shithub: libvpx

Info  •  Files  •  Log  •  Branches

Merge "Move range checks into WRAPLOW"

--- a/vp10/common/vp10_inv_txfm.c

+++ b/vp10/common/vp10_inv_txfm.c

@@ -35,10 +35,10 @@

     c1 = e1 - c1;

     a1 -= b1;

     d1 += c1;

-    op[0] = WRAPLOW(a1, 8);

-    op[1] = WRAPLOW(b1, 8);

-    op[2] = WRAPLOW(c1, 8);

-    op[3] = WRAPLOW(d1, 8);

+    op[0] = WRAPLOW(a1);

+    op[1] = WRAPLOW(b1);

+    op[2] = WRAPLOW(c1);

+    op[3] = WRAPLOW(d1);

     ip += 4;

     op += 4;

   }

@@ -78,8 +78,8 @@

   a1 = ip[0] >> UNIT_QUANT_SHIFT;

   e1 = a1 >> 1;

   a1 -= e1;

-  op[0] = WRAPLOW(a1, 8);

-  op[1] = op[2] = op[3] = WRAPLOW(e1, 8);

+  op[0] = WRAPLOW(a1);

+  op[1] = op[2] = op[3] = WRAPLOW(e1);

   ip = tmp;

   for (i = 0; i < 4; i++) {

@@ -100,18 +100,18 @@

   // stage 1

   temp1 = (input[0] + input[2]) * cospi_16_64;

   temp2 = (input[0] - input[2]) * cospi_16_64;

-  step[0] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step[1] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step[0] = WRAPLOW(dct_const_round_shift(temp1));

+  step[1] = WRAPLOW(dct_const_round_shift(temp2));

   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(dct_const_round_shift(temp1), 8);

-  step[3] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step[2] = WRAPLOW(dct_const_round_shift(temp1));

+  step[3] = WRAPLOW(dct_const_round_shift(temp2));

   // 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);

+  output[0] = WRAPLOW(step[0] + step[3]);

+  output[1] = WRAPLOW(step[1] + step[2]);

+  output[2] = WRAPLOW(step[1] - step[2]);

+  output[3] = WRAPLOW(step[0] - step[3]);

 }

 void vp10_idct4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int stride) {

@@ -143,8 +143,8 @@

                          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);

+  tran_low_t out = WRAPLOW(dct_const_round_shift(input[0] * cospi_16_64));

+  out = WRAPLOW(dct_const_round_shift(out * cospi_16_64));

   a1 = ROUND_POWER_OF_TWO(out, 4);

   for (i = 0; i < 4; i++) {

@@ -166,48 +166,48 @@

   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(dct_const_round_shift(temp1), 8);

-  step1[7] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[4] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[7] = WRAPLOW(dct_const_round_shift(temp2));

   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(dct_const_round_shift(temp1), 8);

-  step1[6] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[5] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[6] = WRAPLOW(dct_const_round_shift(temp2));

   // stage 2

   temp1 = (step1[0] + step1[2]) * cospi_16_64;

   temp2 = (step1[0] - step1[2]) * cospi_16_64;

-  step2[0] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[1] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[0] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[1] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step1[1] * cospi_24_64 - step1[3] * cospi_8_64;

   temp2 = step1[1] * cospi_8_64 + step1[3] * cospi_24_64;

-  step2[2] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[3] = WRAPLOW(dct_const_round_shift(temp2), 8);

-  step2[4] = WRAPLOW(step1[4] + step1[5], 8);

-  step2[5] = WRAPLOW(step1[4] - step1[5], 8);

-  step2[6] = WRAPLOW(-step1[6] + step1[7], 8);

-  step2[7] = WRAPLOW(step1[6] + step1[7], 8);

+  step2[2] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[3] = WRAPLOW(dct_const_round_shift(temp2));

+  step2[4] = WRAPLOW(step1[4] + step1[5]);

+  step2[5] = WRAPLOW(step1[4] - step1[5]);

+  step2[6] = WRAPLOW(-step1[6] + step1[7]);

+  step2[7] = WRAPLOW(step1[6] + step1[7]);

   // stage 3

-  step1[0] = WRAPLOW(step2[0] + step2[3], 8);

-  step1[1] = WRAPLOW(step2[1] + step2[2], 8);

-  step1[2] = WRAPLOW(step2[1] - step2[2], 8);

-  step1[3] = WRAPLOW(step2[0] - step2[3], 8);

+  step1[0] = WRAPLOW(step2[0] + step2[3]);

+  step1[1] = WRAPLOW(step2[1] + step2[2]);

+  step1[2] = WRAPLOW(step2[1] - step2[2]);

+  step1[3] = WRAPLOW(step2[0] - step2[3]);

   step1[4] = step2[4];

   temp1 = (step2[6] - step2[5]) * cospi_16_64;

   temp2 = (step2[5] + step2[6]) * cospi_16_64;

-  step1[5] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[6] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[5] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[6] = WRAPLOW(dct_const_round_shift(temp2));

   step1[7] = step2[7];

   // stage 4

-  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);

+  output[0] = WRAPLOW(step1[0] + step1[7]);

+  output[1] = WRAPLOW(step1[1] + step1[6]);

+  output[2] = WRAPLOW(step1[2] + step1[5]);

+  output[3] = WRAPLOW(step1[3] + step1[4]);

+  output[4] = WRAPLOW(step1[3] - step1[4]);

+  output[5] = WRAPLOW(step1[2] - step1[5]);

+  output[6] = WRAPLOW(step1[1] - step1[6]);

+  output[7] = WRAPLOW(step1[0] - step1[7]);

 }

 void vp10_idct8x8_64_add_c(const tran_low_t *input, uint8_t *dest, int stride) {

@@ -238,8 +238,8 @@

 void vp10_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);

+  tran_low_t out = WRAPLOW(dct_const_round_shift(input[0] * cospi_16_64));

+  out = WRAPLOW(dct_const_round_shift(out * cospi_16_64));

   a1 = ROUND_POWER_OF_TWO(out, 5);

   for (j = 0; j < 8; ++j) {

     for (i = 0; i < 8; ++i)

@@ -279,10 +279,10 @@

   // The overall dynamic range is 14b (input) + 14b (multiplication scaling)

   // + 1b (addition) = 29b.

   // Hence the output bit depth is 15b.

-  output[0] = WRAPLOW(dct_const_round_shift(s0 + s3), 8);

-  output[1] = WRAPLOW(dct_const_round_shift(s1 + s3), 8);

-  output[2] = WRAPLOW(dct_const_round_shift(s2), 8);

-  output[3] = WRAPLOW(dct_const_round_shift(s0 + s1 - s3), 8);

+  output[0] = WRAPLOW(dct_const_round_shift(s0 + s3));

+  output[1] = WRAPLOW(dct_const_round_shift(s1 + s3));

+  output[2] = WRAPLOW(dct_const_round_shift(s2));

+  output[3] = WRAPLOW(dct_const_round_shift(s0 + s1 - s3));

 }

 void vp10_iadst8_c(const tran_low_t *input, tran_low_t *output) {

@@ -313,14 +313,14 @@

   s6 = (int)(cospi_26_64 * x6 + cospi_6_64  * x7);

   s7 = (int)(cospi_6_64  * x6 - cospi_26_64 * x7);

-  x0 = WRAPLOW(dct_const_round_shift(s0 + s4), 8);

-  x1 = WRAPLOW(dct_const_round_shift(s1 + s5), 8);

-  x2 = WRAPLOW(dct_const_round_shift(s2 + s6), 8);

-  x3 = WRAPLOW(dct_const_round_shift(s3 + s7), 8);

-  x4 = WRAPLOW(dct_const_round_shift(s0 - s4), 8);

-  x5 = WRAPLOW(dct_const_round_shift(s1 - s5), 8);

-  x6 = WRAPLOW(dct_const_round_shift(s2 - s6), 8);

-  x7 = WRAPLOW(dct_const_round_shift(s3 - s7), 8);

+  x0 = WRAPLOW(dct_const_round_shift(s0 + s4));

+  x1 = WRAPLOW(dct_const_round_shift(s1 + s5));

+  x2 = WRAPLOW(dct_const_round_shift(s2 + s6));

+  x3 = WRAPLOW(dct_const_round_shift(s3 + s7));

+  x4 = WRAPLOW(dct_const_round_shift(s0 - s4));

+  x5 = WRAPLOW(dct_const_round_shift(s1 - s5));

+  x6 = WRAPLOW(dct_const_round_shift(s2 - s6));

+  x7 = WRAPLOW(dct_const_round_shift(s3 - s7));

   // stage 2

   s0 = (int)x0;

@@ -332,14 +332,14 @@

   s6 = (int)(-cospi_24_64 * x6 + cospi_8_64 * x7);

   s7 = (int)(cospi_8_64 * x6 + cospi_24_64 * x7);

-  x0 = WRAPLOW(s0 + s2, 8);

-  x1 = WRAPLOW(s1 + s3, 8);

-  x2 = WRAPLOW(s0 - s2, 8);

-  x3 = WRAPLOW(s1 - s3, 8);

-  x4 = WRAPLOW(dct_const_round_shift(s4 + s6), 8);

-  x5 = WRAPLOW(dct_const_round_shift(s5 + s7), 8);

-  x6 = WRAPLOW(dct_const_round_shift(s4 - s6), 8);

-  x7 = WRAPLOW(dct_const_round_shift(s5 - s7), 8);

+  x0 = WRAPLOW(s0 + s2);

+  x1 = WRAPLOW(s1 + s3);

+  x2 = WRAPLOW(s0 - s2);

+  x3 = WRAPLOW(s1 - s3);

+  x4 = WRAPLOW(dct_const_round_shift(s4 + s6));

+  x5 = WRAPLOW(dct_const_round_shift(s5 + s7));

+  x6 = WRAPLOW(dct_const_round_shift(s4 - s6));

+  x7 = WRAPLOW(dct_const_round_shift(s5 - s7));

   // stage 3

   s2 = (int)(cospi_16_64 * (x2 + x3));

@@ -347,19 +347,19 @@

   s6 = (int)(cospi_16_64 * (x6 + x7));

   s7 = (int)(cospi_16_64 * (x6 - x7));

-  x2 = WRAPLOW(dct_const_round_shift(s2), 8);

-  x3 = WRAPLOW(dct_const_round_shift(s3), 8);

-  x6 = WRAPLOW(dct_const_round_shift(s6), 8);

-  x7 = WRAPLOW(dct_const_round_shift(s7), 8);

+  x2 = WRAPLOW(dct_const_round_shift(s2));

+  x3 = WRAPLOW(dct_const_round_shift(s3));

+  x6 = WRAPLOW(dct_const_round_shift(s6));

+  x7 = WRAPLOW(dct_const_round_shift(s7));

-  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);

+  output[0] = WRAPLOW(x0);

+  output[1] = WRAPLOW(-x4);

+  output[2] = WRAPLOW(x6);

+  output[3] = WRAPLOW(-x2);

+  output[4] = WRAPLOW(x3);

+  output[5] = WRAPLOW(-x7);

+  output[6] = WRAPLOW(x5);

+  output[7] = WRAPLOW(-x1);

 }

 void vp10_idct8x8_12_add_c(const tran_low_t *input, uint8_t *dest, int stride) {

@@ -422,23 +422,23 @@

   temp1 = step1[8] * cospi_30_64 - step1[15] * cospi_2_64;

   temp2 = step1[8] * cospi_2_64 + step1[15] * cospi_30_64;

-  step2[8] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[15] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[8] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[15] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step1[9] * cospi_14_64 - step1[14] * cospi_18_64;

   temp2 = step1[9] * cospi_18_64 + step1[14] * cospi_14_64;

-  step2[9] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[14] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[9] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[14] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step1[10] * cospi_22_64 - step1[13] * cospi_10_64;

   temp2 = step1[10] * cospi_10_64 + step1[13] * cospi_22_64;

-  step2[10] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[13] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[10] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[13] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step1[11] * cospi_6_64 - step1[12] * cospi_26_64;

   temp2 = step1[11] * cospi_26_64 + step1[12] * cospi_6_64;

-  step2[11] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[12] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[11] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[12] = WRAPLOW(dct_const_round_shift(temp2));

   // stage 3

   step1[0] = step2[0];

@@ -448,109 +448,109 @@

   temp1 = step2[4] * cospi_28_64 - step2[7] * cospi_4_64;

   temp2 = step2[4] * cospi_4_64 + step2[7] * cospi_28_64;

-  step1[4] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[7] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[4] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[7] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step2[5] * cospi_12_64 - step2[6] * cospi_20_64;

   temp2 = step2[5] * cospi_20_64 + step2[6] * cospi_12_64;

-  step1[5] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[6] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[5] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[6] = WRAPLOW(dct_const_round_shift(temp2));

-  step1[8] = WRAPLOW(step2[8] + step2[9], 8);

-  step1[9] = WRAPLOW(step2[8] - step2[9], 8);

-  step1[10] = WRAPLOW(-step2[10] + step2[11], 8);

-  step1[11] = WRAPLOW(step2[10] + step2[11], 8);

-  step1[12] = WRAPLOW(step2[12] + step2[13], 8);

-  step1[13] = WRAPLOW(step2[12] - step2[13], 8);

-  step1[14] = WRAPLOW(-step2[14] + step2[15], 8);

-  step1[15] = WRAPLOW(step2[14] + step2[15], 8);

+  step1[8] = WRAPLOW(step2[8] + step2[9]);

+  step1[9] = WRAPLOW(step2[8] - step2[9]);

+  step1[10] = WRAPLOW(-step2[10] + step2[11]);

+  step1[11] = WRAPLOW(step2[10] + step2[11]);

+  step1[12] = WRAPLOW(step2[12] + step2[13]);

+  step1[13] = WRAPLOW(step2[12] - step2[13]);

+  step1[14] = WRAPLOW(-step2[14] + step2[15]);

+  step1[15] = WRAPLOW(step2[14] + step2[15]);

   // stage 4

   temp1 = (step1[0] + step1[1]) * cospi_16_64;

   temp2 = (step1[0] - step1[1]) * cospi_16_64;

-  step2[0] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[1] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[0] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[1] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step1[2] * cospi_24_64 - step1[3] * cospi_8_64;

   temp2 = step1[2] * cospi_8_64 + step1[3] * cospi_24_64;

-  step2[2] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[3] = WRAPLOW(dct_const_round_shift(temp2), 8);

-  step2[4] = WRAPLOW(step1[4] + step1[5], 8);

-  step2[5] = WRAPLOW(step1[4] - step1[5], 8);

-  step2[6] = WRAPLOW(-step1[6] + step1[7], 8);

-  step2[7] = WRAPLOW(step1[6] + step1[7], 8);

+  step2[2] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[3] = WRAPLOW(dct_const_round_shift(temp2));

+  step2[4] = WRAPLOW(step1[4] + step1[5]);

+  step2[5] = WRAPLOW(step1[4] - step1[5]);

+  step2[6] = WRAPLOW(-step1[6] + step1[7]);

+  step2[7] = WRAPLOW(step1[6] + step1[7]);

   step2[8] = step1[8];

   step2[15] = step1[15];

   temp1 = -step1[9] * cospi_8_64 + step1[14] * cospi_24_64;

   temp2 = step1[9] * cospi_24_64 + step1[14] * cospi_8_64;

-  step2[9] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[14] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[9] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[14] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = -step1[10] * cospi_24_64 - step1[13] * cospi_8_64;

   temp2 = -step1[10] * cospi_8_64 + step1[13] * cospi_24_64;

-  step2[10] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[13] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[10] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[13] = WRAPLOW(dct_const_round_shift(temp2));

   step2[11] = step1[11];

   step2[12] = step1[12];

   // stage 5

-  step1[0] = WRAPLOW(step2[0] + step2[3], 8);

-  step1[1] = WRAPLOW(step2[1] + step2[2], 8);

-  step1[2] = WRAPLOW(step2[1] - step2[2], 8);

-  step1[3] = WRAPLOW(step2[0] - step2[3], 8);

+  step1[0] = WRAPLOW(step2[0] + step2[3]);

+  step1[1] = WRAPLOW(step2[1] + step2[2]);

+  step1[2] = WRAPLOW(step2[1] - step2[2]);

+  step1[3] = WRAPLOW(step2[0] - step2[3]);

   step1[4] = step2[4];

   temp1 = (step2[6] - step2[5]) * cospi_16_64;

   temp2 = (step2[5] + step2[6]) * cospi_16_64;

-  step1[5] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[6] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[5] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[6] = WRAPLOW(dct_const_round_shift(temp2));

   step1[7] = step2[7];

-  step1[8] = WRAPLOW(step2[8] + step2[11], 8);

-  step1[9] = WRAPLOW(step2[9] + step2[10], 8);

-  step1[10] = WRAPLOW(step2[9] - step2[10], 8);

-  step1[11] = WRAPLOW(step2[8] - step2[11], 8);

-  step1[12] = WRAPLOW(-step2[12] + step2[15], 8);

-  step1[13] = WRAPLOW(-step2[13] + step2[14], 8);

-  step1[14] = WRAPLOW(step2[13] + step2[14], 8);

-  step1[15] = WRAPLOW(step2[12] + step2[15], 8);

+  step1[8] = WRAPLOW(step2[8] + step2[11]);

+  step1[9] = WRAPLOW(step2[9] + step2[10]);

+  step1[10] = WRAPLOW(step2[9] - step2[10]);

+  step1[11] = WRAPLOW(step2[8] - step2[11]);

+  step1[12] = WRAPLOW(-step2[12] + step2[15]);

+  step1[13] = WRAPLOW(-step2[13] + step2[14]);

+  step1[14] = WRAPLOW(step2[13] + step2[14]);

+  step1[15] = WRAPLOW(step2[12] + step2[15]);

   // stage 6

-  step2[0] = WRAPLOW(step1[0] + step1[7], 8);

-  step2[1] = WRAPLOW(step1[1] + step1[6], 8);

-  step2[2] = WRAPLOW(step1[2] + step1[5], 8);

-  step2[3] = WRAPLOW(step1[3] + step1[4], 8);

-  step2[4] = WRAPLOW(step1[3] - step1[4], 8);

-  step2[5] = WRAPLOW(step1[2] - step1[5], 8);

-  step2[6] = WRAPLOW(step1[1] - step1[6], 8);

-  step2[7] = WRAPLOW(step1[0] - step1[7], 8);

+  step2[0] = WRAPLOW(step1[0] + step1[7]);

+  step2[1] = WRAPLOW(step1[1] + step1[6]);

+  step2[2] = WRAPLOW(step1[2] + step1[5]);

+  step2[3] = WRAPLOW(step1[3] + step1[4]);

+  step2[4] = WRAPLOW(step1[3] - step1[4]);

+  step2[5] = WRAPLOW(step1[2] - step1[5]);

+  step2[6] = WRAPLOW(step1[1] - step1[6]);

+  step2[7] = WRAPLOW(step1[0] - step1[7]);

   step2[8] = step1[8];

   step2[9] = step1[9];

   temp1 = (-step1[10] + step1[13]) * cospi_16_64;

   temp2 = (step1[10] + step1[13]) * cospi_16_64;

-  step2[10] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[13] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[10] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[13] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = (-step1[11] + step1[12]) * cospi_16_64;

   temp2 = (step1[11] + step1[12]) * cospi_16_64;

-  step2[11] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[12] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[11] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[12] = WRAPLOW(dct_const_round_shift(temp2));

   step2[14] = step1[14];

   step2[15] = step1[15];

   // stage 7

-  output[0] = WRAPLOW(step2[0] + step2[15], 8);

-  output[1] = WRAPLOW(step2[1] + step2[14], 8);

-  output[2] = WRAPLOW(step2[2] + step2[13], 8);

-  output[3] = WRAPLOW(step2[3] + step2[12], 8);

-  output[4] = WRAPLOW(step2[4] + step2[11], 8);

-  output[5] = WRAPLOW(step2[5] + step2[10], 8);

-  output[6] = WRAPLOW(step2[6] + step2[9], 8);

-  output[7] = WRAPLOW(step2[7] + step2[8], 8);

-  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);

+  output[0] = WRAPLOW(step2[0] + step2[15]);

+  output[1] = WRAPLOW(step2[1] + step2[14]);

+  output[2] = WRAPLOW(step2[2] + step2[13]);

+  output[3] = WRAPLOW(step2[3] + step2[12]);

+  output[4] = WRAPLOW(step2[4] + step2[11]);

+  output[5] = WRAPLOW(step2[5] + step2[10]);

+  output[6] = WRAPLOW(step2[6] + step2[9]);

+  output[7] = WRAPLOW(step2[7] + step2[8]);

+  output[8] = WRAPLOW(step2[7] - step2[8]);

+  output[9] = WRAPLOW(step2[6] - step2[9]);

+  output[10] = WRAPLOW(step2[5] - step2[10]);

+  output[11] = WRAPLOW(step2[4] - step2[11]);

+  output[12] = WRAPLOW(step2[3] - step2[12]);

+  output[13] = WRAPLOW(step2[2] - step2[13]);

+  output[14] = WRAPLOW(step2[1] - step2[14]);

+  output[15] = WRAPLOW(step2[0] - step2[15]);

 }

 void vp10_idct16x16_256_add_c(const tran_low_t *input, uint8_t *dest,

@@ -627,22 +627,22 @@

   s14 = x14 * cospi_29_64 + x15 * cospi_3_64;

   s15 = x14 * cospi_3_64  - x15 * cospi_29_64;

-  x0 = WRAPLOW(dct_const_round_shift(s0 + s8), 8);

-  x1 = WRAPLOW(dct_const_round_shift(s1 + s9), 8);

-  x2 = WRAPLOW(dct_const_round_shift(s2 + s10), 8);

-  x3 = WRAPLOW(dct_const_round_shift(s3 + s11), 8);

-  x4 = WRAPLOW(dct_const_round_shift(s4 + s12), 8);

-  x5 = WRAPLOW(dct_const_round_shift(s5 + s13), 8);

-  x6 = WRAPLOW(dct_const_round_shift(s6 + s14), 8);

-  x7 = WRAPLOW(dct_const_round_shift(s7 + s15), 8);

-  x8 = WRAPLOW(dct_const_round_shift(s0 - s8), 8);

-  x9 = WRAPLOW(dct_const_round_shift(s1 - s9), 8);

-  x10 = WRAPLOW(dct_const_round_shift(s2 - s10), 8);

-  x11 = WRAPLOW(dct_const_round_shift(s3 - s11), 8);

-  x12 = WRAPLOW(dct_const_round_shift(s4 - s12), 8);

-  x13 = WRAPLOW(dct_const_round_shift(s5 - s13), 8);

-  x14 = WRAPLOW(dct_const_round_shift(s6 - s14), 8);

-  x15 = WRAPLOW(dct_const_round_shift(s7 - s15), 8);

+  x0 = WRAPLOW(dct_const_round_shift(s0 + s8));

+  x1 = WRAPLOW(dct_const_round_shift(s1 + s9));

+  x2 = WRAPLOW(dct_const_round_shift(s2 + s10));

+  x3 = WRAPLOW(dct_const_round_shift(s3 + s11));

+  x4 = WRAPLOW(dct_const_round_shift(s4 + s12));

+  x5 = WRAPLOW(dct_const_round_shift(s5 + s13));

+  x6 = WRAPLOW(dct_const_round_shift(s6 + s14));

+  x7 = WRAPLOW(dct_const_round_shift(s7 + s15));

+  x8 = WRAPLOW(dct_const_round_shift(s0 - s8));

+  x9 = WRAPLOW(dct_const_round_shift(s1 - s9));

+  x10 = WRAPLOW(dct_const_round_shift(s2 - s10));

+  x11 = WRAPLOW(dct_const_round_shift(s3 - s11));

+  x12 = WRAPLOW(dct_const_round_shift(s4 - s12));

+  x13 = WRAPLOW(dct_const_round_shift(s5 - s13));

+  x14 = WRAPLOW(dct_const_round_shift(s6 - s14));

+  x15 = WRAPLOW(dct_const_round_shift(s7 - s15));

   // stage 2

   s0 = x0;

@@ -662,22 +662,22 @@

   s14 = - x14 * cospi_12_64 + x15 * cospi_20_64;

   s15 =   x14 * cospi_20_64 + x15 * cospi_12_64;

-  x0 = WRAPLOW(s0 + s4, 8);

-  x1 = WRAPLOW(s1 + s5, 8);

-  x2 = WRAPLOW(s2 + s6, 8);

-  x3 = WRAPLOW(s3 + s7, 8);

-  x4 = WRAPLOW(s0 - s4, 8);

-  x5 = WRAPLOW(s1 - s5, 8);

-  x6 = WRAPLOW(s2 - s6, 8);

-  x7 = WRAPLOW(s3 - s7, 8);

-  x8 = WRAPLOW(dct_const_round_shift(s8 + s12), 8);

-  x9 = WRAPLOW(dct_const_round_shift(s9 + s13), 8);

-  x10 = WRAPLOW(dct_const_round_shift(s10 + s14), 8);

-  x11 = WRAPLOW(dct_const_round_shift(s11 + s15), 8);

-  x12 = WRAPLOW(dct_const_round_shift(s8 - s12), 8);

