // stage 3 - odd half
   step1[4] = step2[4];
   temp1 = (step2[6] - step2[5]) * cospi_16_64;
   temp2 = (step2[5] + step2[6]) * cospi_16_64;
   step1[5] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);
   step1[6] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);
   step1[7] = step2[7];
 
   // stage 4
   output[0] = WRAPLOW(step1[0] + step1[7], bd);
   output[1] = WRAPLOW(step1[1] + step1[6], bd);
   output[2] = WRAPLOW(step1[2] + step1[5], bd);
   output[3] = WRAPLOW(step1[3] + step1[4], bd);
   output[4] = WRAPLOW(step1[3] - step1[4], bd);
   output[5] = WRAPLOW(step1[2] - step1[5], bd);
   output[6] = WRAPLOW(step1[1] - step1[6], bd);
   output[7] = WRAPLOW(step1[0] - step1[7], bd);
 }
 
-void vp9_highbd_idct8x8_64_add_c(const tran_low_t *input, uint8_t *dest8,
+void vpx_highbd_idct8x8_64_add_c(const tran_low_t *input, uint8_t *dest8,
                                  int stride, int bd) {
   tran_low_t out[8 * 8];
   tran_low_t *outptr = out;
   int i, j;
   tran_low_t temp_in[8], temp_out[8];
   uint16_t *dest = CONVERT_TO_SHORTPTR(dest8);
 
   // First transform rows.
   for (i = 0; i < 8; ++i) {
-    vp9_highbd_idct8_c(input, outptr, bd);
+    vpx_highbd_idct8_c(input, outptr, bd);
     input += 8;
     outptr += 8;
   }
 
   // Then transform columns.
   for (i = 0; i < 8; ++i) {
     for (j = 0; j < 8; ++j)
       temp_in[j] = out[j * 8 + i];
-    vp9_highbd_idct8_c(temp_in, temp_out, bd);
+    vpx_highbd_idct8_c(temp_in, temp_out, bd);
     for (j = 0; j < 8; ++j) {
       dest[j * stride + i] = highbd_clip_pixel_add(
           dest[j * stride + i], ROUND_POWER_OF_TWO(temp_out[j], 5), bd);
     }
   }
 }
 
-void vp9_highbd_idct8x8_1_add_c(const tran_low_t *input, uint8_t *dest8,
+void vpx_highbd_idct8x8_1_add_c(const tran_low_t *input, uint8_t *dest8,
                                 int stride, int bd) {
   int i, j;
   tran_high_t a1;
   tran_low_t out = WRAPLOW(
       highbd_dct_const_round_shift(input[0] * cospi_16_64, bd), bd);
   uint16_t *dest = CONVERT_TO_SHORTPTR(dest8);
   out = WRAPLOW(highbd_dct_const_round_shift(out * cospi_16_64, bd), bd);
   a1 = ROUND_POWER_OF_TWO(out, 5);
   for (j = 0; j < 8; ++j) {
     for (i = 0; i < 8; ++i)
       dest[i] = highbd_clip_pixel_add(dest[i], a1, bd);
     dest += stride;
   }
 }
 
-void highbd_iadst4_c(const tran_low_t *input, tran_low_t *output, int bd) {
+void vpx_highbd_iadst4_c(const tran_low_t *input, tran_low_t *output, int bd) {
   tran_high_t s0, s1, s2, s3, s4, s5, s6, s7;
 
   tran_low_t x0 = input[0];
   tran_low_t x1 = input[1];
   tran_low_t x2 = input[2];
   tran_low_t x3 = input[3];
   (void) bd;
 
   if (!(x0 | x1 | x2 | x3)) {
     memset(output, 0, 4 * sizeof(*output));
@@ skipping 17 matching lines @@
   // 1-D transform scaling factor is sqrt(2).
   // The overall dynamic range is 14b (input) + 14b (multiplication scaling)
   // + 1b (addition) = 29b.
   // Hence the output bit depth is 15b.
   output[0] = WRAPLOW(highbd_dct_const_round_shift(s0 + s3, bd), bd);
   output[1] = WRAPLOW(highbd_dct_const_round_shift(s1 + s3, bd), bd);
   output[2] = WRAPLOW(highbd_dct_const_round_shift(s2, bd), bd);
   output[3] = WRAPLOW(highbd_dct_const_round_shift(s0 + s1 - s3, bd), bd);
 }
 