-  x13 = WRAPLOW(dct_const_round_shift(s9 - s13), 8);

-  x14 = WRAPLOW(dct_const_round_shift(s10 - s14), 8);

-  x15 = WRAPLOW(dct_const_round_shift(s11 - s15), 8);

+  x0 = WRAPLOW(s0 + s4);

+  x1 = WRAPLOW(s1 + s5);

+  x2 = WRAPLOW(s2 + s6);

+  x3 = WRAPLOW(s3 + s7);

+  x4 = WRAPLOW(s0 - s4);

+  x5 = WRAPLOW(s1 - s5);

+  x6 = WRAPLOW(s2 - s6);

+  x7 = WRAPLOW(s3 - s7);

+  x8 = WRAPLOW(dct_const_round_shift(s8 + s12));

+  x9 = WRAPLOW(dct_const_round_shift(s9 + s13));

+  x10 = WRAPLOW(dct_const_round_shift(s10 + s14));

+  x11 = WRAPLOW(dct_const_round_shift(s11 + s15));

+  x12 = WRAPLOW(dct_const_round_shift(s8 - s12));

+  x13 = WRAPLOW(dct_const_round_shift(s9 - s13));

+  x14 = WRAPLOW(dct_const_round_shift(s10 - s14));

+  x15 = WRAPLOW(dct_const_round_shift(s11 - s15));

   // stage 3

   s0 = x0;

@@ -697,22 +697,22 @@

   s14 = - x14 * cospi_24_64 + x15 * cospi_8_64;

   s15 =   x14 * cospi_8_64  + x15 * cospi_24_64;

-  x0 = WRAPLOW(check_range(s0 + s2), 8);

-  x1 = WRAPLOW(check_range(s1 + s3), 8);

-  x2 = WRAPLOW(check_range(s0 - s2), 8);

-  x3 = WRAPLOW(check_range(s1 - s3), 8);

-  x4 = WRAPLOW(dct_const_round_shift(s4 + s6), 8);

-  x5 = WRAPLOW(dct_const_round_shift(s5 + s7), 8);

-  x6 = WRAPLOW(dct_const_round_shift(s4 - s6), 8);

-  x7 = WRAPLOW(dct_const_round_shift(s5 - s7), 8);

-  x8 = WRAPLOW(check_range(s8 + s10), 8);

-  x9 = WRAPLOW(check_range(s9 + s11), 8);

-  x10 = WRAPLOW(check_range(s8 - s10), 8);

-  x11 = WRAPLOW(check_range(s9 - s11), 8);

-  x12 = WRAPLOW(dct_const_round_shift(s12 + s14), 8);

-  x13 = WRAPLOW(dct_const_round_shift(s13 + s15), 8);

-  x14 = WRAPLOW(dct_const_round_shift(s12 - s14), 8);

-  x15 = WRAPLOW(dct_const_round_shift(s13 - s15), 8);

+  x0 = WRAPLOW(s0 + s2);

+  x1 = WRAPLOW(s1 + s3);

+  x2 = WRAPLOW(s0 - s2);

+  x3 = WRAPLOW(s1 - s3);

+  x4 = WRAPLOW(dct_const_round_shift(s4 + s6));

+  x5 = WRAPLOW(dct_const_round_shift(s5 + s7));

+  x6 = WRAPLOW(dct_const_round_shift(s4 - s6));

+  x7 = WRAPLOW(dct_const_round_shift(s5 - s7));

+  x8 = WRAPLOW(s8 + s10);

+  x9 = WRAPLOW(s9 + s11);

+  x10 = WRAPLOW(s8 - s10);

+  x11 = WRAPLOW(s9 - s11);

+  x12 = WRAPLOW(dct_const_round_shift(s12 + s14));

+  x13 = WRAPLOW(dct_const_round_shift(s13 + s15));

+  x14 = WRAPLOW(dct_const_round_shift(s12 - s14));

+  x15 = WRAPLOW(dct_const_round_shift(s13 - s15));

   // stage 4

   s2 = (- cospi_16_64) * (x2 + x3);

@@ -724,31 +724,31 @@

   s14 = (- cospi_16_64) * (x14 + x15);

   s15 = cospi_16_64 * (x14 - x15);

-  x2 = WRAPLOW(dct_const_round_shift(s2), 8);

-  x3 = WRAPLOW(dct_const_round_shift(s3), 8);

-  x6 = WRAPLOW(dct_const_round_shift(s6), 8);

-  x7 = WRAPLOW(dct_const_round_shift(s7), 8);

-  x10 = WRAPLOW(dct_const_round_shift(s10), 8);

-  x11 = WRAPLOW(dct_const_round_shift(s11), 8);

-  x14 = WRAPLOW(dct_const_round_shift(s14), 8);

-  x15 = WRAPLOW(dct_const_round_shift(s15), 8);

+  x2 = WRAPLOW(dct_const_round_shift(s2));

+  x3 = WRAPLOW(dct_const_round_shift(s3));

+  x6 = WRAPLOW(dct_const_round_shift(s6));

+  x7 = WRAPLOW(dct_const_round_shift(s7));

+  x10 = WRAPLOW(dct_const_round_shift(s10));

+  x11 = WRAPLOW(dct_const_round_shift(s11));

+  x14 = WRAPLOW(dct_const_round_shift(s14));

+  x15 = WRAPLOW(dct_const_round_shift(s15));

-  output[0] = WRAPLOW(x0, 8);

-  output[1] = WRAPLOW(-x8, 8);

-  output[2] = WRAPLOW(x12, 8);

-  output[3] = WRAPLOW(-x4, 8);

-  output[4] = WRAPLOW(x6, 8);

-  output[5] = WRAPLOW(x14, 8);

-  output[6] = WRAPLOW(x10, 8);

-  output[7] = WRAPLOW(x2, 8);

-  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);

+  output[0] = WRAPLOW(x0);

+  output[1] = WRAPLOW(-x8);

+  output[2] = WRAPLOW(x12);

+  output[3] = WRAPLOW(-x4);

+  output[4] = WRAPLOW(x6);

+  output[5] = WRAPLOW(x14);

+  output[6] = WRAPLOW(x10);

+  output[7] = WRAPLOW(x2);

+  output[8] = WRAPLOW(x3);

+  output[9] = WRAPLOW(x11);

+  output[10] = WRAPLOW(x15);

+  output[11] = WRAPLOW(x7);

+  output[12] = WRAPLOW(x5);

+  output[13] = WRAPLOW(-x13);

+  output[14] = WRAPLOW(x9);

+  output[15] = WRAPLOW(-x1);

 }

 void vp10_idct16x16_10_add_c(const tran_low_t *input, uint8_t *dest,

@@ -784,8 +784,8 @@

                             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);

+  tran_low_t out = WRAPLOW(dct_const_round_shift(input[0] * cospi_16_64));

+  out = WRAPLOW(dct_const_round_shift(out * cospi_16_64));

   a1 = ROUND_POWER_OF_TWO(out, 6);

   for (j = 0; j < 16; ++j) {

     for (i = 0; i < 16; ++i)

@@ -818,43 +818,43 @@

   temp1 = input[1] * cospi_31_64 - input[31] * cospi_1_64;

   temp2 = input[1] * cospi_1_64 + input[31] * cospi_31_64;

-  step1[16] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[31] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[16] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[31] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = input[17] * cospi_15_64 - input[15] * cospi_17_64;

   temp2 = input[17] * cospi_17_64 + input[15] * cospi_15_64;

-  step1[17] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[30] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[17] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[30] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = input[9] * cospi_23_64 - input[23] * cospi_9_64;

   temp2 = input[9] * cospi_9_64 + input[23] * cospi_23_64;

-  step1[18] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[29] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[18] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[29] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = input[25] * cospi_7_64 - input[7] * cospi_25_64;

   temp2 = input[25] * cospi_25_64 + input[7] * cospi_7_64;

-  step1[19] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[28] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[19] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[28] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = input[5] * cospi_27_64 - input[27] * cospi_5_64;

   temp2 = input[5] * cospi_5_64 + input[27] * cospi_27_64;

-  step1[20] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[27] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[20] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[27] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = input[21] * cospi_11_64 - input[11] * cospi_21_64;

   temp2 = input[21] * cospi_21_64 + input[11] * cospi_11_64;

-  step1[21] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[26] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[21] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[26] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = input[13] * cospi_19_64 - input[19] * cospi_13_64;

   temp2 = input[13] * cospi_13_64 + input[19] * cospi_19_64;

-  step1[22] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[25] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[22] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[25] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = input[29] * cospi_3_64 - input[3] * cospi_29_64;

   temp2 = input[29] * cospi_29_64 + input[3] * cospi_3_64;

-  step1[23] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[24] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[23] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[24] = WRAPLOW(dct_const_round_shift(temp2));

   // stage 2

   step2[0] = step1[0];

@@ -868,40 +868,40 @@

   temp1 = step1[8] * cospi_30_64 - step1[15] * cospi_2_64;

   temp2 = step1[8] * cospi_2_64 + step1[15] * cospi_30_64;

-  step2[8] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[15] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[8] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[15] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step1[9] * cospi_14_64 - step1[14] * cospi_18_64;

   temp2 = step1[9] * cospi_18_64 + step1[14] * cospi_14_64;

-  step2[9] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[14] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[9] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[14] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step1[10] * cospi_22_64 - step1[13] * cospi_10_64;

   temp2 = step1[10] * cospi_10_64 + step1[13] * cospi_22_64;

-  step2[10] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[13] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[10] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[13] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step1[11] * cospi_6_64 - step1[12] * cospi_26_64;

   temp2 = step1[11] * cospi_26_64 + step1[12] * cospi_6_64;

-  step2[11] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[12] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[11] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[12] = WRAPLOW(dct_const_round_shift(temp2));

-  step2[16] = WRAPLOW(step1[16] + step1[17], 8);

-  step2[17] = WRAPLOW(step1[16] - step1[17], 8);

-  step2[18] = WRAPLOW(-step1[18] + step1[19], 8);

-  step2[19] = WRAPLOW(step1[18] + step1[19], 8);

-  step2[20] = WRAPLOW(step1[20] + step1[21], 8);

-  step2[21] = WRAPLOW(step1[20] - step1[21], 8);

-  step2[22] = WRAPLOW(-step1[22] + step1[23], 8);

-  step2[23] = WRAPLOW(step1[22] + step1[23], 8);

-  step2[24] = WRAPLOW(step1[24] + step1[25], 8);

-  step2[25] = WRAPLOW(step1[24] - step1[25], 8);

-  step2[26] = WRAPLOW(-step1[26] + step1[27], 8);

-  step2[27] = WRAPLOW(step1[26] + step1[27], 8);

-  step2[28] = WRAPLOW(step1[28] + step1[29], 8);

-  step2[29] = WRAPLOW(step1[28] - step1[29], 8);

-  step2[30] = WRAPLOW(-step1[30] + step1[31], 8);

-  step2[31] = WRAPLOW(step1[30] + step1[31], 8);

+  step2[16] = WRAPLOW(step1[16] + step1[17]);

+  step2[17] = WRAPLOW(step1[16] - step1[17]);

+  step2[18] = WRAPLOW(-step1[18] + step1[19]);

+  step2[19] = WRAPLOW(step1[18] + step1[19]);

+  step2[20] = WRAPLOW(step1[20] + step1[21]);

+  step2[21] = WRAPLOW(step1[20] - step1[21]);

+  step2[22] = WRAPLOW(-step1[22] + step1[23]);

+  step2[23] = WRAPLOW(step1[22] + step1[23]);

+  step2[24] = WRAPLOW(step1[24] + step1[25]);

+  step2[25] = WRAPLOW(step1[24] - step1[25]);

+  step2[26] = WRAPLOW(-step1[26] + step1[27]);

+  step2[27] = WRAPLOW(step1[26] + step1[27]);

+  step2[28] = WRAPLOW(step1[28] + step1[29]);

+  step2[29] = WRAPLOW(step1[28] - step1[29]);

+  step2[30] = WRAPLOW(-step1[30] + step1[31]);

+  step2[31] = WRAPLOW(step1[30] + step1[31]);

   // stage 3

   step1[0] = step2[0];

@@ -911,42 +911,42 @@

   temp1 = step2[4] * cospi_28_64 - step2[7] * cospi_4_64;

   temp2 = step2[4] * cospi_4_64 + step2[7] * cospi_28_64;

-  step1[4] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[7] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[4] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[7] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step2[5] * cospi_12_64 - step2[6] * cospi_20_64;

   temp2 = step2[5] * cospi_20_64 + step2[6] * cospi_12_64;

-  step1[5] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[6] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[5] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[6] = WRAPLOW(dct_const_round_shift(temp2));

-  step1[8] = WRAPLOW(step2[8] + step2[9], 8);

-  step1[9] = WRAPLOW(step2[8] - step2[9], 8);

-  step1[10] = WRAPLOW(-step2[10] + step2[11], 8);

-  step1[11] = WRAPLOW(step2[10] + step2[11], 8);

-  step1[12] = WRAPLOW(step2[12] + step2[13], 8);

-  step1[13] = WRAPLOW(step2[12] - step2[13], 8);

-  step1[14] = WRAPLOW(-step2[14] + step2[15], 8);

-  step1[15] = WRAPLOW(step2[14] + step2[15], 8);

+  step1[8] = WRAPLOW(step2[8] + step2[9]);

+  step1[9] = WRAPLOW(step2[8] - step2[9]);

+  step1[10] = WRAPLOW(-step2[10] + step2[11]);

+  step1[11] = WRAPLOW(step2[10] + step2[11]);

+  step1[12] = WRAPLOW(step2[12] + step2[13]);

+  step1[13] = WRAPLOW(step2[12] - step2[13]);

+  step1[14] = WRAPLOW(-step2[14] + step2[15]);

+  step1[15] = WRAPLOW(step2[14] + step2[15]);

   step1[16] = step2[16];

   step1[31] = step2[31];

   temp1 = -step2[17] * cospi_4_64 + step2[30] * cospi_28_64;

   temp2 = step2[17] * cospi_28_64 + step2[30] * cospi_4_64;

-  step1[17] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[30] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[17] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[30] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = -step2[18] * cospi_28_64 - step2[29] * cospi_4_64;

   temp2 = -step2[18] * cospi_4_64 + step2[29] * cospi_28_64;

-  step1[18] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[29] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[18] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[29] = WRAPLOW(dct_const_round_shift(temp2));

   step1[19] = step2[19];

   step1[20] = step2[20];

   temp1 = -step2[21] * cospi_20_64 + step2[26] * cospi_12_64;

   temp2 = step2[21] * cospi_12_64 + step2[26] * cospi_20_64;

-  step1[21] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[26] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[21] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[26] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = -step2[22] * cospi_12_64 - step2[25] * cospi_20_64;

   temp2 = -step2[22] * cospi_20_64 + step2[25] * cospi_12_64;

-  step1[22] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[25] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[22] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[25] = WRAPLOW(dct_const_round_shift(temp2));

   step1[23] = step2[23];

   step1[24] = step2[24];

   step1[27] = step2[27];

@@ -955,87 +955,87 @@

   // stage 4

   temp1 = (step1[0] + step1[1]) * cospi_16_64;

   temp2 = (step1[0] - step1[1]) * cospi_16_64;

-  step2[0] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[1] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[0] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[1] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step1[2] * cospi_24_64 - step1[3] * cospi_8_64;

   temp2 = step1[2] * cospi_8_64 + step1[3] * cospi_24_64;

-  step2[2] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[3] = WRAPLOW(dct_const_round_shift(temp2), 8);

-  step2[4] = WRAPLOW(step1[4] + step1[5], 8);

-  step2[5] = WRAPLOW(step1[4] - step1[5], 8);

-  step2[6] = WRAPLOW(-step1[6] + step1[7], 8);

-  step2[7] = WRAPLOW(step1[6] + step1[7], 8);

+  step2[2] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[3] = WRAPLOW(dct_const_round_shift(temp2));

+  step2[4] = WRAPLOW(step1[4] + step1[5]);

+  step2[5] = WRAPLOW(step1[4] - step1[5]);

+  step2[6] = WRAPLOW(-step1[6] + step1[7]);

+  step2[7] = WRAPLOW(step1[6] + step1[7]);

   step2[8] = step1[8];

   step2[15] = step1[15];

   temp1 = -step1[9] * cospi_8_64 + step1[14] * cospi_24_64;

   temp2 = step1[9] * cospi_24_64 + step1[14] * cospi_8_64;

-  step2[9] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[14] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[9] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[14] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = -step1[10] * cospi_24_64 - step1[13] * cospi_8_64;

   temp2 = -step1[10] * cospi_8_64 + step1[13] * cospi_24_64;

-  step2[10] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[13] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[10] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[13] = WRAPLOW(dct_const_round_shift(temp2));

   step2[11] = step1[11];

   step2[12] = step1[12];

-  step2[16] = WRAPLOW(step1[16] + step1[19], 8);

-  step2[17] = WRAPLOW(step1[17] + step1[18], 8);

-  step2[18] = WRAPLOW(step1[17] - step1[18], 8);

-  step2[19] = WRAPLOW(step1[16] - step1[19], 8);

-  step2[20] = WRAPLOW(-step1[20] + step1[23], 8);

-  step2[21] = WRAPLOW(-step1[21] + step1[22], 8);

-  step2[22] = WRAPLOW(step1[21] + step1[22], 8);

-  step2[23] = WRAPLOW(step1[20] + step1[23], 8);

+  step2[16] = WRAPLOW(step1[16] + step1[19]);

+  step2[17] = WRAPLOW(step1[17] + step1[18]);

+  step2[18] = WRAPLOW(step1[17] - step1[18]);

+  step2[19] = WRAPLOW(step1[16] - step1[19]);

+  step2[20] = WRAPLOW(-step1[20] + step1[23]);

+  step2[21] = WRAPLOW(-step1[21] + step1[22]);

+  step2[22] = WRAPLOW(step1[21] + step1[22]);

+  step2[23] = WRAPLOW(step1[20] + step1[23]);

-  step2[24] = WRAPLOW(step1[24] + step1[27], 8);

-  step2[25] = WRAPLOW(step1[25] + step1[26], 8);

-  step2[26] = WRAPLOW(step1[25] - step1[26], 8);

-  step2[27] = WRAPLOW(step1[24] - step1[27], 8);

-  step2[28] = WRAPLOW(-step1[28] + step1[31], 8);

-  step2[29] = WRAPLOW(-step1[29] + step1[30], 8);

-  step2[30] = WRAPLOW(step1[29] + step1[30], 8);

-  step2[31] = WRAPLOW(step1[28] + step1[31], 8);

+  step2[24] = WRAPLOW(step1[24] + step1[27]);

+  step2[25] = WRAPLOW(step1[25] + step1[26]);

+  step2[26] = WRAPLOW(step1[25] - step1[26]);

+  step2[27] = WRAPLOW(step1[24] - step1[27]);

+  step2[28] = WRAPLOW(-step1[28] + step1[31]);

+  step2[29] = WRAPLOW(-step1[29] + step1[30]);

+  step2[30] = WRAPLOW(step1[29] + step1[30]);

+  step2[31] = WRAPLOW(step1[28] + step1[31]);

   // stage 5

-  step1[0] = WRAPLOW(step2[0] + step2[3], 8);

-  step1[1] = WRAPLOW(step2[1] + step2[2], 8);

-  step1[2] = WRAPLOW(step2[1] - step2[2], 8);

-  step1[3] = WRAPLOW(step2[0] - step2[3], 8);

+  step1[0] = WRAPLOW(step2[0] + step2[3]);

+  step1[1] = WRAPLOW(step2[1] + step2[2]);

+  step1[2] = WRAPLOW(step2[1] - step2[2]);

+  step1[3] = WRAPLOW(step2[0] - step2[3]);

   step1[4] = step2[4];

   temp1 = (step2[6] - step2[5]) * cospi_16_64;

   temp2 = (step2[5] + step2[6]) * cospi_16_64;

-  step1[5] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[6] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[5] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[6] = WRAPLOW(dct_const_round_shift(temp2));

   step1[7] = step2[7];

-  step1[8] = WRAPLOW(step2[8] + step2[11], 8);

-  step1[9] = WRAPLOW(step2[9] + step2[10], 8);

-  step1[10] = WRAPLOW(step2[9] - step2[10], 8);

-  step1[11] = WRAPLOW(step2[8] - step2[11], 8);

-  step1[12] = WRAPLOW(-step2[12] + step2[15], 8);

-  step1[13] = WRAPLOW(-step2[13] + step2[14], 8);

-  step1[14] = WRAPLOW(step2[13] + step2[14], 8);

-  step1[15] = WRAPLOW(step2[12] + step2[15], 8);

+  step1[8] = WRAPLOW(step2[8] + step2[11]);

+  step1[9] = WRAPLOW(step2[9] + step2[10]);

+  step1[10] = WRAPLOW(step2[9] - step2[10]);

+  step1[11] = WRAPLOW(step2[8] - step2[11]);

+  step1[12] = WRAPLOW(-step2[12] + step2[15]);

+  step1[13] = WRAPLOW(-step2[13] + step2[14]);

+  step1[14] = WRAPLOW(step2[13] + step2[14]);

+  step1[15] = WRAPLOW(step2[12] + step2[15]);

   step1[16] = step2[16];

   step1[17] = step2[17];

   temp1 = -step2[18] * cospi_8_64 + step2[29] * cospi_24_64;

   temp2 = step2[18] * cospi_24_64 + step2[29] * cospi_8_64;

-  step1[18] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[29] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[18] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[29] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = -step2[19] * cospi_8_64 + step2[28] * cospi_24_64;

   temp2 = step2[19] * cospi_24_64 + step2[28] * cospi_8_64;

-  step1[19] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[28] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[19] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[28] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = -step2[20] * cospi_24_64 - step2[27] * cospi_8_64;

   temp2 = -step2[20] * cospi_8_64 + step2[27] * cospi_24_64;

-  step1[20] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[27] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[20] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[27] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = -step2[21] * cospi_24_64 - step2[26] * cospi_8_64;

   temp2 = -step2[21] * cospi_8_64 + step2[26] * cospi_24_64;

-  step1[21] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[26] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[21] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[26] = WRAPLOW(dct_const_round_shift(temp2));

   step1[22] = step2[22];

   step1[23] = step2[23];

   step1[24] = step2[24];

@@ -1044,62 +1044,62 @@

   step1[31] = step2[31];

   // stage 6

-  step2[0] = WRAPLOW(step1[0] + step1[7], 8);

-  step2[1] = WRAPLOW(step1[1] + step1[6], 8);

-  step2[2] = WRAPLOW(step1[2] + step1[5], 8);

-  step2[3] = WRAPLOW(step1[3] + step1[4], 8);

-  step2[4] = WRAPLOW(step1[3] - step1[4], 8);

-  step2[5] = WRAPLOW(step1[2] - step1[5], 8);

-  step2[6] = WRAPLOW(step1[1] - step1[6], 8);

-  step2[7] = WRAPLOW(step1[0] - step1[7], 8);

+  step2[0] = WRAPLOW(step1[0] + step1[7]);

+  step2[1] = WRAPLOW(step1[1] + step1[6]);

+  step2[2] = WRAPLOW(step1[2] + step1[5]);

+  step2[3] = WRAPLOW(step1[3] + step1[4]);

+  step2[4] = WRAPLOW(step1[3] - step1[4]);

+  step2[5] = WRAPLOW(step1[2] - step1[5]);

+  step2[6] = WRAPLOW(step1[1] - step1[6]);

+  step2[7] = WRAPLOW(step1[0] - step1[7]);

   step2[8] = step1[8];

   step2[9] = step1[9];

   temp1 = (-step1[10] + step1[13]) * cospi_16_64;

   temp2 = (step1[10] + step1[13]) * cospi_16_64;

-  step2[10] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[13] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[10] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[13] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = (-step1[11] + step1[12]) * cospi_16_64;

   temp2 = (step1[11] + step1[12]) * cospi_16_64;

-  step2[11] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[12] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[11] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[12] = WRAPLOW(dct_const_round_shift(temp2));

   step2[14] = step1[14];

   step2[15] = step1[15];

-  step2[16] = WRAPLOW(step1[16] + step1[23], 8);

-  step2[17] = WRAPLOW(step1[17] + step1[22], 8);

-  step2[18] = WRAPLOW(step1[18] + step1[21], 8);

-  step2[19] = WRAPLOW(step1[19] + step1[20], 8);

-  step2[20] = WRAPLOW(step1[19] - step1[20], 8);

-  step2[21] = WRAPLOW(step1[18] - step1[21], 8);

-  step2[22] = WRAPLOW(step1[17] - step1[22], 8);

-  step2[23] = WRAPLOW(step1[16] - step1[23], 8);

+  step2[16] = WRAPLOW(step1[16] + step1[23]);

+  step2[17] = WRAPLOW(step1[17] + step1[22]);

+  step2[18] = WRAPLOW(step1[18] + step1[21]);

+  step2[19] = WRAPLOW(step1[19] + step1[20]);

+  step2[20] = WRAPLOW(step1[19] - step1[20]);

+  step2[21] = WRAPLOW(step1[18] - step1[21]);

+  step2[22] = WRAPLOW(step1[17] - step1[22]);

+  step2[23] = WRAPLOW(step1[16] - step1[23]);

-  step2[24] = WRAPLOW(-step1[24] + step1[31], 8);

-  step2[25] = WRAPLOW(-step1[25] + step1[30], 8);

-  step2[26] = WRAPLOW(-step1[26] + step1[29], 8);

-  step2[27] = WRAPLOW(-step1[27] + step1[28], 8);

-  step2[28] = WRAPLOW(step1[27] + step1[28], 8);

-  step2[29] = WRAPLOW(step1[26] + step1[29], 8);

-  step2[30] = WRAPLOW(step1[25] + step1[30], 8);

-  step2[31] = WRAPLOW(step1[24] + step1[31], 8);

+  step2[24] = WRAPLOW(-step1[24] + step1[31]);

+  step2[25] = WRAPLOW(-step1[25] + step1[30]);

+  step2[26] = WRAPLOW(-step1[26] + step1[29]);

+  step2[27] = WRAPLOW(-step1[27] + step1[28]);

+  step2[28] = WRAPLOW(step1[27] + step1[28]);

+  step2[29] = WRAPLOW(step1[26] + step1[29]);

+  step2[30] = WRAPLOW(step1[25] + step1[30]);

+  step2[31] = WRAPLOW(step1[24] + step1[31]);

   // stage 7

-  step1[0] = WRAPLOW(step2[0] + step2[15], 8);

-  step1[1] = WRAPLOW(step2[1] + step2[14], 8);

-  step1[2] = WRAPLOW(step2[2] + step2[13], 8);

-  step1[3] = WRAPLOW(step2[3] + step2[12], 8);

-  step1[4] = WRAPLOW(step2[4] + step2[11], 8);

-  step1[5] = WRAPLOW(step2[5] + step2[10], 8);

-  step1[6] = WRAPLOW(step2[6] + step2[9], 8);

-  step1[7] = WRAPLOW(step2[7] + step2[8], 8);

-  step1[8] = WRAPLOW(step2[7] - step2[8], 8);

-  step1[9] = WRAPLOW(step2[6] - step2[9], 8);

-  step1[10] = WRAPLOW(step2[5] - step2[10], 8);

-  step1[11] = WRAPLOW(step2[4] - step2[11], 8);

-  step1[12] = WRAPLOW(step2[3] - step2[12], 8);

-  step1[13] = WRAPLOW(step2[2] - step2[13], 8);

-  step1[14] = WRAPLOW(step2[1] - step2[14], 8);

-  step1[15] = WRAPLOW(step2[0] - step2[15], 8);

+  step1[0] = WRAPLOW(step2[0] + step2[15]);

+  step1[1] = WRAPLOW(step2[1] + step2[14]);

+  step1[2] = WRAPLOW(step2[2] + step2[13]);

+  step1[3] = WRAPLOW(step2[3] + step2[12]);

+  step1[4] = WRAPLOW(step2[4] + step2[11]);

+  step1[5] = WRAPLOW(step2[5] + step2[10]);

+  step1[6] = WRAPLOW(step2[6] + step2[9]);

+  step1[7] = WRAPLOW(step2[7] + step2[8]);

+  step1[8] = WRAPLOW(step2[7] - step2[8]);

+  step1[9] = WRAPLOW(step2[6] - step2[9]);

+  step1[10] = WRAPLOW(step2[5] - step2[10]);

+  step1[11] = WRAPLOW(step2[4] - step2[11]);

+  step1[12] = WRAPLOW(step2[3] - step2[12]);

+  step1[13] = WRAPLOW(step2[2] - step2[13]);

+  step1[14] = WRAPLOW(step2[1] - step2[14]);

+  step1[15] = WRAPLOW(step2[0] - step2[15]);

   step1[16] = step2[16];

   step1[17] = step2[17];

@@ -1107,20 +1107,20 @@

   step1[19] = step2[19];

   temp1 = (-step2[20] + step2[27]) * cospi_16_64;

   temp2 = (step2[20] + step2[27]) * cospi_16_64;

-  step1[20] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[27] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[20] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[27] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = (-step2[21] + step2[26]) * cospi_16_64;

   temp2 = (step2[21] + step2[26]) * cospi_16_64;

-  step1[21] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[26] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[21] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[26] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = (-step2[22] + step2[25]) * cospi_16_64;

   temp2 = (step2[22] + step2[25]) * cospi_16_64;

-  step1[22] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[25] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[22] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[25] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = (-step2[23] + step2[24]) * cospi_16_64;

   temp2 = (step2[23] + step2[24]) * cospi_16_64;

-  step1[23] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[24] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[23] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[24] = WRAPLOW(dct_const_round_shift(temp2));

   step1[28] = step2[28];

   step1[29] = step2[29];

   step1[30] = step2[30];

@@ -1127,38 +1127,38 @@

   step1[31] = step2[31];

   // final stage

-  output[0] = WRAPLOW(step1[0] + step1[31], 8);

-  output[1] = WRAPLOW(step1[1] + step1[30], 8);

-  output[2] = WRAPLOW(step1[2] + step1[29], 8);

-  output[3] = WRAPLOW(step1[3] + step1[28], 8);

-  output[4] = WRAPLOW(step1[4] + step1[27], 8);

-  output[5] = WRAPLOW(step1[5] + step1[26], 8);

-  output[6] = WRAPLOW(step1[6] + step1[25], 8);

-  output[7] = WRAPLOW(step1[7] + step1[24], 8);

-  output[8] = WRAPLOW(step1[8] + step1[23], 8);

-  output[9] = WRAPLOW(step1[9] + step1[22], 8);

-  output[10] = WRAPLOW(step1[10] + step1[21], 8);

-  output[11] = WRAPLOW(step1[11] + step1[20], 8);

-  output[12] = WRAPLOW(step1[12] + step1[19], 8);

-  output[13] = WRAPLOW(step1[13] + step1[18], 8);

-  output[14] = WRAPLOW(step1[14] + step1[17], 8);

-  output[15] = WRAPLOW(step1[15] + step1[16], 8);

-  output[16] = WRAPLOW(step1[15] - step1[16], 8);

-  output[17] = WRAPLOW(step1[14] - step1[17], 8);

-  output[18] = WRAPLOW(step1[13] - step1[18], 8);

-  output[19] = WRAPLOW(step1[12] - step1[19], 8);

-  output[20] = WRAPLOW(step1[11] - step1[20], 8);

-  output[21] = WRAPLOW(step1[10] - step1[21], 8);

-  output[22] = WRAPLOW(step1[9] - step1[22], 8);

-  output[23] = WRAPLOW(step1[8] - step1[23], 8);

-  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);

+  output[0] = WRAPLOW(step1[0] + step1[31]);

+  output[1] = WRAPLOW(step1[1] + step1[30]);

+  output[2] = WRAPLOW(step1[2] + step1[29]);

+  output[3] = WRAPLOW(step1[3] + step1[28]);

+  output[4] = WRAPLOW(step1[4] + step1[27]);

+  output[5] = WRAPLOW(step1[5] + step1[26]);

+  output[6] = WRAPLOW(step1[6] + step1[25]);

+  output[7] = WRAPLOW(step1[7] + step1[24]);

+  output[8] = WRAPLOW(step1[8] + step1[23]);

+  output[9] = WRAPLOW(step1[9] + step1[22]);

+  output[10] = WRAPLOW(step1[10] + step1[21]);

+  output[11] = WRAPLOW(step1[11] + step1[20]);

+  output[12] = WRAPLOW(step1[12] + step1[19]);

+  output[13] = WRAPLOW(step1[13] + step1[18]);

+  output[14] = WRAPLOW(step1[14] + step1[17]);

+  output[15] = WRAPLOW(step1[15] + step1[16]);

+  output[16] = WRAPLOW(step1[15] - step1[16]);

+  output[17] = WRAPLOW(step1[14] - step1[17]);

+  output[18] = WRAPLOW(step1[13] - step1[18]);

+  output[19] = WRAPLOW(step1[12] - step1[19]);

+  output[20] = WRAPLOW(step1[11] - step1[20]);

+  output[21] = WRAPLOW(step1[10] - step1[21]);

+  output[22] = WRAPLOW(step1[9] - step1[22]);

+  output[23] = WRAPLOW(step1[8] - step1[23]);

+  output[24] = WRAPLOW(step1[7] - step1[24]);

+  output[25] = WRAPLOW(step1[6] - step1[25]);

+  output[26] = WRAPLOW(step1[5] - step1[26]);

+  output[27] = WRAPLOW(step1[4] - step1[27]);

+  output[28] = WRAPLOW(step1[3] - step1[28]);

+  output[29] = WRAPLOW(step1[2] - step1[29]);

+  output[30] = WRAPLOW(step1[1] - step1[30]);

+  output[31] = WRAPLOW(step1[0] - step1[31]);

 }

 void vp10_idct32x32_1024_add_c(const tran_low_t *input, uint8_t *dest,

@@ -1234,8 +1234,8 @@

   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);

+  tran_low_t out = WRAPLOW(dct_const_round_shift(input[0] * cospi_16_64));

+  out = WRAPLOW(dct_const_round_shift(out * cospi_16_64));

   a1 = ROUND_POWER_OF_TWO(out, 6);

   for (j = 0; j < 32; ++j) {

@@ -1269,10 +1269,10 @@

     c1 = e1 - c1;

     a1 -= b1;

     d1 += c1;

-    op[0] = WRAPLOW(a1, bd);

-    op[1] = WRAPLOW(b1, bd);

-    op[2] = WRAPLOW(c1, bd);

-    op[3] = WRAPLOW(d1, bd);

+    op[0] = HIGHBD_WRAPLOW(a1, bd);

+    op[1] = HIGHBD_WRAPLOW(b1, bd);

+    op[2] = HIGHBD_WRAPLOW(c1, bd);

+    op[3] = HIGHBD_WRAPLOW(d1, bd);

     ip += 4;

     op += 4;

   }

@@ -1313,8 +1313,8 @@

   a1 = ip[0] >> UNIT_QUANT_SHIFT;

   e1 = a1 >> 1;

   a1 -= e1;

-  op[0] = WRAPLOW(a1, bd);

-  op[1] = op[2] = op[3] = WRAPLOW(e1, bd);

+  op[0] = HIGHBD_WRAPLOW(a1, bd);

+  op[1] = op[2] = op[3] = HIGHBD_WRAPLOW(e1, bd);

   ip = tmp;

   for (i = 0; i < 4; i++) {

@@ -1340,18 +1340,18 @@

   // 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);

+  step[0] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step[1] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), 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);

+  step[2] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step[3] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), 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);

+  output[0] = HIGHBD_WRAPLOW(step[0] + step[3], bd);

+  output[1] = HIGHBD_WRAPLOW(step[1] + step[2], bd);

+  output[2] = HIGHBD_WRAPLOW(step[1] - step[2], bd);

+  output[3] = HIGHBD_WRAPLOW(step[0] - step[3], bd);

 }

 void vp10_highbd_idct4x4_16_add_c(const tran_low_t *input, uint8_t *dest8,

@@ -1385,11 +1385,11 @@

                                 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);

+  tran_low_t out = HIGHBD_WRAPLOW(

+      highbd_dct_const_round_shift(input[0] * cospi_16_64), bd);

   uint16_t *dest = CONVERT_TO_SHORTPTR(dest8);

-  out = WRAPLOW(highbd_dct_const_round_shift(out * cospi_16_64, bd), bd);

+  out = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(out * cospi_16_64), bd);

   a1 = ROUND_POWER_OF_TWO(out, 4);

   for (i = 0; i < 4; i++) {

@@ -1411,39 +1411,39 @@

   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);

+  step1[4] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[7] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), 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);

+  step1[5] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[6] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   // stage 2 & stage 3 - even half

   vp10_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);

+  step2[4] = HIGHBD_WRAPLOW(step1[4] + step1[5], bd);

+  step2[5] = HIGHBD_WRAPLOW(step1[4] - step1[5], bd);

+  step2[6] = HIGHBD_WRAPLOW(-step1[6] + step1[7], bd);

+  step2[7] = HIGHBD_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[5] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[6] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), 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);

+  output[0] = HIGHBD_WRAPLOW(step1[0] + step1[7], bd);

+  output[1] = HIGHBD_WRAPLOW(step1[1] + step1[6], bd);

+  output[2] = HIGHBD_WRAPLOW(step1[2] + step1[5], bd);

+  output[3] = HIGHBD_WRAPLOW(step1[3] + step1[4], bd);

+  output[4] = HIGHBD_WRAPLOW(step1[3] - step1[4], bd);

+  output[5] = HIGHBD_WRAPLOW(step1[2] - step1[5], bd);

+  output[6] = HIGHBD_WRAPLOW(step1[1] - step1[6], bd);

+  output[7] = HIGHBD_WRAPLOW(step1[0] - step1[7], bd);

 }

 void vp10_highbd_idct8x8_64_add_c(const tran_low_t *input, uint8_t *dest8,

@@ -1477,10 +1477,10 @@

                                 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);

+  tran_low_t out = HIGHBD_WRAPLOW(

+      highbd_dct_const_round_shift(input[0] * cospi_16_64), bd);

   uint16_t *dest = CONVERT_TO_SHORTPTR(dest8);

-  out = WRAPLOW(highbd_dct_const_round_shift(out * cospi_16_64, bd), bd);

+  out = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(out * cospi_16_64), bd);

   a1 = ROUND_POWER_OF_TWO(out, 5);

   for (j = 0; j < 8; ++j) {

     for (i = 0; i < 8; ++i)

@@ -1521,10 +1521,10 @@

   // 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);

+  output[0] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s0 + s3), bd);

+  output[1] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s1 + s3), bd);

+  output[2] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s2), bd);

+  output[3] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s0 + s1 - s3), bd);

 }

 void vp10_highbd_iadst8_c(const tran_low_t *input, tran_low_t *output, int bd) {

@@ -1555,14 +1555,14 @@

   s6 = cospi_26_64 * x6 + cospi_6_64  * x7;

   s7 = cospi_6_64  * x6 - cospi_26_64 * x7;

-  x0 = WRAPLOW(highbd_dct_const_round_shift(s0 + s4, bd), bd);

-  x1 = WRAPLOW(highbd_dct_const_round_shift(s1 + s5, bd), bd);

-  x2 = WRAPLOW(highbd_dct_const_round_shift(s2 + s6, bd), bd);

-  x3 = WRAPLOW(highbd_dct_const_round_shift(s3 + s7, bd), bd);

-  x4 = WRAPLOW(highbd_dct_const_round_shift(s0 - s4, bd), bd);

-  x5 = WRAPLOW(highbd_dct_const_round_shift(s1 - s5, bd), bd);

-  x6 = WRAPLOW(highbd_dct_const_round_shift(s2 - s6, bd), bd);

-  x7 = WRAPLOW(highbd_dct_const_round_shift(s3 - s7, bd), bd);

+  x0 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s0 + s4), bd);

+  x1 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s1 + s5), bd);

+  x2 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s2 + s6), bd);

+  x3 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s3 + s7), bd);

+  x4 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s0 - s4), bd);

+  x5 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s1 - s5), bd);

+  x6 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s2 - s6), bd);

+  x7 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s3 - s7), bd);

   // stage 2

   s0 = x0;

@@ -1574,14 +1574,14 @@

   s6 = -cospi_24_64 * x6 + cospi_8_64  * x7;

   s7 =  cospi_8_64  * x6 + cospi_24_64 * x7;

-  x0 = WRAPLOW(s0 + s2, bd);

-  x1 = WRAPLOW(s1 + s3, bd);

-  x2 = WRAPLOW(s0 - s2, bd);

-  x3 = WRAPLOW(s1 - s3, bd);

-  x4 = WRAPLOW(highbd_dct_const_round_shift(s4 + s6, bd), bd);

-  x5 = WRAPLOW(highbd_dct_const_round_shift(s5 + s7, bd), bd);

-  x6 = WRAPLOW(highbd_dct_const_round_shift(s4 - s6, bd), bd);

-  x7 = WRAPLOW(highbd_dct_const_round_shift(s5 - s7, bd), bd);

+  x0 = HIGHBD_WRAPLOW(s0 + s2, bd);

+  x1 = HIGHBD_WRAPLOW(s1 + s3, bd);

+  x2 = HIGHBD_WRAPLOW(s0 - s2, bd);

+  x3 = HIGHBD_WRAPLOW(s1 - s3, bd);

+  x4 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s4 + s6), bd);

+  x5 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s5 + s7), bd);

+  x6 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s4 - s6), bd);

+  x7 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s5 - s7), bd);

   // stage 3

   s2 = cospi_16_64 * (x2 + x3);

@@ -1589,19 +1589,19 @@

   s6 = cospi_16_64 * (x6 + x7);

   s7 = cospi_16_64 * (x6 - x7);

-  x2 = WRAPLOW(highbd_dct_const_round_shift(s2, bd), bd);

-  x3 = WRAPLOW(highbd_dct_const_round_shift(s3, bd), bd);

-  x6 = WRAPLOW(highbd_dct_const_round_shift(s6, bd), bd);

-  x7 = WRAPLOW(highbd_dct_const_round_shift(s7, bd), bd);

+  x2 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s2), bd);

+  x3 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s3), bd);

+  x6 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s6), bd);

+  x7 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s7), bd);

-  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);

+  output[0] = HIGHBD_WRAPLOW(x0, bd);

+  output[1] = HIGHBD_WRAPLOW(-x4, bd);

+  output[2] = HIGHBD_WRAPLOW(x6, bd);

+  output[3] = HIGHBD_WRAPLOW(-x2, bd);

+  output[4] = HIGHBD_WRAPLOW(x3, bd);

+  output[5] = HIGHBD_WRAPLOW(-x7, bd);

+  output[6] = HIGHBD_WRAPLOW(x5, bd);

+  output[7] = HIGHBD_WRAPLOW(-x1, bd);

 }

 void vp10_highbd_idct8x8_10_add_c(const tran_low_t *input, uint8_t *dest8,

@@ -1666,23 +1666,23 @@

   temp1 = step1[8] * cospi_30_64 - step1[15] * cospi_2_64;

   temp2 = step1[8] * cospi_2_64 + step1[15] * cospi_30_64;

-  step2[8] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[15] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[8] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[15] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = step1[9] * cospi_14_64 - step1[14] * cospi_18_64;

   temp2 = step1[9] * cospi_18_64 + step1[14] * cospi_14_64;

-  step2[9] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[14] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[9] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[14] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = step1[10] * cospi_22_64 - step1[13] * cospi_10_64;

   temp2 = step1[10] * cospi_10_64 + step1[13] * cospi_22_64;

-  step2[10] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[13] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[10] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[13] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = step1[11] * cospi_6_64 - step1[12] * cospi_26_64;

   temp2 = step1[11] * cospi_26_64 + step1[12] * cospi_6_64;

-  step2[11] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[12] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[11] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[12] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   // stage 3

   step1[0] = step2[0];

@@ -1692,109 +1692,109 @@

   temp1 = step2[4] * cospi_28_64 - step2[7] * cospi_4_64;

   temp2 = step2[4] * cospi_4_64 + step2[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);

+  step1[4] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[7] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = step2[5] * cospi_12_64 - step2[6] * cospi_20_64;

   temp2 = step2[5] * cospi_20_64 + step2[6] * 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);

+  step1[5] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[6] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

-  step1[8] = WRAPLOW(step2[8] + step2[9], bd);

-  step1[9] = WRAPLOW(step2[8] - step2[9], bd);

-  step1[10] = WRAPLOW(-step2[10] + step2[11], bd);

-  step1[11] = WRAPLOW(step2[10] + step2[11], bd);

-  step1[12] = WRAPLOW(step2[12] + step2[13], bd);

-  step1[13] = WRAPLOW(step2[12] - step2[13], bd);

-  step1[14] = WRAPLOW(-step2[14] + step2[15], bd);

-  step1[15] = WRAPLOW(step2[14] + step2[15], bd);

+  step1[8] = HIGHBD_WRAPLOW(step2[8] + step2[9], bd);

+  step1[9] = HIGHBD_WRAPLOW(step2[8] - step2[9], bd);

+  step1[10] = HIGHBD_WRAPLOW(-step2[10] + step2[11], bd);

+  step1[11] = HIGHBD_WRAPLOW(step2[10] + step2[11], bd);

+  step1[12] = HIGHBD_WRAPLOW(step2[12] + step2[13], bd);

+  step1[13] = HIGHBD_WRAPLOW(step2[12] - step2[13], bd);

+  step1[14] = HIGHBD_WRAPLOW(-step2[14] + step2[15], bd);

+  step1[15] = HIGHBD_WRAPLOW(step2[14] + step2[15], bd);

   // stage 4

   temp1 = (step1[0] + step1[1]) * cospi_16_64;

   temp2 = (step1[0] - step1[1]) * cospi_16_64;

-  step2[0] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[1] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[0] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[1] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = step1[2] * cospi_24_64 - step1[3] * cospi_8_64;

   temp2 = step1[2] * cospi_8_64 + step1[3] * cospi_24_64;

-  step2[2] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[3] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

-  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);

+  step2[2] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[3] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

+  step2[4] = HIGHBD_WRAPLOW(step1[4] + step1[5], bd);

+  step2[5] = HIGHBD_WRAPLOW(step1[4] - step1[5], bd);

+  step2[6] = HIGHBD_WRAPLOW(-step1[6] + step1[7], bd);

+  step2[7] = HIGHBD_WRAPLOW(step1[6] + step1[7], bd);

   step2[8] = step1[8];

   step2[15] = step1[15];

   temp1 = -step1[9] * cospi_8_64 + step1[14] * cospi_24_64;

   temp2 = step1[9] * cospi_24_64 + step1[14] * cospi_8_64;

-  step2[9] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[14] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[9] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[14] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = -step1[10] * cospi_24_64 - step1[13] * cospi_8_64;

   temp2 = -step1[10] * cospi_8_64 + step1[13] * cospi_24_64;

-  step2[10] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[13] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[10] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[13] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   step2[11] = step1[11];

   step2[12] = step1[12];

   // stage 5

-  step1[0] = WRAPLOW(step2[0] + step2[3], bd);

-  step1[1] = WRAPLOW(step2[1] + step2[2], bd);

-  step1[2] = WRAPLOW(step2[1] - step2[2], bd);

-  step1[3] = WRAPLOW(step2[0] - step2[3], bd);

+  step1[0] = HIGHBD_WRAPLOW(step2[0] + step2[3], bd);

+  step1[1] = HIGHBD_WRAPLOW(step2[1] + step2[2], bd);

+  step1[2] = HIGHBD_WRAPLOW(step2[1] - step2[2], bd);

+  step1[3] = HIGHBD_WRAPLOW(step2[0] - step2[3], bd);

   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[5] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[6] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   step1[7] = step2[7];

-  step1[8] = WRAPLOW(step2[8] + step2[11], bd);

-  step1[9] = WRAPLOW(step2[9] + step2[10], bd);

-  step1[10] = WRAPLOW(step2[9] - step2[10], bd);

-  step1[11] = WRAPLOW(step2[8] - step2[11], bd);

-  step1[12] = WRAPLOW(-step2[12] + step2[15], bd);

-  step1[13] = WRAPLOW(-step2[13] + step2[14], bd);

-  step1[14] = WRAPLOW(step2[13] + step2[14], bd);

-  step1[15] = WRAPLOW(step2[12] + step2[15], bd);

+  step1[8] = HIGHBD_WRAPLOW(step2[8] + step2[11], bd);

+  step1[9] = HIGHBD_WRAPLOW(step2[9] + step2[10], bd);

+  step1[10] = HIGHBD_WRAPLOW(step2[9] - step2[10], bd);

+  step1[11] = HIGHBD_WRAPLOW(step2[8] - step2[11], bd);

+  step1[12] = HIGHBD_WRAPLOW(-step2[12] + step2[15], bd);

+  step1[13] = HIGHBD_WRAPLOW(-step2[13] + step2[14], bd);

+  step1[14] = HIGHBD_WRAPLOW(step2[13] + step2[14], bd);

+  step1[15] = HIGHBD_WRAPLOW(step2[12] + step2[15], bd);

   // stage 6

-  step2[0] = WRAPLOW(step1[0] + step1[7], bd);

-  step2[1] = WRAPLOW(step1[1] + step1[6], bd);

-  step2[2] = WRAPLOW(step1[2] + step1[5], bd);

-  step2[3] = WRAPLOW(step1[3] + step1[4], bd);

-  step2[4] = WRAPLOW(step1[3] - step1[4], bd);

-  step2[5] = WRAPLOW(step1[2] - step1[5], bd);

-  step2[6] = WRAPLOW(step1[1] - step1[6], bd);

-  step2[7] = WRAPLOW(step1[0] - step1[7], bd);

+  step2[0] = HIGHBD_WRAPLOW(step1[0] + step1[7], bd);

+  step2[1] = HIGHBD_WRAPLOW(step1[1] + step1[6], bd);

+  step2[2] = HIGHBD_WRAPLOW(step1[2] + step1[5], bd);

+  step2[3] = HIGHBD_WRAPLOW(step1[3] + step1[4], bd);

+  step2[4] = HIGHBD_WRAPLOW(step1[3] - step1[4], bd);

+  step2[5] = HIGHBD_WRAPLOW(step1[2] - step1[5], bd);