-void highbd_iadst8_c(const tran_low_t *input, tran_low_t *output, int bd) {
+void vpx_highbd_iadst8_c(const tran_low_t *input, tran_low_t *output, int bd) {
   tran_high_t s0, s1, s2, s3, s4, s5, s6, s7;
 
   tran_low_t x0 = input[7];
   tran_low_t x1 = input[0];
   tran_low_t x2 = input[5];
   tran_low_t x3 = input[2];
   tran_low_t x4 = input[3];
   tran_low_t x5 = input[4];
   tran_low_t x6 = input[1];
   tran_low_t x7 = input[6];
@@ skipping 56 matching lines @@
   output[0] = WRAPLOW(x0, bd);
   output[1] = WRAPLOW(-x4, bd);
   output[2] = WRAPLOW(x6, bd);
   output[3] = WRAPLOW(-x2, bd);
   output[4] = WRAPLOW(x3, bd);
   output[5] = WRAPLOW(-x7, bd);
   output[6] = WRAPLOW(x5, bd);
   output[7] = WRAPLOW(-x1, bd);
 }
 
-void vp9_highbd_idct8x8_10_add_c(const tran_low_t *input, uint8_t *dest8,
+void vpx_highbd_idct8x8_10_add_c(const tran_low_t *input, uint8_t *dest8,
                                  int stride, int bd) {
   tran_low_t out[8 * 8] = { 0 };
   tran_low_t *outptr = out;
   int i, j;
   tran_low_t temp_in[8], temp_out[8];
   uint16_t *dest = CONVERT_TO_SHORTPTR(dest8);
 
   // First transform rows.
   // Only first 4 row has non-zero coefs.
   for (i = 0; i < 4; ++i) {
-    vp9_highbd_idct8_c(input, outptr, bd);
+    vpx_highbd_idct8_c(input, outptr, bd);
     input += 8;
     outptr += 8;
   }
   // Then transform columns.
   for (i = 0; i < 8; ++i) {
     for (j = 0; j < 8; ++j)
       temp_in[j] = out[j * 8 + i];
-    vp9_highbd_idct8_c(temp_in, temp_out, bd);
+    vpx_highbd_idct8_c(temp_in, temp_out, bd);
     for (j = 0; j < 8; ++j) {
       dest[j * stride + i] = highbd_clip_pixel_add(
           dest[j * stride + i], ROUND_POWER_OF_TWO(temp_out[j], 5), bd);
     }
   }
 }
 
-void vp9_highbd_idct16_c(const tran_low_t *input, tran_low_t *output, int bd) {
+void vpx_highbd_idct16_c(const tran_low_t *input, tran_low_t *output, int bd) {
   tran_low_t step1[16], step2[16];
   tran_high_t temp1, temp2;
   (void) bd;
 
   // stage 1
   step1[0] = input[0/2];
   step1[1] = input[16/2];
   step1[2] = input[8/2];
   step1[3] = input[24/2];
   step1[4] = input[4/2];
@@ skipping 145 matching lines @@
   output[8] = WRAPLOW(step2[7] - step2[8], bd);
   output[9] = WRAPLOW(step2[6] - step2[9], bd);
   output[10] = WRAPLOW(step2[5] - step2[10], bd);
   output[11] = WRAPLOW(step2[4] - step2[11], bd);
   output[12] = WRAPLOW(step2[3] - step2[12], bd);
   output[13] = WRAPLOW(step2[2] - step2[13], bd);
   output[14] = WRAPLOW(step2[1] - step2[14], bd);
   output[15] = WRAPLOW(step2[0] - step2[15], bd);
 }
 
-void vp9_highbd_idct16x16_256_add_c(const tran_low_t *input, uint8_t *dest8,
+void vpx_highbd_idct16x16_256_add_c(const tran_low_t *input, uint8_t *dest8,
                                     int stride, int bd) {
   tran_low_t out[16 * 16];
   tran_low_t *outptr = out;
   int i, j;
   tran_low_t temp_in[16], temp_out[16];
   uint16_t *dest = CONVERT_TO_SHORTPTR(dest8);
 
   // First transform rows.
   for (i = 0; i < 16; ++i) {
-    vp9_highbd_idct16_c(input, outptr, bd);
+    vpx_highbd_idct16_c(input, outptr, bd);
     input += 16;
     outptr += 16;
   }
 
   // Then transform columns.
   for (i = 0; i < 16; ++i) {
     for (j = 0; j < 16; ++j)
       temp_in[j] = out[j * 16 + i];
-    vp9_highbd_idct16_c(temp_in, temp_out, bd);
+    vpx_highbd_idct16_c(temp_in, temp_out, bd);
     for (j = 0; j < 16; ++j) {
       dest[j * stride + i] = highbd_clip_pixel_add(
           dest[j * stride + i], ROUND_POWER_OF_TWO(temp_out[j], 6), bd);
     }
   }
 }
 
-void highbd_iadst16_c(const tran_low_t *input, tran_low_t *output, int bd) {
+void vpx_highbd_iadst16_c(const tran_low_t *input, tran_low_t *output, int bd) {
   tran_high_t s0, s1, s2, s3, s4, s5, s6, s7, s8;
   tran_high_t s9, s10, s11, s12, s13, s14, s15;
 
   tran_low_t x0 = input[15];
   tran_low_t x1 = input[0];
   tran_low_t x2 = input[13];
   tran_low_t x3 = input[2];
   tran_low_t x4 = input[11];
   tran_low_t x5 = input[4];
   tran_low_t x6 = input[9];
@@ skipping 149 matching lines @@
   output[8] = WRAPLOW(x3, bd);
   output[9] = WRAPLOW(x11, bd);
   output[10] = WRAPLOW(x15, bd);
   output[11] = WRAPLOW(x7, bd);
   output[12] = WRAPLOW(x5, bd);
   output[13] = WRAPLOW(-x13, bd);
   output[14] = WRAPLOW(x9, bd);
   output[15] = WRAPLOW(-x1, bd);
 }
 
-void vp9_highbd_idct16x16_10_add_c(const tran_low_t *input, uint8_t *dest8,
+void vpx_highbd_idct16x16_10_add_c(const tran_low_t *input, uint8_t *dest8,
                                    int stride, int bd) {
   tran_low_t out[16 * 16] = { 0 };
   tran_low_t *outptr = out;
   int i, j;
   tran_low_t temp_in[16], temp_out[16];
   uint16_t *dest = CONVERT_TO_SHORTPTR(dest8);
 
   // First transform rows. Since all non-zero dct coefficients are in
   // upper-left 4x4 area, we only need to calculate first 4 rows here.
   for (i = 0; i < 4; ++i) {
-    vp9_highbd_idct16_c(input, outptr, bd);
+    vpx_highbd_idct16_c(input, outptr, bd);
     input += 16;
     outptr += 16;
   }
 
   // Then transform columns.
   for (i = 0; i < 16; ++i) {
     for (j = 0; j < 16; ++j)
       temp_in[j] = out[j*16 + i];
-    vp9_highbd_idct16_c(temp_in, temp_out, bd);
+    vpx_highbd_idct16_c(temp_in, temp_out, bd);
     for (j = 0; j < 16; ++j) {
       dest[j * stride + i] = highbd_clip_pixel_add(
           dest[j * stride + i], ROUND_POWER_OF_TWO(temp_out[j], 6), bd);
     }
   }
 }
 
-void vp9_highbd_idct16x16_1_add_c(const tran_low_t *input, uint8_t *dest8,
+void vpx_highbd_idct16x16_1_add_c(const tran_low_t *input, uint8_t *dest8,
                                   int stride, int bd) {
   int i, j;
   tran_high_t a1;
   tran_low_t out = WRAPLOW(
       highbd_dct_const_round_shift(input[0] * cospi_16_64, bd), bd);
   uint16_t *dest = CONVERT_TO_SHORTPTR(dest8);
 
   out = WRAPLOW(highbd_dct_const_round_shift(out * cospi_16_64, bd), bd);
   a1 = ROUND_POWER_OF_TWO(out, 6);
   for (j = 0; j < 16; ++j) {
     for (i = 0; i < 16; ++i)
       dest[i] = highbd_clip_pixel_add(dest[i], a1, bd);
     dest += stride;
   }
 }
 