+  step2[6] = HIGHBD_WRAPLOW(step1[1] - step1[6], bd);

+  step2[7] = HIGHBD_WRAPLOW(step1[0] - step1[7], bd);

   step2[8] = step1[8];

   step2[9] = step1[9];

   temp1 = (-step1[10] + step1[13]) * cospi_16_64;

   temp2 = (step1[10] + step1[13]) * cospi_16_64;

-  step2[10] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[13] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[10] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[13] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = (-step1[11] + step1[12]) * cospi_16_64;

   temp2 = (step1[11] + step1[12]) * cospi_16_64;

-  step2[11] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[12] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[11] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[12] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   step2[14] = step1[14];

   step2[15] = step1[15];

   // stage 7

-  output[0] = WRAPLOW(step2[0] + step2[15], bd);

-  output[1] = WRAPLOW(step2[1] + step2[14], bd);

-  output[2] = WRAPLOW(step2[2] + step2[13], bd);

-  output[3] = WRAPLOW(step2[3] + step2[12], bd);

-  output[4] = WRAPLOW(step2[4] + step2[11], bd);

-  output[5] = WRAPLOW(step2[5] + step2[10], bd);

-  output[6] = WRAPLOW(step2[6] + step2[9], bd);

-  output[7] = WRAPLOW(step2[7] + step2[8], bd);

-  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);

+  output[0] = HIGHBD_WRAPLOW(step2[0] + step2[15], bd);

+  output[1] = HIGHBD_WRAPLOW(step2[1] + step2[14], bd);

+  output[2] = HIGHBD_WRAPLOW(step2[2] + step2[13], bd);

+  output[3] = HIGHBD_WRAPLOW(step2[3] + step2[12], bd);

+  output[4] = HIGHBD_WRAPLOW(step2[4] + step2[11], bd);

+  output[5] = HIGHBD_WRAPLOW(step2[5] + step2[10], bd);

+  output[6] = HIGHBD_WRAPLOW(step2[6] + step2[9], bd);

+  output[7] = HIGHBD_WRAPLOW(step2[7] + step2[8], bd);

+  output[8] = HIGHBD_WRAPLOW(step2[7] - step2[8], bd);

+  output[9] = HIGHBD_WRAPLOW(step2[6] - step2[9], bd);

+  output[10] = HIGHBD_WRAPLOW(step2[5] - step2[10], bd);

+  output[11] = HIGHBD_WRAPLOW(step2[4] - step2[11], bd);

+  output[12] = HIGHBD_WRAPLOW(step2[3] - step2[12], bd);

+  output[13] = HIGHBD_WRAPLOW(step2[2] - step2[13], bd);

+  output[14] = HIGHBD_WRAPLOW(step2[1] - step2[14], bd);

+  output[15] = HIGHBD_WRAPLOW(step2[0] - step2[15], bd);

 }

 void vp10_highbd_idct16x16_256_add_c(const tran_low_t *input, uint8_t *dest8,

@@ -1874,22 +1874,22 @@

   s14 = x14 * cospi_29_64 + x15 * cospi_3_64;

   s15 = x14 * cospi_3_64  - x15 * cospi_29_64;

-  x0 = WRAPLOW(highbd_dct_const_round_shift(s0 + s8, bd), bd);

-  x1 = WRAPLOW(highbd_dct_const_round_shift(s1 + s9, bd), bd);

-  x2 = WRAPLOW(highbd_dct_const_round_shift(s2 + s10, bd), bd);

-  x3 = WRAPLOW(highbd_dct_const_round_shift(s3 + s11, bd), bd);

-  x4 = WRAPLOW(highbd_dct_const_round_shift(s4 + s12, bd), bd);

-  x5 = WRAPLOW(highbd_dct_const_round_shift(s5 + s13, bd), bd);

-  x6 = WRAPLOW(highbd_dct_const_round_shift(s6 + s14, bd), bd);

-  x7 = WRAPLOW(highbd_dct_const_round_shift(s7 + s15, bd), bd);

-  x8  = WRAPLOW(highbd_dct_const_round_shift(s0 - s8, bd), bd);

-  x9  = WRAPLOW(highbd_dct_const_round_shift(s1 - s9, bd), bd);

-  x10 = WRAPLOW(highbd_dct_const_round_shift(s2 - s10, bd), bd);

-  x11 = WRAPLOW(highbd_dct_const_round_shift(s3 - s11, bd), bd);

-  x12 = WRAPLOW(highbd_dct_const_round_shift(s4 - s12, bd), bd);

-  x13 = WRAPLOW(highbd_dct_const_round_shift(s5 - s13, bd), bd);

-  x14 = WRAPLOW(highbd_dct_const_round_shift(s6 - s14, bd), bd);

-  x15 = WRAPLOW(highbd_dct_const_round_shift(s7 - s15, bd), bd);

+  x0 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s0 + s8), bd);

+  x1 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s1 + s9), bd);

+  x2 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s2 + s10), bd);

+  x3 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s3 + s11), bd);

+  x4 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s4 + s12), bd);

+  x5 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s5 + s13), bd);

+  x6 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s6 + s14), bd);

+  x7 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s7 + s15), bd);

+  x8  = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s0 - s8), bd);

+  x9  = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s1 - s9), bd);

+  x10 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s2 - s10), bd);

+  x11 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s3 - s11), bd);

+  x12 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s4 - s12), bd);

+  x13 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s5 - s13), bd);

+  x14 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s6 - s14), bd);

+  x15 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s7 - s15), bd);

   // stage 2

   s0 = x0;

@@ -1909,22 +1909,22 @@

   s14 = -x14 * cospi_12_64 + x15 * cospi_20_64;

   s15 = x14 * cospi_20_64 + x15 * cospi_12_64;

-  x0 = WRAPLOW(s0 + s4, bd);

-  x1 = WRAPLOW(s1 + s5, bd);

-  x2 = WRAPLOW(s2 + s6, bd);

-  x3 = WRAPLOW(s3 + s7, bd);

-  x4 = WRAPLOW(s0 - s4, bd);

-  x5 = WRAPLOW(s1 - s5, bd);

-  x6 = WRAPLOW(s2 - s6, bd);

-  x7 = WRAPLOW(s3 - s7, bd);

-  x8 = WRAPLOW(highbd_dct_const_round_shift(s8 + s12, bd), bd);

-  x9 = WRAPLOW(highbd_dct_const_round_shift(s9 + s13, bd), bd);

-  x10 = WRAPLOW(highbd_dct_const_round_shift(s10 + s14, bd), bd);

-  x11 = WRAPLOW(highbd_dct_const_round_shift(s11 + s15, bd), bd);

-  x12 = WRAPLOW(highbd_dct_const_round_shift(s8 - s12, bd), bd);

-  x13 = WRAPLOW(highbd_dct_const_round_shift(s9 - s13, bd), bd);

-  x14 = WRAPLOW(highbd_dct_const_round_shift(s10 - s14, bd), bd);

-  x15 = WRAPLOW(highbd_dct_const_round_shift(s11 - s15, bd), bd);

+  x0 = HIGHBD_WRAPLOW(s0 + s4, bd);

+  x1 = HIGHBD_WRAPLOW(s1 + s5, bd);

+  x2 = HIGHBD_WRAPLOW(s2 + s6, bd);

+  x3 = HIGHBD_WRAPLOW(s3 + s7, bd);

+  x4 = HIGHBD_WRAPLOW(s0 - s4, bd);

+  x5 = HIGHBD_WRAPLOW(s1 - s5, bd);

+  x6 = HIGHBD_WRAPLOW(s2 - s6, bd);

+  x7 = HIGHBD_WRAPLOW(s3 - s7, bd);

+  x8 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s8 + s12), bd);

+  x9 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s9 + s13), bd);

+  x10 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s10 + s14), bd);

+  x11 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s11 + s15), bd);

+  x12 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s8 - s12), bd);

+  x13 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s9 - s13), bd);

+  x14 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s10 - s14), bd);

+  x15 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s11 - s15), bd);

   // stage 3

   s0 = x0;

@@ -1944,22 +1944,22 @@

   s14 = -x14 * cospi_24_64 + x15 * cospi_8_64;

   s15 = x14 * cospi_8_64 + x15 * cospi_24_64;

-  x0 = WRAPLOW(s0 + s2, bd);

-  x1 = WRAPLOW(s1 + s3, bd);

-  x2 = WRAPLOW(s0 - s2, bd);

-  x3 = WRAPLOW(s1 - s3, bd);

-  x4 = WRAPLOW(highbd_dct_const_round_shift(s4 + s6, bd), bd);

-  x5 = WRAPLOW(highbd_dct_const_round_shift(s5 + s7, bd), bd);

-  x6 = WRAPLOW(highbd_dct_const_round_shift(s4 - s6, bd), bd);

-  x7 = WRAPLOW(highbd_dct_const_round_shift(s5 - s7, bd), bd);

-  x8 = WRAPLOW(s8 + s10, bd);

-  x9 = WRAPLOW(s9 + s11, bd);

-  x10 = WRAPLOW(s8 - s10, bd);

-  x11 = WRAPLOW(s9 - s11, bd);

-  x12 = WRAPLOW(highbd_dct_const_round_shift(s12 + s14, bd), bd);

-  x13 = WRAPLOW(highbd_dct_const_round_shift(s13 + s15, bd), bd);

-  x14 = WRAPLOW(highbd_dct_const_round_shift(s12 - s14, bd), bd);

-  x15 = WRAPLOW(highbd_dct_const_round_shift(s13 - s15, bd), bd);

+  x0 = HIGHBD_WRAPLOW(s0 + s2, bd);

+  x1 = HIGHBD_WRAPLOW(s1 + s3, bd);

+  x2 = HIGHBD_WRAPLOW(s0 - s2, bd);

+  x3 = HIGHBD_WRAPLOW(s1 - s3, bd);

+  x4 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s4 + s6), bd);

+  x5 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s5 + s7), bd);

+  x6 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s4 - s6), bd);

+  x7 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s5 - s7), bd);

+  x8 = HIGHBD_WRAPLOW(s8 + s10, bd);

+  x9 = HIGHBD_WRAPLOW(s9 + s11, bd);

+  x10 = HIGHBD_WRAPLOW(s8 - s10, bd);

+  x11 = HIGHBD_WRAPLOW(s9 - s11, bd);

+  x12 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s12 + s14), bd);

+  x13 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s13 + s15), bd);

+  x14 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s12 - s14), bd);

+  x15 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s13 - s15), bd);

   // stage 4

   s2 = (- cospi_16_64) * (x2 + x3);

@@ -1971,31 +1971,31 @@

   s14 = (- cospi_16_64) * (x14 + x15);

   s15 = cospi_16_64 * (x14 - x15);

-  x2 = WRAPLOW(highbd_dct_const_round_shift(s2, bd), bd);

-  x3 = WRAPLOW(highbd_dct_const_round_shift(s3, bd), bd);

-  x6 = WRAPLOW(highbd_dct_const_round_shift(s6, bd), bd);

-  x7 = WRAPLOW(highbd_dct_const_round_shift(s7, bd), bd);

-  x10 = WRAPLOW(highbd_dct_const_round_shift(s10, bd), bd);

-  x11 = WRAPLOW(highbd_dct_const_round_shift(s11, bd), bd);

-  x14 = WRAPLOW(highbd_dct_const_round_shift(s14, bd), bd);

-  x15 = WRAPLOW(highbd_dct_const_round_shift(s15, bd), bd);

+  x2 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s2), bd);

+  x3 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s3), bd);

+  x6 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s6), bd);

+  x7 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s7), bd);

+  x10 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s10), bd);

+  x11 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s11), bd);

+  x14 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s14), bd);

+  x15 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s15), bd);

-  output[0] = WRAPLOW(x0, bd);

-  output[1] = WRAPLOW(-x8, bd);

-  output[2] = WRAPLOW(x12, bd);

-  output[3] = WRAPLOW(-x4, bd);

-  output[4] = WRAPLOW(x6, bd);

-  output[5] = WRAPLOW(x14, bd);

-  output[6] = WRAPLOW(x10, bd);

-  output[7] = WRAPLOW(x2, bd);

-  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);

+  output[0] = HIGHBD_WRAPLOW(x0, bd);

+  output[1] = HIGHBD_WRAPLOW(-x8, bd);

+  output[2] = HIGHBD_WRAPLOW(x12, bd);

+  output[3] = HIGHBD_WRAPLOW(-x4, bd);

+  output[4] = HIGHBD_WRAPLOW(x6, bd);

+  output[5] = HIGHBD_WRAPLOW(x14, bd);

+  output[6] = HIGHBD_WRAPLOW(x10, bd);

+  output[7] = HIGHBD_WRAPLOW(x2, bd);

+  output[8] = HIGHBD_WRAPLOW(x3, bd);

+  output[9] = HIGHBD_WRAPLOW(x11, bd);

+  output[10] = HIGHBD_WRAPLOW(x15, bd);

+  output[11] = HIGHBD_WRAPLOW(x7, bd);

+  output[12] = HIGHBD_WRAPLOW(x5, bd);

+  output[13] = HIGHBD_WRAPLOW(-x13, bd);

+  output[14] = HIGHBD_WRAPLOW(x9, bd);

+  output[15] = HIGHBD_WRAPLOW(-x1, bd);

 }

 void vp10_highbd_idct16x16_10_add_c(const tran_low_t *input, uint8_t *dest8,

@@ -2030,11 +2030,11 @@

                                   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);

+  tran_low_t out = HIGHBD_WRAPLOW(

+      highbd_dct_const_round_shift(input[0] * cospi_16_64), bd);

   uint16_t *dest = CONVERT_TO_SHORTPTR(dest8);

-  out = WRAPLOW(highbd_dct_const_round_shift(out * cospi_16_64, bd), bd);

+  out = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(out * cospi_16_64), bd);

   a1 = ROUND_POWER_OF_TWO(out, 6);

   for (j = 0; j < 16; ++j) {

     for (i = 0; i < 16; ++i)

@@ -2069,43 +2069,43 @@

   temp1 = input[1] * cospi_31_64 - input[31] * cospi_1_64;

   temp2 = input[1] * cospi_1_64 + input[31] * cospi_31_64;

-  step1[16] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[31] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[16] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[31] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = input[17] * cospi_15_64 - input[15] * cospi_17_64;

   temp2 = input[17] * cospi_17_64 + input[15] * cospi_15_64;

-  step1[17] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[30] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[17] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[30] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = input[9] * cospi_23_64 - input[23] * cospi_9_64;

   temp2 = input[9] * cospi_9_64 + input[23] * cospi_23_64;

-  step1[18] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[29] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[18] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[29] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = input[25] * cospi_7_64 - input[7] * cospi_25_64;

   temp2 = input[25] * cospi_25_64 + input[7] * cospi_7_64;

-  step1[19] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[28] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[19] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[28] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = input[5] * cospi_27_64 - input[27] * cospi_5_64;

   temp2 = input[5] * cospi_5_64 + input[27] * cospi_27_64;

-  step1[20] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[27] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[20] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[27] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = input[21] * cospi_11_64 - input[11] * cospi_21_64;

   temp2 = input[21] * cospi_21_64 + input[11] * cospi_11_64;

-  step1[21] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[26] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[21] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[26] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = input[13] * cospi_19_64 - input[19] * cospi_13_64;

   temp2 = input[13] * cospi_13_64 + input[19] * cospi_19_64;

-  step1[22] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[25] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[22] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[25] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = input[29] * cospi_3_64 - input[3] * cospi_29_64;

   temp2 = input[29] * cospi_29_64 + input[3] * cospi_3_64;

-  step1[23] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[24] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[23] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[24] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   // stage 2

   step2[0] = step1[0];

@@ -2119,40 +2119,40 @@

   temp1 = step1[8] * cospi_30_64 - step1[15] * cospi_2_64;

   temp2 = step1[8] * cospi_2_64 + step1[15] * cospi_30_64;

-  step2[8] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[15] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[8] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[15] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = step1[9] * cospi_14_64 - step1[14] * cospi_18_64;

   temp2 = step1[9] * cospi_18_64 + step1[14] * cospi_14_64;

-  step2[9] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[14] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[9] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[14] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = step1[10] * cospi_22_64 - step1[13] * cospi_10_64;

   temp2 = step1[10] * cospi_10_64 + step1[13] * cospi_22_64;

-  step2[10] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[13] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[10] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[13] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = step1[11] * cospi_6_64 - step1[12] * cospi_26_64;

   temp2 = step1[11] * cospi_26_64 + step1[12] * cospi_6_64;

-  step2[11] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[12] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[11] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[12] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

-  step2[16] = WRAPLOW(step1[16] + step1[17], bd);

-  step2[17] = WRAPLOW(step1[16] - step1[17], bd);

-  step2[18] = WRAPLOW(-step1[18] + step1[19], bd);

-  step2[19] = WRAPLOW(step1[18] + step1[19], bd);

-  step2[20] = WRAPLOW(step1[20] + step1[21], bd);

-  step2[21] = WRAPLOW(step1[20] - step1[21], bd);

-  step2[22] = WRAPLOW(-step1[22] + step1[23], bd);

-  step2[23] = WRAPLOW(step1[22] + step1[23], bd);

-  step2[24] = WRAPLOW(step1[24] + step1[25], bd);

-  step2[25] = WRAPLOW(step1[24] - step1[25], bd);

-  step2[26] = WRAPLOW(-step1[26] + step1[27], bd);

-  step2[27] = WRAPLOW(step1[26] + step1[27], bd);

-  step2[28] = WRAPLOW(step1[28] + step1[29], bd);

-  step2[29] = WRAPLOW(step1[28] - step1[29], bd);

-  step2[30] = WRAPLOW(-step1[30] + step1[31], bd);

-  step2[31] = WRAPLOW(step1[30] + step1[31], bd);

+  step2[16] = HIGHBD_WRAPLOW(step1[16] + step1[17], bd);

+  step2[17] = HIGHBD_WRAPLOW(step1[16] - step1[17], bd);

+  step2[18] = HIGHBD_WRAPLOW(-step1[18] + step1[19], bd);

+  step2[19] = HIGHBD_WRAPLOW(step1[18] + step1[19], bd);

+  step2[20] = HIGHBD_WRAPLOW(step1[20] + step1[21], bd);

+  step2[21] = HIGHBD_WRAPLOW(step1[20] - step1[21], bd);

+  step2[22] = HIGHBD_WRAPLOW(-step1[22] + step1[23], bd);

+  step2[23] = HIGHBD_WRAPLOW(step1[22] + step1[23], bd);

+  step2[24] = HIGHBD_WRAPLOW(step1[24] + step1[25], bd);

+  step2[25] = HIGHBD_WRAPLOW(step1[24] - step1[25], bd);

+  step2[26] = HIGHBD_WRAPLOW(-step1[26] + step1[27], bd);

+  step2[27] = HIGHBD_WRAPLOW(step1[26] + step1[27], bd);

+  step2[28] = HIGHBD_WRAPLOW(step1[28] + step1[29], bd);

+  step2[29] = HIGHBD_WRAPLOW(step1[28] - step1[29], bd);

+  step2[30] = HIGHBD_WRAPLOW(-step1[30] + step1[31], bd);

+  step2[31] = HIGHBD_WRAPLOW(step1[30] + step1[31], bd);

   // stage 3

   step1[0] = step2[0];

@@ -2162,42 +2162,42 @@

   temp1 = step2[4] * cospi_28_64 - step2[7] * cospi_4_64;

   temp2 = step2[4] * cospi_4_64 + step2[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);

+  step1[4] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[7] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = step2[5] * cospi_12_64 - step2[6] * cospi_20_64;

   temp2 = step2[5] * cospi_20_64 + step2[6] * 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);

+  step1[5] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[6] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

-  step1[8] = WRAPLOW(step2[8] + step2[9], bd);

-  step1[9] = WRAPLOW(step2[8] - step2[9], bd);

-  step1[10] = WRAPLOW(-step2[10] + step2[11], bd);

-  step1[11] = WRAPLOW(step2[10] + step2[11], bd);

-  step1[12] = WRAPLOW(step2[12] + step2[13], bd);

-  step1[13] = WRAPLOW(step2[12] - step2[13], bd);

-  step1[14] = WRAPLOW(-step2[14] + step2[15], bd);

-  step1[15] = WRAPLOW(step2[14] + step2[15], bd);

+  step1[8] = HIGHBD_WRAPLOW(step2[8] + step2[9], bd);

+  step1[9] = HIGHBD_WRAPLOW(step2[8] - step2[9], bd);

+  step1[10] = HIGHBD_WRAPLOW(-step2[10] + step2[11], bd);

+  step1[11] = HIGHBD_WRAPLOW(step2[10] + step2[11], bd);

+  step1[12] = HIGHBD_WRAPLOW(step2[12] + step2[13], bd);

+  step1[13] = HIGHBD_WRAPLOW(step2[12] - step2[13], bd);

+  step1[14] = HIGHBD_WRAPLOW(-step2[14] + step2[15], bd);

+  step1[15] = HIGHBD_WRAPLOW(step2[14] + step2[15], bd);

   step1[16] = step2[16];

   step1[31] = step2[31];

   temp1 = -step2[17] * cospi_4_64 + step2[30] * cospi_28_64;

   temp2 = step2[17] * cospi_28_64 + step2[30] * cospi_4_64;

-  step1[17] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[30] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[17] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[30] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = -step2[18] * cospi_28_64 - step2[29] * cospi_4_64;

   temp2 = -step2[18] * cospi_4_64 + step2[29] * cospi_28_64;

-  step1[18] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[29] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[18] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[29] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   step1[19] = step2[19];

   step1[20] = step2[20];

   temp1 = -step2[21] * cospi_20_64 + step2[26] * cospi_12_64;

   temp2 = step2[21] * cospi_12_64 + step2[26] * cospi_20_64;

-  step1[21] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[26] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[21] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[26] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = -step2[22] * cospi_12_64 - step2[25] * cospi_20_64;

   temp2 = -step2[22] * cospi_20_64 + step2[25] * cospi_12_64;

-  step1[22] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[25] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[22] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[25] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   step1[23] = step2[23];

   step1[24] = step2[24];

   step1[27] = step2[27];

@@ -2206,87 +2206,87 @@

   // stage 4

   temp1 = (step1[0] + step1[1]) * cospi_16_64;

   temp2 = (step1[0] - step1[1]) * cospi_16_64;

-  step2[0] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[1] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[0] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[1] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = step1[2] * cospi_24_64 - step1[3] * cospi_8_64;

   temp2 = step1[2] * cospi_8_64 + step1[3] * cospi_24_64;

-  step2[2] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[3] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

-  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);

+  step2[2] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[3] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

+  step2[4] = HIGHBD_WRAPLOW(step1[4] + step1[5], bd);

+  step2[5] = HIGHBD_WRAPLOW(step1[4] - step1[5], bd);

+  step2[6] = HIGHBD_WRAPLOW(-step1[6] + step1[7], bd);

+  step2[7] = HIGHBD_WRAPLOW(step1[6] + step1[7], bd);

   step2[8] = step1[8];

   step2[15] = step1[15];

   temp1 = -step1[9] * cospi_8_64 + step1[14] * cospi_24_64;

   temp2 = step1[9] * cospi_24_64 + step1[14] * cospi_8_64;

-  step2[9] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[14] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[9] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[14] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = -step1[10] * cospi_24_64 - step1[13] * cospi_8_64;

   temp2 = -step1[10] * cospi_8_64 + step1[13] * cospi_24_64;

-  step2[10] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[13] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[10] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[13] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   step2[11] = step1[11];

   step2[12] = step1[12];

-  step2[16] = WRAPLOW(step1[16] + step1[19], bd);

-  step2[17] = WRAPLOW(step1[17] + step1[18], bd);

-  step2[18] = WRAPLOW(step1[17] - step1[18], bd);

-  step2[19] = WRAPLOW(step1[16] - step1[19], bd);

-  step2[20] = WRAPLOW(-step1[20] + step1[23], bd);

-  step2[21] = WRAPLOW(-step1[21] + step1[22], bd);

-  step2[22] = WRAPLOW(step1[21] + step1[22], bd);

-  step2[23] = WRAPLOW(step1[20] + step1[23], bd);

+  step2[16] = HIGHBD_WRAPLOW(step1[16] + step1[19], bd);

+  step2[17] = HIGHBD_WRAPLOW(step1[17] + step1[18], bd);

+  step2[18] = HIGHBD_WRAPLOW(step1[17] - step1[18], bd);

+  step2[19] = HIGHBD_WRAPLOW(step1[16] - step1[19], bd);

+  step2[20] = HIGHBD_WRAPLOW(-step1[20] + step1[23], bd);

+  step2[21] = HIGHBD_WRAPLOW(-step1[21] + step1[22], bd);

+  step2[22] = HIGHBD_WRAPLOW(step1[21] + step1[22], bd);

+  step2[23] = HIGHBD_WRAPLOW(step1[20] + step1[23], bd);

-  step2[24] = WRAPLOW(step1[24] + step1[27], bd);

-  step2[25] = WRAPLOW(step1[25] + step1[26], bd);

-  step2[26] = WRAPLOW(step1[25] - step1[26], bd);

-  step2[27] = WRAPLOW(step1[24] - step1[27], bd);