-void highbd_idct32_c(const tran_low_t *input, tran_low_t *output, int bd) {
+static void highbd_idct32_c(const tran_low_t *input,
+                            tran_low_t *output, int bd) {
   tran_low_t step1[32], step2[32];
   tran_high_t temp1, temp2;
   (void) bd;
 
   // stage 1
   step1[0] = input[0];
   step1[1] = input[16];
   step1[2] = input[8];
   step1[3] = input[24];
   step1[4] = input[4];
@@ skipping 347 matching lines @@
   output[24] = WRAPLOW(step1[7] - step1[24], bd);
   output[25] = WRAPLOW(step1[6] - step1[25], bd);
   output[26] = WRAPLOW(step1[5] - step1[26], bd);
   output[27] = WRAPLOW(step1[4] - step1[27], bd);
   output[28] = WRAPLOW(step1[3] - step1[28], bd);
   output[29] = WRAPLOW(step1[2] - step1[29], bd);
   output[30] = WRAPLOW(step1[1] - step1[30], bd);
   output[31] = WRAPLOW(step1[0] - step1[31], bd);
 }
 
-void vp9_highbd_idct32x32_1024_add_c(const tran_low_t *input, uint8_t *dest8,
+void vpx_highbd_idct32x32_1024_add_c(const tran_low_t *input, uint8_t *dest8,
                                      int stride, int bd) {
   tran_low_t out[32 * 32];
   tran_low_t *outptr = out;
   int i, j;
   tran_low_t temp_in[32], temp_out[32];
   uint16_t *dest = CONVERT_TO_SHORTPTR(dest8);
 
   // Rows
   for (i = 0; i < 32; ++i) {
     tran_low_t zero_coeff[16];
@@ skipping 19 matching lines @@
     for (j = 0; j < 32; ++j)
       temp_in[j] = out[j * 32 + i];
     highbd_idct32_c(temp_in, temp_out, bd);
     for (j = 0; j < 32; ++j) {
       dest[j * stride + i] = highbd_clip_pixel_add(
           dest[j * stride + i], ROUND_POWER_OF_TWO(temp_out[j], 6), bd);
     }
   }
 }
 
-void vp9_highbd_idct32x32_34_add_c(const tran_low_t *input, uint8_t *dest8,
+void vpx_highbd_idct32x32_34_add_c(const tran_low_t *input, uint8_t *dest8,
                                    int stride, int bd) {
   tran_low_t out[32 * 32] = {0};
   tran_low_t *outptr = out;
   int i, j;
   tran_low_t temp_in[32], temp_out[32];
   uint16_t *dest = CONVERT_TO_SHORTPTR(dest8);
 
   // Rows
   // Only upper-left 8x8 has non-zero coeff.
   for (i = 0; i < 8; ++i) {
     highbd_idct32_c(input, outptr, bd);
     input += 32;
     outptr += 32;
   }
   // Columns
   for (i = 0; i < 32; ++i) {
     for (j = 0; j < 32; ++j)
       temp_in[j] = out[j * 32 + i];
     highbd_idct32_c(temp_in, temp_out, bd);
     for (j = 0; j < 32; ++j) {
       dest[j * stride + i] = highbd_clip_pixel_add(
           dest[j * stride + i], ROUND_POWER_OF_TWO(temp_out[j], 6), bd);
     }
   }
 }
 
-void vp9_highbd_idct32x32_1_add_c(const tran_low_t *input, uint8_t *dest8,
+void vpx_highbd_idct32x32_1_add_c(const tran_low_t *input, uint8_t *dest8,
                                   int stride, int bd) {
   int i, j;
   int a1;
   uint16_t *dest = CONVERT_TO_SHORTPTR(dest8);
 
   tran_low_t out = WRAPLOW(
       highbd_dct_const_round_shift(input[0] * cospi_16_64, bd), bd);
   out = WRAPLOW(highbd_dct_const_round_shift(out * cospi_16_64, bd), bd);
   a1 = ROUND_POWER_OF_TWO(out, 6);
 
   for (j = 0; j < 32; ++j) {
     for (i = 0; i < 32; ++i)
       dest[i] = highbd_clip_pixel_add(dest[i], a1, bd);
     dest += stride;
   }
 }
 #endif  // CONFIG_VP9_HIGHBITDEPTH