-  step2[28] = WRAPLOW(-step1[28] + step1[31], bd);

-  step2[29] = WRAPLOW(-step1[29] + step1[30], bd);

-  step2[30] = WRAPLOW(step1[29] + step1[30], bd);

-  step2[31] = WRAPLOW(step1[28] + step1[31], bd);

+  step2[24] = HIGHBD_WRAPLOW(step1[24] + step1[27], bd);

+  step2[25] = HIGHBD_WRAPLOW(step1[25] + step1[26], bd);

+  step2[26] = HIGHBD_WRAPLOW(step1[25] - step1[26], bd);

+  step2[27] = HIGHBD_WRAPLOW(step1[24] - step1[27], bd);

+  step2[28] = HIGHBD_WRAPLOW(-step1[28] + step1[31], bd);

+  step2[29] = HIGHBD_WRAPLOW(-step1[29] + step1[30], bd);

+  step2[30] = HIGHBD_WRAPLOW(step1[29] + step1[30], bd);

+  step2[31] = HIGHBD_WRAPLOW(step1[28] + step1[31], bd);

   // stage 5

-  step1[0] = WRAPLOW(step2[0] + step2[3], bd);

-  step1[1] = WRAPLOW(step2[1] + step2[2], bd);

-  step1[2] = WRAPLOW(step2[1] - step2[2], bd);

-  step1[3] = WRAPLOW(step2[0] - step2[3], bd);

+  step1[0] = HIGHBD_WRAPLOW(step2[0] + step2[3], bd);

+  step1[1] = HIGHBD_WRAPLOW(step2[1] + step2[2], bd);

+  step1[2] = HIGHBD_WRAPLOW(step2[1] - step2[2], bd);

+  step1[3] = HIGHBD_WRAPLOW(step2[0] - step2[3], bd);

   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[5] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[6] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   step1[7] = step2[7];

-  step1[8] = WRAPLOW(step2[8] + step2[11], bd);

-  step1[9] = WRAPLOW(step2[9] + step2[10], bd);

-  step1[10] = WRAPLOW(step2[9] - step2[10], bd);

-  step1[11] = WRAPLOW(step2[8] - step2[11], bd);

-  step1[12] = WRAPLOW(-step2[12] + step2[15], bd);

-  step1[13] = WRAPLOW(-step2[13] + step2[14], bd);

-  step1[14] = WRAPLOW(step2[13] + step2[14], bd);

-  step1[15] = WRAPLOW(step2[12] + step2[15], bd);

+  step1[8] = HIGHBD_WRAPLOW(step2[8] + step2[11], bd);

+  step1[9] = HIGHBD_WRAPLOW(step2[9] + step2[10], bd);

+  step1[10] = HIGHBD_WRAPLOW(step2[9] - step2[10], bd);

+  step1[11] = HIGHBD_WRAPLOW(step2[8] - step2[11], bd);

+  step1[12] = HIGHBD_WRAPLOW(-step2[12] + step2[15], bd);

+  step1[13] = HIGHBD_WRAPLOW(-step2[13] + step2[14], bd);

+  step1[14] = HIGHBD_WRAPLOW(step2[13] + step2[14], bd);

+  step1[15] = HIGHBD_WRAPLOW(step2[12] + step2[15], bd);

   step1[16] = step2[16];

   step1[17] = step2[17];

   temp1 = -step2[18] * cospi_8_64 + step2[29] * cospi_24_64;

   temp2 = step2[18] * cospi_24_64 + step2[29] * cospi_8_64;

-  step1[18] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[29] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[18] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[29] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = -step2[19] * cospi_8_64 + step2[28] * cospi_24_64;

   temp2 = step2[19] * cospi_24_64 + step2[28] * cospi_8_64;

-  step1[19] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[28] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[19] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[28] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = -step2[20] * cospi_24_64 - step2[27] * cospi_8_64;

   temp2 = -step2[20] * cospi_8_64 + step2[27] * cospi_24_64;

-  step1[20] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[27] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[20] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[27] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = -step2[21] * cospi_24_64 - step2[26] * cospi_8_64;

   temp2 = -step2[21] * cospi_8_64 + step2[26] * cospi_24_64;

-  step1[21] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[26] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[21] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[26] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   step1[22] = step2[22];

   step1[23] = step2[23];

   step1[24] = step2[24];

@@ -2295,62 +2295,62 @@

   step1[31] = step2[31];

   // stage 6

-  step2[0] = WRAPLOW(step1[0] + step1[7], bd);

-  step2[1] = WRAPLOW(step1[1] + step1[6], bd);

-  step2[2] = WRAPLOW(step1[2] + step1[5], bd);

-  step2[3] = WRAPLOW(step1[3] + step1[4], bd);

-  step2[4] = WRAPLOW(step1[3] - step1[4], bd);

-  step2[5] = WRAPLOW(step1[2] - step1[5], bd);

-  step2[6] = WRAPLOW(step1[1] - step1[6], bd);

-  step2[7] = WRAPLOW(step1[0] - step1[7], bd);

+  step2[0] = HIGHBD_WRAPLOW(step1[0] + step1[7], bd);

+  step2[1] = HIGHBD_WRAPLOW(step1[1] + step1[6], bd);

+  step2[2] = HIGHBD_WRAPLOW(step1[2] + step1[5], bd);

+  step2[3] = HIGHBD_WRAPLOW(step1[3] + step1[4], bd);

+  step2[4] = HIGHBD_WRAPLOW(step1[3] - step1[4], bd);

+  step2[5] = HIGHBD_WRAPLOW(step1[2] - step1[5], bd);

+  step2[6] = HIGHBD_WRAPLOW(step1[1] - step1[6], bd);

+  step2[7] = HIGHBD_WRAPLOW(step1[0] - step1[7], bd);

   step2[8] = step1[8];

   step2[9] = step1[9];

   temp1 = (-step1[10] + step1[13]) * cospi_16_64;

   temp2 = (step1[10] + step1[13]) * cospi_16_64;

-  step2[10] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[13] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[10] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[13] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = (-step1[11] + step1[12]) * cospi_16_64;

   temp2 = (step1[11] + step1[12]) * cospi_16_64;

-  step2[11] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[12] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[11] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[12] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   step2[14] = step1[14];

   step2[15] = step1[15];

-  step2[16] = WRAPLOW(step1[16] + step1[23], bd);

-  step2[17] = WRAPLOW(step1[17] + step1[22], bd);

-  step2[18] = WRAPLOW(step1[18] + step1[21], bd);

-  step2[19] = WRAPLOW(step1[19] + step1[20], bd);

-  step2[20] = WRAPLOW(step1[19] - step1[20], bd);

-  step2[21] = WRAPLOW(step1[18] - step1[21], bd);

-  step2[22] = WRAPLOW(step1[17] - step1[22], bd);

-  step2[23] = WRAPLOW(step1[16] - step1[23], bd);

+  step2[16] = HIGHBD_WRAPLOW(step1[16] + step1[23], bd);

+  step2[17] = HIGHBD_WRAPLOW(step1[17] + step1[22], bd);

+  step2[18] = HIGHBD_WRAPLOW(step1[18] + step1[21], bd);

+  step2[19] = HIGHBD_WRAPLOW(step1[19] + step1[20], bd);

+  step2[20] = HIGHBD_WRAPLOW(step1[19] - step1[20], bd);

+  step2[21] = HIGHBD_WRAPLOW(step1[18] - step1[21], bd);

+  step2[22] = HIGHBD_WRAPLOW(step1[17] - step1[22], bd);

+  step2[23] = HIGHBD_WRAPLOW(step1[16] - step1[23], bd);

-  step2[24] = WRAPLOW(-step1[24] + step1[31], bd);

-  step2[25] = WRAPLOW(-step1[25] + step1[30], bd);

-  step2[26] = WRAPLOW(-step1[26] + step1[29], bd);

-  step2[27] = WRAPLOW(-step1[27] + step1[28], bd);

-  step2[28] = WRAPLOW(step1[27] + step1[28], bd);

-  step2[29] = WRAPLOW(step1[26] + step1[29], bd);

-  step2[30] = WRAPLOW(step1[25] + step1[30], bd);

-  step2[31] = WRAPLOW(step1[24] + step1[31], bd);

+  step2[24] = HIGHBD_WRAPLOW(-step1[24] + step1[31], bd);

+  step2[25] = HIGHBD_WRAPLOW(-step1[25] + step1[30], bd);

+  step2[26] = HIGHBD_WRAPLOW(-step1[26] + step1[29], bd);

+  step2[27] = HIGHBD_WRAPLOW(-step1[27] + step1[28], bd);

+  step2[28] = HIGHBD_WRAPLOW(step1[27] + step1[28], bd);

+  step2[29] = HIGHBD_WRAPLOW(step1[26] + step1[29], bd);

+  step2[30] = HIGHBD_WRAPLOW(step1[25] + step1[30], bd);

+  step2[31] = HIGHBD_WRAPLOW(step1[24] + step1[31], bd);

   // stage 7

-  step1[0] = WRAPLOW(step2[0] + step2[15], bd);

-  step1[1] = WRAPLOW(step2[1] + step2[14], bd);

-  step1[2] = WRAPLOW(step2[2] + step2[13], bd);

-  step1[3] = WRAPLOW(step2[3] + step2[12], bd);

-  step1[4] = WRAPLOW(step2[4] + step2[11], bd);

-  step1[5] = WRAPLOW(step2[5] + step2[10], bd);

-  step1[6] = WRAPLOW(step2[6] + step2[9], bd);

-  step1[7] = WRAPLOW(step2[7] + step2[8], bd);

-  step1[8] = WRAPLOW(step2[7] - step2[8], bd);

-  step1[9] = WRAPLOW(step2[6] - step2[9], bd);

-  step1[10] = WRAPLOW(step2[5] - step2[10], bd);

-  step1[11] = WRAPLOW(step2[4] - step2[11], bd);

-  step1[12] = WRAPLOW(step2[3] - step2[12], bd);

-  step1[13] = WRAPLOW(step2[2] - step2[13], bd);

-  step1[14] = WRAPLOW(step2[1] - step2[14], bd);

-  step1[15] = WRAPLOW(step2[0] - step2[15], bd);

+  step1[0] = HIGHBD_WRAPLOW(step2[0] + step2[15], bd);

+  step1[1] = HIGHBD_WRAPLOW(step2[1] + step2[14], bd);

+  step1[2] = HIGHBD_WRAPLOW(step2[2] + step2[13], bd);

+  step1[3] = HIGHBD_WRAPLOW(step2[3] + step2[12], bd);

+  step1[4] = HIGHBD_WRAPLOW(step2[4] + step2[11], bd);

+  step1[5] = HIGHBD_WRAPLOW(step2[5] + step2[10], bd);

+  step1[6] = HIGHBD_WRAPLOW(step2[6] + step2[9], bd);

+  step1[7] = HIGHBD_WRAPLOW(step2[7] + step2[8], bd);

+  step1[8] = HIGHBD_WRAPLOW(step2[7] - step2[8], bd);

+  step1[9] = HIGHBD_WRAPLOW(step2[6] - step2[9], bd);

+  step1[10] = HIGHBD_WRAPLOW(step2[5] - step2[10], bd);

+  step1[11] = HIGHBD_WRAPLOW(step2[4] - step2[11], bd);

+  step1[12] = HIGHBD_WRAPLOW(step2[3] - step2[12], bd);

+  step1[13] = HIGHBD_WRAPLOW(step2[2] - step2[13], bd);

+  step1[14] = HIGHBD_WRAPLOW(step2[1] - step2[14], bd);

+  step1[15] = HIGHBD_WRAPLOW(step2[0] - step2[15], bd);

   step1[16] = step2[16];

   step1[17] = step2[17];

@@ -2358,20 +2358,20 @@

   step1[19] = step2[19];

   temp1 = (-step2[20] + step2[27]) * cospi_16_64;

   temp2 = (step2[20] + step2[27]) * cospi_16_64;

-  step1[20] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[27] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[20] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[27] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = (-step2[21] + step2[26]) * cospi_16_64;

   temp2 = (step2[21] + step2[26]) * cospi_16_64;

-  step1[21] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[26] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[21] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[26] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = (-step2[22] + step2[25]) * cospi_16_64;

   temp2 = (step2[22] + step2[25]) * cospi_16_64;

-  step1[22] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[25] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[22] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[25] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = (-step2[23] + step2[24]) * cospi_16_64;

   temp2 = (step2[23] + step2[24]) * cospi_16_64;

-  step1[23] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[24] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[23] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[24] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   step1[28] = step2[28];

   step1[29] = step2[29];

   step1[30] = step2[30];

@@ -2378,38 +2378,38 @@

   step1[31] = step2[31];

   // final stage

-  output[0] = WRAPLOW(step1[0] + step1[31], bd);

-  output[1] = WRAPLOW(step1[1] + step1[30], bd);

-  output[2] = WRAPLOW(step1[2] + step1[29], bd);

-  output[3] = WRAPLOW(step1[3] + step1[28], bd);

-  output[4] = WRAPLOW(step1[4] + step1[27], bd);

-  output[5] = WRAPLOW(step1[5] + step1[26], bd);

-  output[6] = WRAPLOW(step1[6] + step1[25], bd);

-  output[7] = WRAPLOW(step1[7] + step1[24], bd);

-  output[8] = WRAPLOW(step1[8] + step1[23], bd);

-  output[9] = WRAPLOW(step1[9] + step1[22], bd);

-  output[10] = WRAPLOW(step1[10] + step1[21], bd);

-  output[11] = WRAPLOW(step1[11] + step1[20], bd);

-  output[12] = WRAPLOW(step1[12] + step1[19], bd);

-  output[13] = WRAPLOW(step1[13] + step1[18], bd);

-  output[14] = WRAPLOW(step1[14] + step1[17], bd);

-  output[15] = WRAPLOW(step1[15] + step1[16], bd);

-  output[16] = WRAPLOW(step1[15] - step1[16], bd);

-  output[17] = WRAPLOW(step1[14] - step1[17], bd);

-  output[18] = WRAPLOW(step1[13] - step1[18], bd);

-  output[19] = WRAPLOW(step1[12] - step1[19], bd);

-  output[20] = WRAPLOW(step1[11] - step1[20], bd);

-  output[21] = WRAPLOW(step1[10] - step1[21], bd);

-  output[22] = WRAPLOW(step1[9] - step1[22], bd);

-  output[23] = WRAPLOW(step1[8] - step1[23], bd);

-  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);

+  output[0] = HIGHBD_WRAPLOW(step1[0] + step1[31], bd);

+  output[1] = HIGHBD_WRAPLOW(step1[1] + step1[30], bd);

+  output[2] = HIGHBD_WRAPLOW(step1[2] + step1[29], bd);

+  output[3] = HIGHBD_WRAPLOW(step1[3] + step1[28], bd);

+  output[4] = HIGHBD_WRAPLOW(step1[4] + step1[27], bd);

+  output[5] = HIGHBD_WRAPLOW(step1[5] + step1[26], bd);

+  output[6] = HIGHBD_WRAPLOW(step1[6] + step1[25], bd);

+  output[7] = HIGHBD_WRAPLOW(step1[7] + step1[24], bd);

+  output[8] = HIGHBD_WRAPLOW(step1[8] + step1[23], bd);

+  output[9] = HIGHBD_WRAPLOW(step1[9] + step1[22], bd);

+  output[10] = HIGHBD_WRAPLOW(step1[10] + step1[21], bd);

+  output[11] = HIGHBD_WRAPLOW(step1[11] + step1[20], bd);

+  output[12] = HIGHBD_WRAPLOW(step1[12] + step1[19], bd);

+  output[13] = HIGHBD_WRAPLOW(step1[13] + step1[18], bd);

+  output[14] = HIGHBD_WRAPLOW(step1[14] + step1[17], bd);

+  output[15] = HIGHBD_WRAPLOW(step1[15] + step1[16], bd);

+  output[16] = HIGHBD_WRAPLOW(step1[15] - step1[16], bd);

+  output[17] = HIGHBD_WRAPLOW(step1[14] - step1[17], bd);

+  output[18] = HIGHBD_WRAPLOW(step1[13] - step1[18], bd);

+  output[19] = HIGHBD_WRAPLOW(step1[12] - step1[19], bd);

+  output[20] = HIGHBD_WRAPLOW(step1[11] - step1[20], bd);

+  output[21] = HIGHBD_WRAPLOW(step1[10] - step1[21], bd);

+  output[22] = HIGHBD_WRAPLOW(step1[9] - step1[22], bd);

+  output[23] = HIGHBD_WRAPLOW(step1[8] - step1[23], bd);

+  output[24] = HIGHBD_WRAPLOW(step1[7] - step1[24], bd);

+  output[25] = HIGHBD_WRAPLOW(step1[6] - step1[25], bd);

+  output[26] = HIGHBD_WRAPLOW(step1[5] - step1[26], bd);

+  output[27] = HIGHBD_WRAPLOW(step1[4] - step1[27], bd);

+  output[28] = HIGHBD_WRAPLOW(step1[3] - step1[28], bd);

+  output[29] = HIGHBD_WRAPLOW(step1[2] - step1[29], bd);

+  output[30] = HIGHBD_WRAPLOW(step1[1] - step1[30], bd);

+  output[31] = HIGHBD_WRAPLOW(step1[0] - step1[31], bd);

 }

 void vp10_highbd_idct32x32_1024_add_c(const tran_low_t *input, uint8_t *dest8,

@@ -2485,9 +2485,9 @@

   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);

+  tran_low_t out = HIGHBD_WRAPLOW(

+      highbd_dct_const_round_shift(input[0] * cospi_16_64), bd);

+  out = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(out * cospi_16_64), bd);

   a1 = ROUND_POWER_OF_TWO(out, 6);

   for (j = 0; j < 32; ++j) {

--- a/vp10/common/vp10_inv_txfm.h

+++ b/vp10/common/vp10_inv_txfm.h

@@ -15,6 +15,7 @@

 #include "./vpx_config.h"

 #include "vpx_dsp/txfm_common.h"

+#include "vpx_dsp/inv_txfm.h"

 #include "vpx_ports/mem.h"

 #ifdef __cplusplus

@@ -21,7 +22,7 @@

 extern "C" {

 #endif

-static INLINE tran_low_t check_range(tran_high_t input) {

+static INLINE tran_high_t check_range(tran_high_t input) {

 #if CONFIG_COEFFICIENT_RANGE_CHECKING

   // For valid VP9 input streams, intermediate stage coefficients should always

   // stay within the range of a signed 16 bit integer. Coefficients can go out

@@ -32,17 +33,17 @@

   assert(INT16_MIN <= input);

   assert(input <= INT16_MAX);

 #endif  // CONFIG_COEFFICIENT_RANGE_CHECKING

-  return (tran_low_t)input;

+  return input;

 }

-static INLINE tran_low_t dct_const_round_shift(tran_high_t input) {

+static INLINE tran_high_t dct_const_round_shift(tran_high_t input) {

   tran_high_t rv = ROUND_POWER_OF_TWO(input, DCT_CONST_BITS);

-  return check_range(rv);

+  return rv;

 }

 #if CONFIG_VP9_HIGHBITDEPTH

-static INLINE tran_low_t highbd_check_range(tran_high_t input,

-                                            int bd) {

+static INLINE tran_high_t highbd_check_range(tran_high_t input,

+                                             int bd) {

 #if CONFIG_COEFFICIENT_RANGE_CHECKING

   // For valid highbitdepth VP9 streams, intermediate stage coefficients will

   // stay within the ranges:

@@ -56,13 +57,12 @@

   (void) int_min;

 #endif  // CONFIG_COEFFICIENT_RANGE_CHECKING

   (void) bd;

-  return (tran_low_t)input;

+  return input;

 }

-static INLINE tran_low_t highbd_dct_const_round_shift(tran_high_t input,

-                                                      int bd) {

+static INLINE tran_high_t highbd_dct_const_round_shift(tran_high_t input) {

   tran_high_t rv = ROUND_POWER_OF_TWO(input, DCT_CONST_BITS);

-  return highbd_check_range(rv, bd);

+  return rv;

 }

 #endif  // CONFIG_VP9_HIGHBITDEPTH

@@ -83,9 +83,21 @@

 // bd of 10 uses trans_low with 18bits, need to remove 14bits

 // bd of 12 uses trans_low with 20bits, need to remove 12bits

 // bd of x uses trans_low with 8+x bits, need to remove 24-x bits

-#define WRAPLOW(x, bd) ((((int32_t)(x)) << (24 - bd)) >> (24 - bd))

-#else

-#define WRAPLOW(x, bd) ((int32_t)(x))

+

+#define WRAPLOW(x) ((((int32_t)check_range(x)) << 16) >> 16)

+#if CONFIG_VP9_HIGHBITDEPTH

+#define HIGHBD_WRAPLOW(x, bd) \

+        ((((int32_t)highbd_check_range((x), bd)) << (24 - bd)) >> (24 - bd))

+#endif  // CONFIG_VP9_HIGHBITDEPTH

+

+#else   // CONFIG_EMULATE_HARDWARE

+

+#define WRAPLOW(x) ((int32_t)check_range(x))

+#if CONFIG_VP9_HIGHBITDEPTH

+#define HIGHBD_WRAPLOW(x, bd) \

+        ((int32_t)highbd_check_range((x), bd))

+#endif  // CONFIG_VP9_HIGHBITDEPTH

+

 #endif  // CONFIG_EMULATE_HARDWARE

 void vp10_idct4_c(const tran_low_t *input, tran_low_t *output);

@@ -107,14 +119,14 @@

 static INLINE uint16_t highbd_clip_pixel_add(uint16_t dest, tran_high_t trans,

                                              int bd) {

-  trans = WRAPLOW(trans, bd);

-  return clip_pixel_highbd(WRAPLOW(dest + trans, bd), bd);

+  trans = HIGHBD_WRAPLOW(trans, bd);

+  return clip_pixel_highbd(dest + trans, bd);

 }

 #endif

 static INLINE uint8_t clip_pixel_add(uint8_t dest, tran_high_t trans) {

-  trans = WRAPLOW(trans, 8);

-  return clip_pixel(WRAPLOW(dest + trans, 8));

+  trans = WRAPLOW(trans);

+  return clip_pixel(dest + trans);

 }

 #ifdef __cplusplus

 }  // extern "C"

--- a/vpx_dsp/inv_txfm.c

+++ b/vpx_dsp/inv_txfm.c

@@ -35,10 +35,10 @@

     c1 = e1 - c1;

     a1 -= b1;

     d1 += c1;

-    op[0] = WRAPLOW(a1, 8);

-    op[1] = WRAPLOW(b1, 8);

-    op[2] = WRAPLOW(c1, 8);

-    op[3] = WRAPLOW(d1, 8);

+    op[0] = WRAPLOW(a1);

+    op[1] = WRAPLOW(b1);

+    op[2] = WRAPLOW(c1);

+    op[3] = WRAPLOW(d1);

     ip += 4;

     op += 4;

   }

@@ -76,8 +76,8 @@

   a1 = ip[0] >> UNIT_QUANT_SHIFT;

   e1 = a1 >> 1;

   a1 -= e1;

-  op[0] = WRAPLOW(a1, 8);

-  op[1] = op[2] = op[3] = WRAPLOW(e1, 8);

+  op[0] = WRAPLOW(a1);

+  op[1] = op[2] = op[3] = WRAPLOW(e1);

   ip = tmp;

   for (i = 0; i < 4; i++) {

@@ -98,18 +98,18 @@

   // stage 1

   temp1 = (input[0] + input[2]) * cospi_16_64;

   temp2 = (input[0] - input[2]) * cospi_16_64;

-  step[0] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step[1] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step[0] = WRAPLOW(dct_const_round_shift(temp1));

+  step[1] = WRAPLOW(dct_const_round_shift(temp2));

   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(dct_const_round_shift(temp1), 8);

-  step[3] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step[2] = WRAPLOW(dct_const_round_shift(temp1));

+  step[3] = WRAPLOW(dct_const_round_shift(temp2));

   // 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);

+  output[0] = WRAPLOW(step[0] + step[3]);

+  output[1] = WRAPLOW(step[1] + step[2]);

+  output[2] = WRAPLOW(step[1] - step[2]);

+  output[3] = WRAPLOW(step[0] - step[3]);

 }

 void vpx_idct4x4_16_add_c(const tran_low_t *input, uint8_t *dest, int stride) {

@@ -141,8 +141,8 @@

                          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);

+  tran_low_t out = WRAPLOW(dct_const_round_shift(input[0] * cospi_16_64));

+  out = WRAPLOW(dct_const_round_shift(out * cospi_16_64));

   a1 = ROUND_POWER_OF_TWO(out, 4);

   for (i = 0; i < 4; i++) {

@@ -164,48 +164,48 @@

   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(dct_const_round_shift(temp1), 8);

-  step1[7] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[4] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[7] = WRAPLOW(dct_const_round_shift(temp2));

   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(dct_const_round_shift(temp1), 8);

-  step1[6] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[5] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[6] = WRAPLOW(dct_const_round_shift(temp2));

   // stage 2

   temp1 = (step1[0] + step1[2]) * cospi_16_64;

   temp2 = (step1[0] - step1[2]) * cospi_16_64;

-  step2[0] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[1] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[0] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[1] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step1[1] * cospi_24_64 - step1[3] * cospi_8_64;

   temp2 = step1[1] * cospi_8_64 + step1[3] * cospi_24_64;

-  step2[2] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[3] = WRAPLOW(dct_const_round_shift(temp2), 8);

-  step2[4] = WRAPLOW(step1[4] + step1[5], 8);

-  step2[5] = WRAPLOW(step1[4] - step1[5], 8);

-  step2[6] = WRAPLOW(-step1[6] + step1[7], 8);

-  step2[7] = WRAPLOW(step1[6] + step1[7], 8);

+  step2[2] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[3] = WRAPLOW(dct_const_round_shift(temp2));

+  step2[4] = WRAPLOW(step1[4] + step1[5]);

+  step2[5] = WRAPLOW(step1[4] - step1[5]);

+  step2[6] = WRAPLOW(-step1[6] + step1[7]);

+  step2[7] = WRAPLOW(step1[6] + step1[7]);

   // stage 3

-  step1[0] = WRAPLOW(step2[0] + step2[3], 8);

-  step1[1] = WRAPLOW(step2[1] + step2[2], 8);

-  step1[2] = WRAPLOW(step2[1] - step2[2], 8);

-  step1[3] = WRAPLOW(step2[0] - step2[3], 8);

+  step1[0] = WRAPLOW(step2[0] + step2[3]);

+  step1[1] = WRAPLOW(step2[1] + step2[2]);

+  step1[2] = WRAPLOW(step2[1] - step2[2]);

+  step1[3] = WRAPLOW(step2[0] - step2[3]);

   step1[4] = step2[4];

   temp1 = (step2[6] - step2[5]) * cospi_16_64;

   temp2 = (step2[5] + step2[6]) * cospi_16_64;

-  step1[5] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[6] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[5] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[6] = WRAPLOW(dct_const_round_shift(temp2));

   step1[7] = step2[7];

   // stage 4

-  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);

+  output[0] = WRAPLOW(step1[0] + step1[7]);

+  output[1] = WRAPLOW(step1[1] + step1[6]);

+  output[2] = WRAPLOW(step1[2] + step1[5]);

+  output[3] = WRAPLOW(step1[3] + step1[4]);

+  output[4] = WRAPLOW(step1[3] - step1[4]);

+  output[5] = WRAPLOW(step1[2] - step1[5]);

+  output[6] = WRAPLOW(step1[1] - step1[6]);

+  output[7] = WRAPLOW(step1[0] - step1[7]);

 }

 void vpx_idct8x8_64_add_c(const tran_low_t *input, uint8_t *dest, int stride) {

@@ -236,8 +236,8 @@

 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);

+  tran_low_t out = WRAPLOW(dct_const_round_shift(input[0] * cospi_16_64));

+  out = WRAPLOW(dct_const_round_shift(out * cospi_16_64));

   a1 = ROUND_POWER_OF_TWO(out, 5);

   for (j = 0; j < 8; ++j) {

     for (i = 0; i < 8; ++i)

@@ -277,10 +277,10 @@

   // The overall dynamic range is 14b (input) + 14b (multiplication scaling)

   // + 1b (addition) = 29b.

   // Hence the output bit depth is 15b.

-  output[0] = WRAPLOW(dct_const_round_shift(s0 + s3), 8);

-  output[1] = WRAPLOW(dct_const_round_shift(s1 + s3), 8);

-  output[2] = WRAPLOW(dct_const_round_shift(s2), 8);

-  output[3] = WRAPLOW(dct_const_round_shift(s0 + s1 - s3), 8);

+  output[0] = WRAPLOW(dct_const_round_shift(s0 + s3));

+  output[1] = WRAPLOW(dct_const_round_shift(s1 + s3));

+  output[2] = WRAPLOW(dct_const_round_shift(s2));

+  output[3] = WRAPLOW(dct_const_round_shift(s0 + s1 - s3));

 }

 void iadst8_c(const tran_low_t *input, tran_low_t *output) {

@@ -311,14 +311,14 @@

   s6 = (int)(cospi_26_64 * x6 + cospi_6_64  * x7);

   s7 = (int)(cospi_6_64  * x6 - cospi_26_64 * x7);

-  x0 = WRAPLOW(dct_const_round_shift(s0 + s4), 8);

-  x1 = WRAPLOW(dct_const_round_shift(s1 + s5), 8);

-  x2 = WRAPLOW(dct_const_round_shift(s2 + s6), 8);

-  x3 = WRAPLOW(dct_const_round_shift(s3 + s7), 8);

-  x4 = WRAPLOW(dct_const_round_shift(s0 - s4), 8);

-  x5 = WRAPLOW(dct_const_round_shift(s1 - s5), 8);

-  x6 = WRAPLOW(dct_const_round_shift(s2 - s6), 8);

-  x7 = WRAPLOW(dct_const_round_shift(s3 - s7), 8);

+  x0 = WRAPLOW(dct_const_round_shift(s0 + s4));

+  x1 = WRAPLOW(dct_const_round_shift(s1 + s5));

+  x2 = WRAPLOW(dct_const_round_shift(s2 + s6));

+  x3 = WRAPLOW(dct_const_round_shift(s3 + s7));

+  x4 = WRAPLOW(dct_const_round_shift(s0 - s4));

+  x5 = WRAPLOW(dct_const_round_shift(s1 - s5));

+  x6 = WRAPLOW(dct_const_round_shift(s2 - s6));

+  x7 = WRAPLOW(dct_const_round_shift(s3 - s7));

   // stage 2

   s0 = (int)x0;

@@ -330,14 +330,14 @@

   s6 = (int)(-cospi_24_64 * x6 + cospi_8_64 * x7);

   s7 = (int)(cospi_8_64 * x6 + cospi_24_64 * x7);

-  x0 = WRAPLOW(s0 + s2, 8);

-  x1 = WRAPLOW(s1 + s3, 8);

-  x2 = WRAPLOW(s0 - s2, 8);

-  x3 = WRAPLOW(s1 - s3, 8);

-  x4 = WRAPLOW(dct_const_round_shift(s4 + s6), 8);

-  x5 = WRAPLOW(dct_const_round_shift(s5 + s7), 8);

-  x6 = WRAPLOW(dct_const_round_shift(s4 - s6), 8);

-  x7 = WRAPLOW(dct_const_round_shift(s5 - s7), 8);

+  x0 = WRAPLOW(s0 + s2);

+  x1 = WRAPLOW(s1 + s3);

+  x2 = WRAPLOW(s0 - s2);

+  x3 = WRAPLOW(s1 - s3);

+  x4 = WRAPLOW(dct_const_round_shift(s4 + s6));

+  x5 = WRAPLOW(dct_const_round_shift(s5 + s7));

+  x6 = WRAPLOW(dct_const_round_shift(s4 - s6));

+  x7 = WRAPLOW(dct_const_round_shift(s5 - s7));

   // stage 3

   s2 = (int)(cospi_16_64 * (x2 + x3));

@@ -345,19 +345,19 @@

   s6 = (int)(cospi_16_64 * (x6 + x7));

   s7 = (int)(cospi_16_64 * (x6 - x7));

-  x2 = WRAPLOW(dct_const_round_shift(s2), 8);

-  x3 = WRAPLOW(dct_const_round_shift(s3), 8);

-  x6 = WRAPLOW(dct_const_round_shift(s6), 8);

-  x7 = WRAPLOW(dct_const_round_shift(s7), 8);

+  x2 = WRAPLOW(dct_const_round_shift(s2));

+  x3 = WRAPLOW(dct_const_round_shift(s3));

+  x6 = WRAPLOW(dct_const_round_shift(s6));

+  x7 = WRAPLOW(dct_const_round_shift(s7));

-  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);

+  output[0] = WRAPLOW(x0);

+  output[1] = WRAPLOW(-x4);

+  output[2] = WRAPLOW(x6);

+  output[3] = WRAPLOW(-x2);

+  output[4] = WRAPLOW(x3);

+  output[5] = WRAPLOW(-x7);

+  output[6] = WRAPLOW(x5);

+  output[7] = WRAPLOW(-x1);

 }

 void vpx_idct8x8_12_add_c(const tran_low_t *input, uint8_t *dest, int stride) {

@@ -420,23 +420,23 @@

   temp1 = step1[8] * cospi_30_64 - step1[15] * cospi_2_64;

   temp2 = step1[8] * cospi_2_64 + step1[15] * cospi_30_64;

-  step2[8] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[15] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[8] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[15] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step1[9] * cospi_14_64 - step1[14] * cospi_18_64;

   temp2 = step1[9] * cospi_18_64 + step1[14] * cospi_14_64;

-  step2[9] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[14] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[9] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[14] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step1[10] * cospi_22_64 - step1[13] * cospi_10_64;

   temp2 = step1[10] * cospi_10_64 + step1[13] * cospi_22_64;

-  step2[10] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[13] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[10] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[13] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step1[11] * cospi_6_64 - step1[12] * cospi_26_64;

   temp2 = step1[11] * cospi_26_64 + step1[12] * cospi_6_64;

-  step2[11] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[12] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[11] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[12] = WRAPLOW(dct_const_round_shift(temp2));

   // stage 3

   step1[0] = step2[0];

@@ -446,109 +446,109 @@

   temp1 = step2[4] * cospi_28_64 - step2[7] * cospi_4_64;

   temp2 = step2[4] * cospi_4_64 + step2[7] * cospi_28_64;

-  step1[4] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[7] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[4] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[7] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step2[5] * cospi_12_64 - step2[6] * cospi_20_64;

   temp2 = step2[5] * cospi_20_64 + step2[6] * cospi_12_64;

-  step1[5] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[6] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[5] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[6] = WRAPLOW(dct_const_round_shift(temp2));

-  step1[8] = WRAPLOW(step2[8] + step2[9], 8);

-  step1[9] = WRAPLOW(step2[8] - step2[9], 8);

-  step1[10] = WRAPLOW(-step2[10] + step2[11], 8);

-  step1[11] = WRAPLOW(step2[10] + step2[11], 8);

-  step1[12] = WRAPLOW(step2[12] + step2[13], 8);

-  step1[13] = WRAPLOW(step2[12] - step2[13], 8);

-  step1[14] = WRAPLOW(-step2[14] + step2[15], 8);

-  step1[15] = WRAPLOW(step2[14] + step2[15], 8);

+  step1[8] = WRAPLOW(step2[8] + step2[9]);

+  step1[9] = WRAPLOW(step2[8] - step2[9]);

+  step1[10] = WRAPLOW(-step2[10] + step2[11]);

+  step1[11] = WRAPLOW(step2[10] + step2[11]);

+  step1[12] = WRAPLOW(step2[12] + step2[13]);

+  step1[13] = WRAPLOW(step2[12] - step2[13]);

+  step1[14] = WRAPLOW(-step2[14] + step2[15]);

+  step1[15] = WRAPLOW(step2[14] + step2[15]);

   // stage 4

   temp1 = (step1[0] + step1[1]) * cospi_16_64;

   temp2 = (step1[0] - step1[1]) * cospi_16_64;

-  step2[0] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[1] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[0] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[1] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step1[2] * cospi_24_64 - step1[3] * cospi_8_64;

   temp2 = step1[2] * cospi_8_64 + step1[3] * cospi_24_64;

-  step2[2] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[3] = WRAPLOW(dct_const_round_shift(temp2), 8);

-  step2[4] = WRAPLOW(step1[4] + step1[5], 8);

-  step2[5] = WRAPLOW(step1[4] - step1[5], 8);

-  step2[6] = WRAPLOW(-step1[6] + step1[7], 8);

-  step2[7] = WRAPLOW(step1[6] + step1[7], 8);

+  step2[2] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[3] = WRAPLOW(dct_const_round_shift(temp2));

+  step2[4] = WRAPLOW(step1[4] + step1[5]);

+  step2[5] = WRAPLOW(step1[4] - step1[5]);

+  step2[6] = WRAPLOW(-step1[6] + step1[7]);

+  step2[7] = WRAPLOW(step1[6] + step1[7]);

   step2[8] = step1[8];

   step2[15] = step1[15];

   temp1 = -step1[9] * cospi_8_64 + step1[14] * cospi_24_64;

   temp2 = step1[9] * cospi_24_64 + step1[14] * cospi_8_64;

-  step2[9] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[14] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[9] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[14] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = -step1[10] * cospi_24_64 - step1[13] * cospi_8_64;

   temp2 = -step1[10] * cospi_8_64 + step1[13] * cospi_24_64;

-  step2[10] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[13] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[10] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[13] = WRAPLOW(dct_const_round_shift(temp2));

   step2[11] = step1[11];

   step2[12] = step1[12];

   // stage 5

-  step1[0] = WRAPLOW(step2[0] + step2[3], 8);

-  step1[1] = WRAPLOW(step2[1] + step2[2], 8);

-  step1[2] = WRAPLOW(step2[1] - step2[2], 8);

-  step1[3] = WRAPLOW(step2[0] - step2[3], 8);

+  step1[0] = WRAPLOW(step2[0] + step2[3]);

+  step1[1] = WRAPLOW(step2[1] + step2[2]);

+  step1[2] = WRAPLOW(step2[1] - step2[2]);

+  step1[3] = WRAPLOW(step2[0] - step2[3]);

   step1[4] = step2[4];

   temp1 = (step2[6] - step2[5]) * cospi_16_64;

   temp2 = (step2[5] + step2[6]) * cospi_16_64;

-  step1[5] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[6] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[5] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[6] = WRAPLOW(dct_const_round_shift(temp2));

   step1[7] = step2[7];

-  step1[8] = WRAPLOW(step2[8] + step2[11], 8);

-  step1[9] = WRAPLOW(step2[9] + step2[10], 8);

-  step1[10] = WRAPLOW(step2[9] - step2[10], 8);

-  step1[11] = WRAPLOW(step2[8] - step2[11], 8);

-  step1[12] = WRAPLOW(-step2[12] + step2[15], 8);

-  step1[13] = WRAPLOW(-step2[13] + step2[14], 8);

-  step1[14] = WRAPLOW(step2[13] + step2[14], 8);

-  step1[15] = WRAPLOW(step2[12] + step2[15], 8);

+  step1[8] = WRAPLOW(step2[8] + step2[11]);

+  step1[9] = WRAPLOW(step2[9] + step2[10]);

+  step1[10] = WRAPLOW(step2[9] - step2[10]);

+  step1[11] = WRAPLOW(step2[8] - step2[11]);

+  step1[12] = WRAPLOW(-step2[12] + step2[15]);

+  step1[13] = WRAPLOW(-step2[13] + step2[14]);

+  step1[14] = WRAPLOW(step2[13] + step2[14]);

+  step1[15] = WRAPLOW(step2[12] + step2[15]);

   // stage 6

-  step2[0] = WRAPLOW(step1[0] + step1[7], 8);

-  step2[1] = WRAPLOW(step1[1] + step1[6], 8);

-  step2[2] = WRAPLOW(step1[2] + step1[5], 8);

-  step2[3] = WRAPLOW(step1[3] + step1[4], 8);

-  step2[4] = WRAPLOW(step1[3] - step1[4], 8);

-  step2[5] = WRAPLOW(step1[2] - step1[5], 8);

-  step2[6] = WRAPLOW(step1[1] - step1[6], 8);

-  step2[7] = WRAPLOW(step1[0] - step1[7], 8);

+  step2[0] = WRAPLOW(step1[0] + step1[7]);

+  step2[1] = WRAPLOW(step1[1] + step1[6]);

+  step2[2] = WRAPLOW(step1[2] + step1[5]);

+  step2[3] = WRAPLOW(step1[3] + step1[4]);

+  step2[4] = WRAPLOW(step1[3] - step1[4]);

+  step2[5] = WRAPLOW(step1[2] - step1[5]);

+  step2[6] = WRAPLOW(step1[1] - step1[6]);

+  step2[7] = WRAPLOW(step1[0] - step1[7]);

   step2[8] = step1[8];

   step2[9] = step1[9];

   temp1 = (-step1[10] + step1[13]) * cospi_16_64;

   temp2 = (step1[10] + step1[13]) * cospi_16_64;

-  step2[10] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[13] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[10] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[13] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = (-step1[11] + step1[12]) * cospi_16_64;

   temp2 = (step1[11] + step1[12]) * cospi_16_64;

-  step2[11] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[12] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[11] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[12] = WRAPLOW(dct_const_round_shift(temp2));

   step2[14] = step1[14];

   step2[15] = step1[15];

   // stage 7

-  output[0] = WRAPLOW(step2[0] + step2[15], 8);

-  output[1] = WRAPLOW(step2[1] + step2[14], 8);

-  output[2] = WRAPLOW(step2[2] + step2[13], 8);

-  output[3] = WRAPLOW(step2[3] + step2[12], 8);

-  output[4] = WRAPLOW(step2[4] + step2[11], 8);

-  output[5] = WRAPLOW(step2[5] + step2[10], 8);

-  output[6] = WRAPLOW(step2[6] + step2[9], 8);

-  output[7] = WRAPLOW(step2[7] + step2[8], 8);

-  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);

+  output[0] = WRAPLOW(step2[0] + step2[15]);

+  output[1] = WRAPLOW(step2[1] + step2[14]);

+  output[2] = WRAPLOW(step2[2] + step2[13]);

+  output[3] = WRAPLOW(step2[3] + step2[12]);

+  output[4] = WRAPLOW(step2[4] + step2[11]);

+  output[5] = WRAPLOW(step2[5] + step2[10]);

+  output[6] = WRAPLOW(step2[6] + step2[9]);

+  output[7] = WRAPLOW(step2[7] + step2[8]);

+  output[8] = WRAPLOW(step2[7] - step2[8]);

+  output[9] = WRAPLOW(step2[6] - step2[9]);

+  output[10] = WRAPLOW(step2[5] - step2[10]);

+  output[11] = WRAPLOW(step2[4] - step2[11]);

+  output[12] = WRAPLOW(step2[3] - step2[12]);

+  output[13] = WRAPLOW(step2[2] - step2[13]);

+  output[14] = WRAPLOW(step2[1] - step2[14]);

+  output[15] = WRAPLOW(step2[0] - step2[15]);

 }

 void vpx_idct16x16_256_add_c(const tran_low_t *input, uint8_t *dest,

@@ -625,22 +625,22 @@

   s14 = x14 * cospi_29_64 + x15 * cospi_3_64;

   s15 = x14 * cospi_3_64  - x15 * cospi_29_64;

-  x0 = WRAPLOW(dct_const_round_shift(s0 + s8), 8);

-  x1 = WRAPLOW(dct_const_round_shift(s1 + s9), 8);

-  x2 = WRAPLOW(dct_const_round_shift(s2 + s10), 8);

-  x3 = WRAPLOW(dct_const_round_shift(s3 + s11), 8);

-  x4 = WRAPLOW(dct_const_round_shift(s4 + s12), 8);

-  x5 = WRAPLOW(dct_const_round_shift(s5 + s13), 8);

-  x6 = WRAPLOW(dct_const_round_shift(s6 + s14), 8);

-  x7 = WRAPLOW(dct_const_round_shift(s7 + s15), 8);

-  x8 = WRAPLOW(dct_const_round_shift(s0 - s8), 8);

-  x9 = WRAPLOW(dct_const_round_shift(s1 - s9), 8);

-  x10 = WRAPLOW(dct_const_round_shift(s2 - s10), 8);

-  x11 = WRAPLOW(dct_const_round_shift(s3 - s11), 8);

-  x12 = WRAPLOW(dct_const_round_shift(s4 - s12), 8);

-  x13 = WRAPLOW(dct_const_round_shift(s5 - s13), 8);

-  x14 = WRAPLOW(dct_const_round_shift(s6 - s14), 8);

-  x15 = WRAPLOW(dct_const_round_shift(s7 - s15), 8);

+  x0 = WRAPLOW(dct_const_round_shift(s0 + s8));

+  x1 = WRAPLOW(dct_const_round_shift(s1 + s9));

+  x2 = WRAPLOW(dct_const_round_shift(s2 + s10));

+  x3 = WRAPLOW(dct_const_round_shift(s3 + s11));

+  x4 = WRAPLOW(dct_const_round_shift(s4 + s12));

+  x5 = WRAPLOW(dct_const_round_shift(s5 + s13));

+  x6 = WRAPLOW(dct_const_round_shift(s6 + s14));

+  x7 = WRAPLOW(dct_const_round_shift(s7 + s15));

+  x8 = WRAPLOW(dct_const_round_shift(s0 - s8));

+  x9 = WRAPLOW(dct_const_round_shift(s1 - s9));

+  x10 = WRAPLOW(dct_const_round_shift(s2 - s10));

+  x11 = WRAPLOW(dct_const_round_shift(s3 - s11));

+  x12 = WRAPLOW(dct_const_round_shift(s4 - s12));

+  x13 = WRAPLOW(dct_const_round_shift(s5 - s13));

+  x14 = WRAPLOW(dct_const_round_shift(s6 - s14));

+  x15 = WRAPLOW(dct_const_round_shift(s7 - s15));

   // stage 2

   s0 = x0;

@@ -660,22 +660,22 @@

   s14 = - x14 * cospi_12_64 + x15 * cospi_20_64;

   s15 =   x14 * cospi_20_64 + x15 * cospi_12_64;

-  x0 = WRAPLOW(s0 + s4, 8);

-  x1 = WRAPLOW(s1 + s5, 8);

-  x2 = WRAPLOW(s2 + s6, 8);

-  x3 = WRAPLOW(s3 + s7, 8);

-  x4 = WRAPLOW(s0 - s4, 8);

-  x5 = WRAPLOW(s1 - s5, 8);

-  x6 = WRAPLOW(s2 - s6, 8);

-  x7 = WRAPLOW(s3 - s7, 8);

-  x8 = WRAPLOW(dct_const_round_shift(s8 + s12), 8);

-  x9 = WRAPLOW(dct_const_round_shift(s9 + s13), 8);

-  x10 = WRAPLOW(dct_const_round_shift(s10 + s14), 8);

-  x11 = WRAPLOW(dct_const_round_shift(s11 + s15), 8);

-  x12 = WRAPLOW(dct_const_round_shift(s8 - s12), 8);

-  x13 = WRAPLOW(dct_const_round_shift(s9 - s13), 8);

-  x14 = WRAPLOW(dct_const_round_shift(s10 - s14), 8);

-  x15 = WRAPLOW(dct_const_round_shift(s11 - s15), 8);

+  x0 = WRAPLOW(s0 + s4);

+  x1 = WRAPLOW(s1 + s5);

+  x2 = WRAPLOW(s2 + s6);

+  x3 = WRAPLOW(s3 + s7);

+  x4 = WRAPLOW(s0 - s4);

+  x5 = WRAPLOW(s1 - s5);

+  x6 = WRAPLOW(s2 - s6);

+  x7 = WRAPLOW(s3 - s7);

+  x8 = WRAPLOW(dct_const_round_shift(s8 + s12));

+  x9 = WRAPLOW(dct_const_round_shift(s9 + s13));

+  x10 = WRAPLOW(dct_const_round_shift(s10 + s14));

+  x11 = WRAPLOW(dct_const_round_shift(s11 + s15));

+  x12 = WRAPLOW(dct_const_round_shift(s8 - s12));

+  x13 = WRAPLOW(dct_const_round_shift(s9 - s13));

+  x14 = WRAPLOW(dct_const_round_shift(s10 - s14));

+  x15 = WRAPLOW(dct_const_round_shift(s11 - s15));

   // stage 3

   s0 = x0;

@@ -695,22 +695,22 @@

   s14 = - x14 * cospi_24_64 + x15 * cospi_8_64;

   s15 =   x14 * cospi_8_64  + x15 * cospi_24_64;

-  x0 = WRAPLOW(check_range(s0 + s2), 8);

-  x1 = WRAPLOW(check_range(s1 + s3), 8);

-  x2 = WRAPLOW(check_range(s0 - s2), 8);

-  x3 = WRAPLOW(check_range(s1 - s3), 8);

-  x4 = WRAPLOW(dct_const_round_shift(s4 + s6), 8);

-  x5 = WRAPLOW(dct_const_round_shift(s5 + s7), 8);

-  x6 = WRAPLOW(dct_const_round_shift(s4 - s6), 8);

-  x7 = WRAPLOW(dct_const_round_shift(s5 - s7), 8);

-  x8 = WRAPLOW(check_range(s8 + s10), 8);

-  x9 = WRAPLOW(check_range(s9 + s11), 8);

-  x10 = WRAPLOW(check_range(s8 - s10), 8);

-  x11 = WRAPLOW(check_range(s9 - s11), 8);

-  x12 = WRAPLOW(dct_const_round_shift(s12 + s14), 8);

-  x13 = WRAPLOW(dct_const_round_shift(s13 + s15), 8);

-  x14 = WRAPLOW(dct_const_round_shift(s12 - s14), 8);

-  x15 = WRAPLOW(dct_const_round_shift(s13 - s15), 8);

+  x0 = WRAPLOW(s0 + s2);

+  x1 = WRAPLOW(s1 + s3);

+  x2 = WRAPLOW(s0 - s2);

+  x3 = WRAPLOW(s1 - s3);

+  x4 = WRAPLOW(dct_const_round_shift(s4 + s6));

+  x5 = WRAPLOW(dct_const_round_shift(s5 + s7));

+  x6 = WRAPLOW(dct_const_round_shift(s4 - s6));

+  x7 = WRAPLOW(dct_const_round_shift(s5 - s7));

+  x8 = WRAPLOW(s8 + s10);

+  x9 = WRAPLOW(s9 + s11);

+  x10 = WRAPLOW(s8 - s10);

+  x11 = WRAPLOW(s9 - s11);

+  x12 = WRAPLOW(dct_const_round_shift(s12 + s14));

+  x13 = WRAPLOW(dct_const_round_shift(s13 + s15));

+  x14 = WRAPLOW(dct_const_round_shift(s12 - s14));

+  x15 = WRAPLOW(dct_const_round_shift(s13 - s15));

   // stage 4

   s2 = (- cospi_16_64) * (x2 + x3);

@@ -722,31 +722,31 @@

   s14 = (- cospi_16_64) * (x14 + x15);

   s15 = cospi_16_64 * (x14 - x15);

-  x2 = WRAPLOW(dct_const_round_shift(s2), 8);

-  x3 = WRAPLOW(dct_const_round_shift(s3), 8);

-  x6 = WRAPLOW(dct_const_round_shift(s6), 8);

-  x7 = WRAPLOW(dct_const_round_shift(s7), 8);

-  x10 = WRAPLOW(dct_const_round_shift(s10), 8);

-  x11 = WRAPLOW(dct_const_round_shift(s11), 8);

-  x14 = WRAPLOW(dct_const_round_shift(s14), 8);

-  x15 = WRAPLOW(dct_const_round_shift(s15), 8);

+  x2 = WRAPLOW(dct_const_round_shift(s2));

+  x3 = WRAPLOW(dct_const_round_shift(s3));

+  x6 = WRAPLOW(dct_const_round_shift(s6));

+  x7 = WRAPLOW(dct_const_round_shift(s7));

+  x10 = WRAPLOW(dct_const_round_shift(s10));

+  x11 = WRAPLOW(dct_const_round_shift(s11));

+  x14 = WRAPLOW(dct_const_round_shift(s14));

+  x15 = WRAPLOW(dct_const_round_shift(s15));

-  output[0] = WRAPLOW(x0, 8);

-  output[1] = WRAPLOW(-x8, 8);

-  output[2] = WRAPLOW(x12, 8);

-  output[3] = WRAPLOW(-x4, 8);

-  output[4] = WRAPLOW(x6, 8);

-  output[5] = WRAPLOW(x14, 8);

-  output[6] = WRAPLOW(x10, 8);

-  output[7] = WRAPLOW(x2, 8);

-  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);

+  output[0] = WRAPLOW(x0);

+  output[1] = WRAPLOW(-x8);

+  output[2] = WRAPLOW(x12);

+  output[3] = WRAPLOW(-x4);

+  output[4] = WRAPLOW(x6);

+  output[5] = WRAPLOW(x14);

+  output[6] = WRAPLOW(x10);

+  output[7] = WRAPLOW(x2);

+  output[8] = WRAPLOW(x3);

+  output[9] = WRAPLOW(x11);

+  output[10] = WRAPLOW(x15);

+  output[11] = WRAPLOW(x7);

+  output[12] = WRAPLOW(x5);

+  output[13] = WRAPLOW(-x13);

+  output[14] = WRAPLOW(x9);

+  output[15] = WRAPLOW(-x1);

 }

 void vpx_idct16x16_10_add_c(const tran_low_t *input, uint8_t *dest,

@@ -779,8 +779,8 @@

 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);

+  tran_low_t out = WRAPLOW(dct_const_round_shift(input[0] * cospi_16_64));

+  out = WRAPLOW(dct_const_round_shift(out * cospi_16_64));

   a1 = ROUND_POWER_OF_TWO(out, 6);

   for (j = 0; j < 16; ++j) {

     for (i = 0; i < 16; ++i)

@@ -813,43 +813,43 @@

   temp1 = input[1] * cospi_31_64 - input[31] * cospi_1_64;

   temp2 = input[1] * cospi_1_64 + input[31] * cospi_31_64;

-  step1[16] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[31] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[16] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[31] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = input[17] * cospi_15_64 - input[15] * cospi_17_64;

   temp2 = input[17] * cospi_17_64 + input[15] * cospi_15_64;

-  step1[17] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[30] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[17] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[30] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = input[9] * cospi_23_64 - input[23] * cospi_9_64;

   temp2 = input[9] * cospi_9_64 + input[23] * cospi_23_64;

-  step1[18] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[29] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[18] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[29] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = input[25] * cospi_7_64 - input[7] * cospi_25_64;

   temp2 = input[25] * cospi_25_64 + input[7] * cospi_7_64;

-  step1[19] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[28] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[19] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[28] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = input[5] * cospi_27_64 - input[27] * cospi_5_64;

   temp2 = input[5] * cospi_5_64 + input[27] * cospi_27_64;

-  step1[20] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[27] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[20] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[27] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = input[21] * cospi_11_64 - input[11] * cospi_21_64;

   temp2 = input[21] * cospi_21_64 + input[11] * cospi_11_64;

-  step1[21] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[26] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[21] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[26] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = input[13] * cospi_19_64 - input[19] * cospi_13_64;

   temp2 = input[13] * cospi_13_64 + input[19] * cospi_19_64;

-  step1[22] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[25] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[22] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[25] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = input[29] * cospi_3_64 - input[3] * cospi_29_64;

   temp2 = input[29] * cospi_29_64 + input[3] * cospi_3_64;

-  step1[23] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[24] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[23] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[24] = WRAPLOW(dct_const_round_shift(temp2));

   // stage 2

   step2[0] = step1[0];

@@ -863,40 +863,40 @@

   temp1 = step1[8] * cospi_30_64 - step1[15] * cospi_2_64;

   temp2 = step1[8] * cospi_2_64 + step1[15] * cospi_30_64;

-  step2[8] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[15] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[8] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[15] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step1[9] * cospi_14_64 - step1[14] * cospi_18_64;

   temp2 = step1[9] * cospi_18_64 + step1[14] * cospi_14_64;

-  step2[9] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[14] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[9] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[14] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step1[10] * cospi_22_64 - step1[13] * cospi_10_64;

   temp2 = step1[10] * cospi_10_64 + step1[13] * cospi_22_64;

-  step2[10] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[13] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[10] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[13] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step1[11] * cospi_6_64 - step1[12] * cospi_26_64;

   temp2 = step1[11] * cospi_26_64 + step1[12] * cospi_6_64;

-  step2[11] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[12] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[11] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[12] = WRAPLOW(dct_const_round_shift(temp2));

-  step2[16] = WRAPLOW(step1[16] + step1[17], 8);

-  step2[17] = WRAPLOW(step1[16] - step1[17], 8);

-  step2[18] = WRAPLOW(-step1[18] + step1[19], 8);

-  step2[19] = WRAPLOW(step1[18] + step1[19], 8);

-  step2[20] = WRAPLOW(step1[20] + step1[21], 8);

-  step2[21] = WRAPLOW(step1[20] - step1[21], 8);

-  step2[22] = WRAPLOW(-step1[22] + step1[23], 8);

-  step2[23] = WRAPLOW(step1[22] + step1[23], 8);

-  step2[24] = WRAPLOW(step1[24] + step1[25], 8);

-  step2[25] = WRAPLOW(step1[24] - step1[25], 8);

-  step2[26] = WRAPLOW(-step1[26] + step1[27], 8);

-  step2[27] = WRAPLOW(step1[26] + step1[27], 8);

-  step2[28] = WRAPLOW(step1[28] + step1[29], 8);

-  step2[29] = WRAPLOW(step1[28] - step1[29], 8);

-  step2[30] = WRAPLOW(-step1[30] + step1[31], 8);

-  step2[31] = WRAPLOW(step1[30] + step1[31], 8);

+  step2[16] = WRAPLOW(step1[16] + step1[17]);

+  step2[17] = WRAPLOW(step1[16] - step1[17]);

+  step2[18] = WRAPLOW(-step1[18] + step1[19]);

+  step2[19] = WRAPLOW(step1[18] + step1[19]);

+  step2[20] = WRAPLOW(step1[20] + step1[21]);

+  step2[21] = WRAPLOW(step1[20] - step1[21]);

+  step2[22] = WRAPLOW(-step1[22] + step1[23]);

+  step2[23] = WRAPLOW(step1[22] + step1[23]);

+  step2[24] = WRAPLOW(step1[24] + step1[25]);

+  step2[25] = WRAPLOW(step1[24] - step1[25]);

+  step2[26] = WRAPLOW(-step1[26] + step1[27]);

+  step2[27] = WRAPLOW(step1[26] + step1[27]);

+  step2[28] = WRAPLOW(step1[28] + step1[29]);

+  step2[29] = WRAPLOW(step1[28] - step1[29]);

+  step2[30] = WRAPLOW(-step1[30] + step1[31]);

+  step2[31] = WRAPLOW(step1[30] + step1[31]);

   // stage 3

   step1[0] = step2[0];

@@ -906,42 +906,42 @@

   temp1 = step2[4] * cospi_28_64 - step2[7] * cospi_4_64;

   temp2 = step2[4] * cospi_4_64 + step2[7] * cospi_28_64;

-  step1[4] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[7] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[4] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[7] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step2[5] * cospi_12_64 - step2[6] * cospi_20_64;

   temp2 = step2[5] * cospi_20_64 + step2[6] * cospi_12_64;

-  step1[5] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[6] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[5] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[6] = WRAPLOW(dct_const_round_shift(temp2));

-  step1[8] = WRAPLOW(step2[8] + step2[9], 8);

-  step1[9] = WRAPLOW(step2[8] - step2[9], 8);

-  step1[10] = WRAPLOW(-step2[10] + step2[11], 8);

-  step1[11] = WRAPLOW(step2[10] + step2[11], 8);

-  step1[12] = WRAPLOW(step2[12] + step2[13], 8);

-  step1[13] = WRAPLOW(step2[12] - step2[13], 8);

-  step1[14] = WRAPLOW(-step2[14] + step2[15], 8);

-  step1[15] = WRAPLOW(step2[14] + step2[15], 8);

+  step1[8] = WRAPLOW(step2[8] + step2[9]);

+  step1[9] = WRAPLOW(step2[8] - step2[9]);

+  step1[10] = WRAPLOW(-step2[10] + step2[11]);

+  step1[11] = WRAPLOW(step2[10] + step2[11]);

+  step1[12] = WRAPLOW(step2[12] + step2[13]);

+  step1[13] = WRAPLOW(step2[12] - step2[13]);

+  step1[14] = WRAPLOW(-step2[14] + step2[15]);

+  step1[15] = WRAPLOW(step2[14] + step2[15]);

   step1[16] = step2[16];

   step1[31] = step2[31];

   temp1 = -step2[17] * cospi_4_64 + step2[30] * cospi_28_64;

   temp2 = step2[17] * cospi_28_64 + step2[30] * cospi_4_64;

-  step1[17] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[30] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[17] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[30] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = -step2[18] * cospi_28_64 - step2[29] * cospi_4_64;

   temp2 = -step2[18] * cospi_4_64 + step2[29] * cospi_28_64;

-  step1[18] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[29] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[18] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[29] = WRAPLOW(dct_const_round_shift(temp2));

   step1[19] = step2[19];

   step1[20] = step2[20];

   temp1 = -step2[21] * cospi_20_64 + step2[26] * cospi_12_64;

   temp2 = step2[21] * cospi_12_64 + step2[26] * cospi_20_64;

-  step1[21] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[26] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[21] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[26] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = -step2[22] * cospi_12_64 - step2[25] * cospi_20_64;

   temp2 = -step2[22] * cospi_20_64 + step2[25] * cospi_12_64;

-  step1[22] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[25] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[22] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[25] = WRAPLOW(dct_const_round_shift(temp2));

   step1[23] = step2[23];

   step1[24] = step2[24];

   step1[27] = step2[27];

@@ -950,87 +950,87 @@

   // stage 4

   temp1 = (step1[0] + step1[1]) * cospi_16_64;

   temp2 = (step1[0] - step1[1]) * cospi_16_64;

-  step2[0] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[1] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[0] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[1] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = step1[2] * cospi_24_64 - step1[3] * cospi_8_64;

   temp2 = step1[2] * cospi_8_64 + step1[3] * cospi_24_64;

-  step2[2] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[3] = WRAPLOW(dct_const_round_shift(temp2), 8);

-  step2[4] = WRAPLOW(step1[4] + step1[5], 8);

-  step2[5] = WRAPLOW(step1[4] - step1[5], 8);

-  step2[6] = WRAPLOW(-step1[6] + step1[7], 8);

-  step2[7] = WRAPLOW(step1[6] + step1[7], 8);

+  step2[2] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[3] = WRAPLOW(dct_const_round_shift(temp2));

+  step2[4] = WRAPLOW(step1[4] + step1[5]);

+  step2[5] = WRAPLOW(step1[4] - step1[5]);

+  step2[6] = WRAPLOW(-step1[6] + step1[7]);

+  step2[7] = WRAPLOW(step1[6] + step1[7]);

   step2[8] = step1[8];

   step2[15] = step1[15];

   temp1 = -step1[9] * cospi_8_64 + step1[14] * cospi_24_64;

   temp2 = step1[9] * cospi_24_64 + step1[14] * cospi_8_64;

-  step2[9] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[14] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[9] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[14] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = -step1[10] * cospi_24_64 - step1[13] * cospi_8_64;

   temp2 = -step1[10] * cospi_8_64 + step1[13] * cospi_24_64;

-  step2[10] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[13] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[10] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[13] = WRAPLOW(dct_const_round_shift(temp2));

   step2[11] = step1[11];

   step2[12] = step1[12];

-  step2[16] = WRAPLOW(step1[16] + step1[19], 8);

-  step2[17] = WRAPLOW(step1[17] + step1[18], 8);

-  step2[18] = WRAPLOW(step1[17] - step1[18], 8);

-  step2[19] = WRAPLOW(step1[16] - step1[19], 8);

-  step2[20] = WRAPLOW(-step1[20] + step1[23], 8);

-  step2[21] = WRAPLOW(-step1[21] + step1[22], 8);

-  step2[22] = WRAPLOW(step1[21] + step1[22], 8);

-  step2[23] = WRAPLOW(step1[20] + step1[23], 8);

+  step2[16] = WRAPLOW(step1[16] + step1[19]);

+  step2[17] = WRAPLOW(step1[17] + step1[18]);

+  step2[18] = WRAPLOW(step1[17] - step1[18]);

+  step2[19] = WRAPLOW(step1[16] - step1[19]);

+  step2[20] = WRAPLOW(-step1[20] + step1[23]);

+  step2[21] = WRAPLOW(-step1[21] + step1[22]);

+  step2[22] = WRAPLOW(step1[21] + step1[22]);

+  step2[23] = WRAPLOW(step1[20] + step1[23]);

-  step2[24] = WRAPLOW(step1[24] + step1[27], 8);

-  step2[25] = WRAPLOW(step1[25] + step1[26], 8);

-  step2[26] = WRAPLOW(step1[25] - step1[26], 8);

-  step2[27] = WRAPLOW(step1[24] - step1[27], 8);

-  step2[28] = WRAPLOW(-step1[28] + step1[31], 8);

-  step2[29] = WRAPLOW(-step1[29] + step1[30], 8);

-  step2[30] = WRAPLOW(step1[29] + step1[30], 8);

-  step2[31] = WRAPLOW(step1[28] + step1[31], 8);

+  step2[24] = WRAPLOW(step1[24] + step1[27]);

+  step2[25] = WRAPLOW(step1[25] + step1[26]);

+  step2[26] = WRAPLOW(step1[25] - step1[26]);

+  step2[27] = WRAPLOW(step1[24] - step1[27]);

+  step2[28] = WRAPLOW(-step1[28] + step1[31]);

+  step2[29] = WRAPLOW(-step1[29] + step1[30]);

+  step2[30] = WRAPLOW(step1[29] + step1[30]);

+  step2[31] = WRAPLOW(step1[28] + step1[31]);

   // stage 5

-  step1[0] = WRAPLOW(step2[0] + step2[3], 8);

-  step1[1] = WRAPLOW(step2[1] + step2[2], 8);

-  step1[2] = WRAPLOW(step2[1] - step2[2], 8);

-  step1[3] = WRAPLOW(step2[0] - step2[3], 8);

+  step1[0] = WRAPLOW(step2[0] + step2[3]);

+  step1[1] = WRAPLOW(step2[1] + step2[2]);

+  step1[2] = WRAPLOW(step2[1] - step2[2]);

+  step1[3] = WRAPLOW(step2[0] - step2[3]);

   step1[4] = step2[4];

   temp1 = (step2[6] - step2[5]) * cospi_16_64;

   temp2 = (step2[5] + step2[6]) * cospi_16_64;

-  step1[5] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[6] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[5] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[6] = WRAPLOW(dct_const_round_shift(temp2));

   step1[7] = step2[7];

-  step1[8] = WRAPLOW(step2[8] + step2[11], 8);

-  step1[9] = WRAPLOW(step2[9] + step2[10], 8);

-  step1[10] = WRAPLOW(step2[9] - step2[10], 8);

-  step1[11] = WRAPLOW(step2[8] - step2[11], 8);

-  step1[12] = WRAPLOW(-step2[12] + step2[15], 8);

-  step1[13] = WRAPLOW(-step2[13] + step2[14], 8);

-  step1[14] = WRAPLOW(step2[13] + step2[14], 8);

-  step1[15] = WRAPLOW(step2[12] + step2[15], 8);

+  step1[8] = WRAPLOW(step2[8] + step2[11]);

+  step1[9] = WRAPLOW(step2[9] + step2[10]);

+  step1[10] = WRAPLOW(step2[9] - step2[10]);

+  step1[11] = WRAPLOW(step2[8] - step2[11]);

+  step1[12] = WRAPLOW(-step2[12] + step2[15]);

+  step1[13] = WRAPLOW(-step2[13] + step2[14]);

+  step1[14] = WRAPLOW(step2[13] + step2[14]);

+  step1[15] = WRAPLOW(step2[12] + step2[15]);

   step1[16] = step2[16];

   step1[17] = step2[17];

   temp1 = -step2[18] * cospi_8_64 + step2[29] * cospi_24_64;

   temp2 = step2[18] * cospi_24_64 + step2[29] * cospi_8_64;

-  step1[18] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[29] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[18] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[29] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = -step2[19] * cospi_8_64 + step2[28] * cospi_24_64;

   temp2 = step2[19] * cospi_24_64 + step2[28] * cospi_8_64;

-  step1[19] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[28] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[19] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[28] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = -step2[20] * cospi_24_64 - step2[27] * cospi_8_64;

   temp2 = -step2[20] * cospi_8_64 + step2[27] * cospi_24_64;

-  step1[20] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[27] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[20] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[27] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = -step2[21] * cospi_24_64 - step2[26] * cospi_8_64;

   temp2 = -step2[21] * cospi_8_64 + step2[26] * cospi_24_64;

-  step1[21] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[26] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[21] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[26] = WRAPLOW(dct_const_round_shift(temp2));

   step1[22] = step2[22];

   step1[23] = step2[23];

   step1[24] = step2[24];

@@ -1039,62 +1039,62 @@

   step1[31] = step2[31];

   // stage 6

-  step2[0] = WRAPLOW(step1[0] + step1[7], 8);

-  step2[1] = WRAPLOW(step1[1] + step1[6], 8);

-  step2[2] = WRAPLOW(step1[2] + step1[5], 8);

-  step2[3] = WRAPLOW(step1[3] + step1[4], 8);

-  step2[4] = WRAPLOW(step1[3] - step1[4], 8);

-  step2[5] = WRAPLOW(step1[2] - step1[5], 8);

-  step2[6] = WRAPLOW(step1[1] - step1[6], 8);

-  step2[7] = WRAPLOW(step1[0] - step1[7], 8);

+  step2[0] = WRAPLOW(step1[0] + step1[7]);

+  step2[1] = WRAPLOW(step1[1] + step1[6]);

+  step2[2] = WRAPLOW(step1[2] + step1[5]);

+  step2[3] = WRAPLOW(step1[3] + step1[4]);

+  step2[4] = WRAPLOW(step1[3] - step1[4]);

+  step2[5] = WRAPLOW(step1[2] - step1[5]);

+  step2[6] = WRAPLOW(step1[1] - step1[6]);

+  step2[7] = WRAPLOW(step1[0] - step1[7]);

   step2[8] = step1[8];

   step2[9] = step1[9];

   temp1 = (-step1[10] + step1[13]) * cospi_16_64;

   temp2 = (step1[10] + step1[13]) * cospi_16_64;

-  step2[10] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[13] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[10] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[13] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = (-step1[11] + step1[12]) * cospi_16_64;

   temp2 = (step1[11] + step1[12]) * cospi_16_64;

-  step2[11] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step2[12] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step2[11] = WRAPLOW(dct_const_round_shift(temp1));

+  step2[12] = WRAPLOW(dct_const_round_shift(temp2));

   step2[14] = step1[14];

   step2[15] = step1[15];

-  step2[16] = WRAPLOW(step1[16] + step1[23], 8);

-  step2[17] = WRAPLOW(step1[17] + step1[22], 8);

-  step2[18] = WRAPLOW(step1[18] + step1[21], 8);

-  step2[19] = WRAPLOW(step1[19] + step1[20], 8);

-  step2[20] = WRAPLOW(step1[19] - step1[20], 8);

-  step2[21] = WRAPLOW(step1[18] - step1[21], 8);

-  step2[22] = WRAPLOW(step1[17] - step1[22], 8);

-  step2[23] = WRAPLOW(step1[16] - step1[23], 8);

+  step2[16] = WRAPLOW(step1[16] + step1[23]);

+  step2[17] = WRAPLOW(step1[17] + step1[22]);

+  step2[18] = WRAPLOW(step1[18] + step1[21]);

+  step2[19] = WRAPLOW(step1[19] + step1[20]);

+  step2[20] = WRAPLOW(step1[19] - step1[20]);

+  step2[21] = WRAPLOW(step1[18] - step1[21]);

+  step2[22] = WRAPLOW(step1[17] - step1[22]);

+  step2[23] = WRAPLOW(step1[16] - step1[23]);

-  step2[24] = WRAPLOW(-step1[24] + step1[31], 8);

-  step2[25] = WRAPLOW(-step1[25] + step1[30], 8);

-  step2[26] = WRAPLOW(-step1[26] + step1[29], 8);

-  step2[27] = WRAPLOW(-step1[27] + step1[28], 8);

-  step2[28] = WRAPLOW(step1[27] + step1[28], 8);

-  step2[29] = WRAPLOW(step1[26] + step1[29], 8);

-  step2[30] = WRAPLOW(step1[25] + step1[30], 8);

-  step2[31] = WRAPLOW(step1[24] + step1[31], 8);

+  step2[24] = WRAPLOW(-step1[24] + step1[31]);

+  step2[25] = WRAPLOW(-step1[25] + step1[30]);

+  step2[26] = WRAPLOW(-step1[26] + step1[29]);

+  step2[27] = WRAPLOW(-step1[27] + step1[28]);

+  step2[28] = WRAPLOW(step1[27] + step1[28]);

+  step2[29] = WRAPLOW(step1[26] + step1[29]);

+  step2[30] = WRAPLOW(step1[25] + step1[30]);

+  step2[31] = WRAPLOW(step1[24] + step1[31]);

   // stage 7

-  step1[0] = WRAPLOW(step2[0] + step2[15], 8);

-  step1[1] = WRAPLOW(step2[1] + step2[14], 8);

-  step1[2] = WRAPLOW(step2[2] + step2[13], 8);

-  step1[3] = WRAPLOW(step2[3] + step2[12], 8);

-  step1[4] = WRAPLOW(step2[4] + step2[11], 8);

-  step1[5] = WRAPLOW(step2[5] + step2[10], 8);

-  step1[6] = WRAPLOW(step2[6] + step2[9], 8);

-  step1[7] = WRAPLOW(step2[7] + step2[8], 8);

-  step1[8] = WRAPLOW(step2[7] - step2[8], 8);

-  step1[9] = WRAPLOW(step2[6] - step2[9], 8);

-  step1[10] = WRAPLOW(step2[5] - step2[10], 8);

-  step1[11] = WRAPLOW(step2[4] - step2[11], 8);

-  step1[12] = WRAPLOW(step2[3] - step2[12], 8);

-  step1[13] = WRAPLOW(step2[2] - step2[13], 8);

-  step1[14] = WRAPLOW(step2[1] - step2[14], 8);

-  step1[15] = WRAPLOW(step2[0] - step2[15], 8);

+  step1[0] = WRAPLOW(step2[0] + step2[15]);

+  step1[1] = WRAPLOW(step2[1] + step2[14]);

+  step1[2] = WRAPLOW(step2[2] + step2[13]);

+  step1[3] = WRAPLOW(step2[3] + step2[12]);

+  step1[4] = WRAPLOW(step2[4] + step2[11]);

+  step1[5] = WRAPLOW(step2[5] + step2[10]);

+  step1[6] = WRAPLOW(step2[6] + step2[9]);

+  step1[7] = WRAPLOW(step2[7] + step2[8]);

+  step1[8] = WRAPLOW(step2[7] - step2[8]);

+  step1[9] = WRAPLOW(step2[6] - step2[9]);

+  step1[10] = WRAPLOW(step2[5] - step2[10]);

+  step1[11] = WRAPLOW(step2[4] - step2[11]);

+  step1[12] = WRAPLOW(step2[3] - step2[12]);

+  step1[13] = WRAPLOW(step2[2] - step2[13]);

+  step1[14] = WRAPLOW(step2[1] - step2[14]);

+  step1[15] = WRAPLOW(step2[0] - step2[15]);

   step1[16] = step2[16];

   step1[17] = step2[17];

@@ -1102,20 +1102,20 @@

   step1[19] = step2[19];

   temp1 = (-step2[20] + step2[27]) * cospi_16_64;

   temp2 = (step2[20] + step2[27]) * cospi_16_64;

-  step1[20] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[27] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[20] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[27] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = (-step2[21] + step2[26]) * cospi_16_64;

   temp2 = (step2[21] + step2[26]) * cospi_16_64;

-  step1[21] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[26] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[21] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[26] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = (-step2[22] + step2[25]) * cospi_16_64;

   temp2 = (step2[22] + step2[25]) * cospi_16_64;

-  step1[22] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[25] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[22] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[25] = WRAPLOW(dct_const_round_shift(temp2));

   temp1 = (-step2[23] + step2[24]) * cospi_16_64;

   temp2 = (step2[23] + step2[24]) * cospi_16_64;

-  step1[23] = WRAPLOW(dct_const_round_shift(temp1), 8);

-  step1[24] = WRAPLOW(dct_const_round_shift(temp2), 8);

+  step1[23] = WRAPLOW(dct_const_round_shift(temp1));

+  step1[24] = WRAPLOW(dct_const_round_shift(temp2));

   step1[28] = step2[28];

   step1[29] = step2[29];

   step1[30] = step2[30];

@@ -1122,38 +1122,38 @@

   step1[31] = step2[31];

   // final stage

-  output[0] = WRAPLOW(step1[0] + step1[31], 8);

-  output[1] = WRAPLOW(step1[1] + step1[30], 8);

-  output[2] = WRAPLOW(step1[2] + step1[29], 8);

-  output[3] = WRAPLOW(step1[3] + step1[28], 8);

-  output[4] = WRAPLOW(step1[4] + step1[27], 8);

-  output[5] = WRAPLOW(step1[5] + step1[26], 8);

-  output[6] = WRAPLOW(step1[6] + step1[25], 8);

-  output[7] = WRAPLOW(step1[7] + step1[24], 8);

-  output[8] = WRAPLOW(step1[8] + step1[23], 8);

-  output[9] = WRAPLOW(step1[9] + step1[22], 8);

-  output[10] = WRAPLOW(step1[10] + step1[21], 8);

-  output[11] = WRAPLOW(step1[11] + step1[20], 8);

-  output[12] = WRAPLOW(step1[12] + step1[19], 8);

-  output[13] = WRAPLOW(step1[13] + step1[18], 8);

-  output[14] = WRAPLOW(step1[14] + step1[17], 8);

-  output[15] = WRAPLOW(step1[15] + step1[16], 8);

-  output[16] = WRAPLOW(step1[15] - step1[16], 8);

-  output[17] = WRAPLOW(step1[14] - step1[17], 8);

-  output[18] = WRAPLOW(step1[13] - step1[18], 8);

-  output[19] = WRAPLOW(step1[12] - step1[19], 8);

-  output[20] = WRAPLOW(step1[11] - step1[20], 8);

-  output[21] = WRAPLOW(step1[10] - step1[21], 8);

-  output[22] = WRAPLOW(step1[9] - step1[22], 8);

-  output[23] = WRAPLOW(step1[8] - step1[23], 8);

-  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);

+  output[0] = WRAPLOW(step1[0] + step1[31]);

+  output[1] = WRAPLOW(step1[1] + step1[30]);

+  output[2] = WRAPLOW(step1[2] + step1[29]);

+  output[3] = WRAPLOW(step1[3] + step1[28]);

+  output[4] = WRAPLOW(step1[4] + step1[27]);

+  output[5] = WRAPLOW(step1[5] + step1[26]);

+  output[6] = WRAPLOW(step1[6] + step1[25]);

+  output[7] = WRAPLOW(step1[7] + step1[24]);

+  output[8] = WRAPLOW(step1[8] + step1[23]);

+  output[9] = WRAPLOW(step1[9] + step1[22]);

+  output[10] = WRAPLOW(step1[10] + step1[21]);

+  output[11] = WRAPLOW(step1[11] + step1[20]);

+  output[12] = WRAPLOW(step1[12] + step1[19]);

+  output[13] = WRAPLOW(step1[13] + step1[18]);

+  output[14] = WRAPLOW(step1[14] + step1[17]);

+  output[15] = WRAPLOW(step1[15] + step1[16]);

+  output[16] = WRAPLOW(step1[15] - step1[16]);

+  output[17] = WRAPLOW(step1[14] - step1[17]);

+  output[18] = WRAPLOW(step1[13] - step1[18]);

+  output[19] = WRAPLOW(step1[12] - step1[19]);

+  output[20] = WRAPLOW(step1[11] - step1[20]);

+  output[21] = WRAPLOW(step1[10] - step1[21]);

+  output[22] = WRAPLOW(step1[9] - step1[22]);

+  output[23] = WRAPLOW(step1[8] - step1[23]);

+  output[24] = WRAPLOW(step1[7] - step1[24]);

+  output[25] = WRAPLOW(step1[6] - step1[25]);

+  output[26] = WRAPLOW(step1[5] - step1[26]);

+  output[27] = WRAPLOW(step1[4] - step1[27]);

+  output[28] = WRAPLOW(step1[3] - step1[28]);

+  output[29] = WRAPLOW(step1[2] - step1[29]);

+  output[30] = WRAPLOW(step1[1] - step1[30]);

+  output[31] = WRAPLOW(step1[0] - step1[31]);

 }

 void vpx_idct32x32_1024_add_c(const tran_low_t *input, uint8_t *dest,

@@ -1253,8 +1253,8 @@

   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);

+  tran_low_t out = WRAPLOW(dct_const_round_shift(input[0] * cospi_16_64));

+  out = WRAPLOW(dct_const_round_shift(out * cospi_16_64));

   a1 = ROUND_POWER_OF_TWO(out, 6);

   for (j = 0; j < 32; ++j) {

@@ -1288,10 +1288,10 @@

     c1 = e1 - c1;

     a1 -= b1;

     d1 += c1;

-    op[0] = WRAPLOW(a1, bd);

-    op[1] = WRAPLOW(b1, bd);

-    op[2] = WRAPLOW(c1, bd);

-    op[3] = WRAPLOW(d1, bd);

+    op[0] = HIGHBD_WRAPLOW(a1, bd);

+    op[1] = HIGHBD_WRAPLOW(b1, bd);

+    op[2] = HIGHBD_WRAPLOW(c1, bd);

+    op[3] = HIGHBD_WRAPLOW(d1, bd);

     ip += 4;

     op += 4;

   }

@@ -1332,8 +1332,8 @@

   a1 = ip[0] >> UNIT_QUANT_SHIFT;

   e1 = a1 >> 1;

   a1 -= e1;

-  op[0] = WRAPLOW(a1, bd);

-  op[1] = op[2] = op[3] = WRAPLOW(e1, bd);

+  op[0] = HIGHBD_WRAPLOW(a1, bd);

+  op[1] = op[2] = op[3] = HIGHBD_WRAPLOW(e1, bd);

   ip = tmp;

   for (i = 0; i < 4; i++) {

@@ -1359,18 +1359,18 @@

   // 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);

+  step[0] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step[1] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), 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);

+  step[2] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step[3] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), 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);

+  output[0] = HIGHBD_WRAPLOW(step[0] + step[3], bd);

+  output[1] = HIGHBD_WRAPLOW(step[1] + step[2], bd);

+  output[2] = HIGHBD_WRAPLOW(step[1] - step[2], bd);

+  output[3] = HIGHBD_WRAPLOW(step[0] - step[3], bd);

 }

 void vpx_highbd_idct4x4_16_add_c(const tran_low_t *input, uint8_t *dest8,

@@ -1404,11 +1404,11 @@

                                 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);

+  tran_low_t out = HIGHBD_WRAPLOW(

+      highbd_dct_const_round_shift(input[0] * cospi_16_64), bd);

   uint16_t *dest = CONVERT_TO_SHORTPTR(dest8);

-  out = WRAPLOW(highbd_dct_const_round_shift(out * cospi_16_64, bd), bd);

+  out = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(out * cospi_16_64), bd);

   a1 = ROUND_POWER_OF_TWO(out, 4);

   for (i = 0; i < 4; i++) {

@@ -1430,39 +1430,39 @@

   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);

+  step1[4] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[7] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), 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);

+  step1[5] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[6] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   // stage 2 & stage 3 - even half

   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);

+  step2[4] = HIGHBD_WRAPLOW(step1[4] + step1[5], bd);

+  step2[5] = HIGHBD_WRAPLOW(step1[4] - step1[5], bd);

+  step2[6] = HIGHBD_WRAPLOW(-step1[6] + step1[7], bd);

+  step2[7] = HIGHBD_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[5] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[6] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), 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);

+  output[0] = HIGHBD_WRAPLOW(step1[0] + step1[7], bd);

+  output[1] = HIGHBD_WRAPLOW(step1[1] + step1[6], bd);

+  output[2] = HIGHBD_WRAPLOW(step1[2] + step1[5], bd);

+  output[3] = HIGHBD_WRAPLOW(step1[3] + step1[4], bd);

+  output[4] = HIGHBD_WRAPLOW(step1[3] - step1[4], bd);

+  output[5] = HIGHBD_WRAPLOW(step1[2] - step1[5], bd);

+  output[6] = HIGHBD_WRAPLOW(step1[1] - step1[6], bd);

+  output[7] = HIGHBD_WRAPLOW(step1[0] - step1[7], bd);

 }

 void vpx_highbd_idct8x8_64_add_c(const tran_low_t *input, uint8_t *dest8,

@@ -1496,10 +1496,10 @@

                                 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);

+  tran_low_t out = HIGHBD_WRAPLOW(

+      highbd_dct_const_round_shift(input[0] * cospi_16_64), bd);

   uint16_t *dest = CONVERT_TO_SHORTPTR(dest8);

-  out = WRAPLOW(highbd_dct_const_round_shift(out * cospi_16_64, bd), bd);

+  out = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(out * cospi_16_64), bd);

   a1 = ROUND_POWER_OF_TWO(out, 5);

   for (j = 0; j < 8; ++j) {

     for (i = 0; i < 8; ++i)

@@ -1540,10 +1540,10 @@

   // 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);

+  output[0] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s0 + s3), bd);

+  output[1] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s1 + s3), bd);

+  output[2] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s2), bd);

+  output[3] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s0 + s1 - s3), bd);

 }

 void vpx_highbd_iadst8_c(const tran_low_t *input, tran_low_t *output, int bd) {

@@ -1574,14 +1574,14 @@

   s6 = cospi_26_64 * x6 + cospi_6_64  * x7;

   s7 = cospi_6_64  * x6 - cospi_26_64 * x7;

-  x0 = WRAPLOW(highbd_dct_const_round_shift(s0 + s4, bd), bd);

-  x1 = WRAPLOW(highbd_dct_const_round_shift(s1 + s5, bd), bd);

-  x2 = WRAPLOW(highbd_dct_const_round_shift(s2 + s6, bd), bd);

-  x3 = WRAPLOW(highbd_dct_const_round_shift(s3 + s7, bd), bd);

-  x4 = WRAPLOW(highbd_dct_const_round_shift(s0 - s4, bd), bd);

-  x5 = WRAPLOW(highbd_dct_const_round_shift(s1 - s5, bd), bd);

-  x6 = WRAPLOW(highbd_dct_const_round_shift(s2 - s6, bd), bd);

-  x7 = WRAPLOW(highbd_dct_const_round_shift(s3 - s7, bd), bd);

+  x0 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s0 + s4), bd);

+  x1 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s1 + s5), bd);

+  x2 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s2 + s6), bd);

+  x3 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s3 + s7), bd);

+  x4 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s0 - s4), bd);

+  x5 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s1 - s5), bd);

+  x6 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s2 - s6), bd);

+  x7 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s3 - s7), bd);

   // stage 2

   s0 = x0;

@@ -1593,14 +1593,14 @@

   s6 = -cospi_24_64 * x6 + cospi_8_64  * x7;

   s7 =  cospi_8_64  * x6 + cospi_24_64 * x7;

-  x0 = WRAPLOW(s0 + s2, bd);

-  x1 = WRAPLOW(s1 + s3, bd);

-  x2 = WRAPLOW(s0 - s2, bd);

-  x3 = WRAPLOW(s1 - s3, bd);

-  x4 = WRAPLOW(highbd_dct_const_round_shift(s4 + s6, bd), bd);

-  x5 = WRAPLOW(highbd_dct_const_round_shift(s5 + s7, bd), bd);

-  x6 = WRAPLOW(highbd_dct_const_round_shift(s4 - s6, bd), bd);

-  x7 = WRAPLOW(highbd_dct_const_round_shift(s5 - s7, bd), bd);

+  x0 = HIGHBD_WRAPLOW(s0 + s2, bd);

+  x1 = HIGHBD_WRAPLOW(s1 + s3, bd);

+  x2 = HIGHBD_WRAPLOW(s0 - s2, bd);

+  x3 = HIGHBD_WRAPLOW(s1 - s3, bd);

+  x4 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s4 + s6), bd);

+  x5 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s5 + s7), bd);

+  x6 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s4 - s6), bd);

+  x7 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s5 - s7), bd);

   // stage 3

   s2 = cospi_16_64 * (x2 + x3);

@@ -1608,19 +1608,19 @@

   s6 = cospi_16_64 * (x6 + x7);

   s7 = cospi_16_64 * (x6 - x7);

-  x2 = WRAPLOW(highbd_dct_const_round_shift(s2, bd), bd);

-  x3 = WRAPLOW(highbd_dct_const_round_shift(s3, bd), bd);

-  x6 = WRAPLOW(highbd_dct_const_round_shift(s6, bd), bd);

-  x7 = WRAPLOW(highbd_dct_const_round_shift(s7, bd), bd);

+  x2 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s2), bd);

+  x3 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s3), bd);

+  x6 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s6), bd);

+  x7 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s7), bd);

-  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);

+  output[0] = HIGHBD_WRAPLOW(x0, bd);

+  output[1] = HIGHBD_WRAPLOW(-x4, bd);

+  output[2] = HIGHBD_WRAPLOW(x6, bd);

+  output[3] = HIGHBD_WRAPLOW(-x2, bd);

+  output[4] = HIGHBD_WRAPLOW(x3, bd);

+  output[5] = HIGHBD_WRAPLOW(-x7, bd);

+  output[6] = HIGHBD_WRAPLOW(x5, bd);

+  output[7] = HIGHBD_WRAPLOW(-x1, bd);

 }

 void vpx_highbd_idct8x8_10_add_c(const tran_low_t *input, uint8_t *dest8,

@@ -1685,23 +1685,23 @@

   temp1 = step1[8] * cospi_30_64 - step1[15] * cospi_2_64;

   temp2 = step1[8] * cospi_2_64 + step1[15] * cospi_30_64;

-  step2[8] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[15] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[8] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[15] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = step1[9] * cospi_14_64 - step1[14] * cospi_18_64;

   temp2 = step1[9] * cospi_18_64 + step1[14] * cospi_14_64;

-  step2[9] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[14] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[9] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[14] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = step1[10] * cospi_22_64 - step1[13] * cospi_10_64;

   temp2 = step1[10] * cospi_10_64 + step1[13] * cospi_22_64;

-  step2[10] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[13] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[10] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[13] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = step1[11] * cospi_6_64 - step1[12] * cospi_26_64;

   temp2 = step1[11] * cospi_26_64 + step1[12] * cospi_6_64;

-  step2[11] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[12] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[11] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[12] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   // stage 3

   step1[0] = step2[0];

@@ -1711,109 +1711,109 @@

   temp1 = step2[4] * cospi_28_64 - step2[7] * cospi_4_64;

   temp2 = step2[4] * cospi_4_64 + step2[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);

+  step1[4] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[7] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = step2[5] * cospi_12_64 - step2[6] * cospi_20_64;

   temp2 = step2[5] * cospi_20_64 + step2[6] * 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);

+  step1[5] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[6] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

-  step1[8] = WRAPLOW(step2[8] + step2[9], bd);

-  step1[9] = WRAPLOW(step2[8] - step2[9], bd);

-  step1[10] = WRAPLOW(-step2[10] + step2[11], bd);

-  step1[11] = WRAPLOW(step2[10] + step2[11], bd);

-  step1[12] = WRAPLOW(step2[12] + step2[13], bd);

-  step1[13] = WRAPLOW(step2[12] - step2[13], bd);

-  step1[14] = WRAPLOW(-step2[14] + step2[15], bd);

-  step1[15] = WRAPLOW(step2[14] + step2[15], bd);

+  step1[8] = HIGHBD_WRAPLOW(step2[8] + step2[9], bd);

+  step1[9] = HIGHBD_WRAPLOW(step2[8] - step2[9], bd);

+  step1[10] = HIGHBD_WRAPLOW(-step2[10] + step2[11], bd);

+  step1[11] = HIGHBD_WRAPLOW(step2[10] + step2[11], bd);

+  step1[12] = HIGHBD_WRAPLOW(step2[12] + step2[13], bd);

+  step1[13] = HIGHBD_WRAPLOW(step2[12] - step2[13], bd);

+  step1[14] = HIGHBD_WRAPLOW(-step2[14] + step2[15], bd);

+  step1[15] = HIGHBD_WRAPLOW(step2[14] + step2[15], bd);

   // stage 4

   temp1 = (step1[0] + step1[1]) * cospi_16_64;

   temp2 = (step1[0] - step1[1]) * cospi_16_64;

-  step2[0] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[1] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[0] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[1] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = step1[2] * cospi_24_64 - step1[3] * cospi_8_64;

   temp2 = step1[2] * cospi_8_64 + step1[3] * cospi_24_64;

-  step2[2] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[3] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

-  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);

+  step2[2] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[3] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

+  step2[4] = HIGHBD_WRAPLOW(step1[4] + step1[5], bd);

+  step2[5] = HIGHBD_WRAPLOW(step1[4] - step1[5], bd);

+  step2[6] = HIGHBD_WRAPLOW(-step1[6] + step1[7], bd);

+  step2[7] = HIGHBD_WRAPLOW(step1[6] + step1[7], bd);

   step2[8] = step1[8];

   step2[15] = step1[15];

   temp1 = -step1[9] * cospi_8_64 + step1[14] * cospi_24_64;

   temp2 = step1[9] * cospi_24_64 + step1[14] * cospi_8_64;

-  step2[9] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[14] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[9] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[14] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = -step1[10] * cospi_24_64 - step1[13] * cospi_8_64;

   temp2 = -step1[10] * cospi_8_64 + step1[13] * cospi_24_64;

-  step2[10] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[13] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[10] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[13] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   step2[11] = step1[11];

   step2[12] = step1[12];

   // stage 5

-  step1[0] = WRAPLOW(step2[0] + step2[3], bd);

-  step1[1] = WRAPLOW(step2[1] + step2[2], bd);

-  step1[2] = WRAPLOW(step2[1] - step2[2], bd);

-  step1[3] = WRAPLOW(step2[0] - step2[3], bd);

+  step1[0] = HIGHBD_WRAPLOW(step2[0] + step2[3], bd);

+  step1[1] = HIGHBD_WRAPLOW(step2[1] + step2[2], bd);

+  step1[2] = HIGHBD_WRAPLOW(step2[1] - step2[2], bd);

+  step1[3] = HIGHBD_WRAPLOW(step2[0] - step2[3], bd);

   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[5] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[6] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   step1[7] = step2[7];

-  step1[8] = WRAPLOW(step2[8] + step2[11], bd);

-  step1[9] = WRAPLOW(step2[9] + step2[10], bd);

-  step1[10] = WRAPLOW(step2[9] - step2[10], bd);

-  step1[11] = WRAPLOW(step2[8] - step2[11], bd);

-  step1[12] = WRAPLOW(-step2[12] + step2[15], bd);

-  step1[13] = WRAPLOW(-step2[13] + step2[14], bd);

-  step1[14] = WRAPLOW(step2[13] + step2[14], bd);

-  step1[15] = WRAPLOW(step2[12] + step2[15], bd);

+  step1[8] = HIGHBD_WRAPLOW(step2[8] + step2[11], bd);

+  step1[9] = HIGHBD_WRAPLOW(step2[9] + step2[10], bd);

+  step1[10] = HIGHBD_WRAPLOW(step2[9] - step2[10], bd);

+  step1[11] = HIGHBD_WRAPLOW(step2[8] - step2[11], bd);

+  step1[12] = HIGHBD_WRAPLOW(-step2[12] + step2[15], bd);

+  step1[13] = HIGHBD_WRAPLOW(-step2[13] + step2[14], bd);

+  step1[14] = HIGHBD_WRAPLOW(step2[13] + step2[14], bd);

+  step1[15] = HIGHBD_WRAPLOW(step2[12] + step2[15], bd);

   // stage 6

-  step2[0] = WRAPLOW(step1[0] + step1[7], bd);

-  step2[1] = WRAPLOW(step1[1] + step1[6], bd);

-  step2[2] = WRAPLOW(step1[2] + step1[5], bd);

-  step2[3] = WRAPLOW(step1[3] + step1[4], bd);

-  step2[4] = WRAPLOW(step1[3] - step1[4], bd);

-  step2[5] = WRAPLOW(step1[2] - step1[5], bd);

-  step2[6] = WRAPLOW(step1[1] - step1[6], bd);

-  step2[7] = WRAPLOW(step1[0] - step1[7], bd);

+  step2[0] = HIGHBD_WRAPLOW(step1[0] + step1[7], bd);

+  step2[1] = HIGHBD_WRAPLOW(step1[1] + step1[6], bd);

+  step2[2] = HIGHBD_WRAPLOW(step1[2] + step1[5], bd);

+  step2[3] = HIGHBD_WRAPLOW(step1[3] + step1[4], bd);

+  step2[4] = HIGHBD_WRAPLOW(step1[3] - step1[4], bd);

+  step2[5] = HIGHBD_WRAPLOW(step1[2] - step1[5], bd);

+  step2[6] = HIGHBD_WRAPLOW(step1[1] - step1[6], bd);

+  step2[7] = HIGHBD_WRAPLOW(step1[0] - step1[7], bd);

   step2[8] = step1[8];

   step2[9] = step1[9];

   temp1 = (-step1[10] + step1[13]) * cospi_16_64;

   temp2 = (step1[10] + step1[13]) * cospi_16_64;

-  step2[10] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[13] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[10] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[13] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = (-step1[11] + step1[12]) * cospi_16_64;

   temp2 = (step1[11] + step1[12]) * cospi_16_64;

-  step2[11] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[12] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[11] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[12] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   step2[14] = step1[14];

   step2[15] = step1[15];

   // stage 7

-  output[0] = WRAPLOW(step2[0] + step2[15], bd);

-  output[1] = WRAPLOW(step2[1] + step2[14], bd);

-  output[2] = WRAPLOW(step2[2] + step2[13], bd);

-  output[3] = WRAPLOW(step2[3] + step2[12], bd);

-  output[4] = WRAPLOW(step2[4] + step2[11], bd);

-  output[5] = WRAPLOW(step2[5] + step2[10], bd);

-  output[6] = WRAPLOW(step2[6] + step2[9], bd);

-  output[7] = WRAPLOW(step2[7] + step2[8], bd);

-  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);

+  output[0] = HIGHBD_WRAPLOW(step2[0] + step2[15], bd);

+  output[1] = HIGHBD_WRAPLOW(step2[1] + step2[14], bd);

+  output[2] = HIGHBD_WRAPLOW(step2[2] + step2[13], bd);

+  output[3] = HIGHBD_WRAPLOW(step2[3] + step2[12], bd);

+  output[4] = HIGHBD_WRAPLOW(step2[4] + step2[11], bd);

+  output[5] = HIGHBD_WRAPLOW(step2[5] + step2[10], bd);

+  output[6] = HIGHBD_WRAPLOW(step2[6] + step2[9], bd);

+  output[7] = HIGHBD_WRAPLOW(step2[7] + step2[8], bd);

+  output[8] = HIGHBD_WRAPLOW(step2[7] - step2[8], bd);

+  output[9] = HIGHBD_WRAPLOW(step2[6] - step2[9], bd);

+  output[10] = HIGHBD_WRAPLOW(step2[5] - step2[10], bd);

+  output[11] = HIGHBD_WRAPLOW(step2[4] - step2[11], bd);

+  output[12] = HIGHBD_WRAPLOW(step2[3] - step2[12], bd);

+  output[13] = HIGHBD_WRAPLOW(step2[2] - step2[13], bd);

+  output[14] = HIGHBD_WRAPLOW(step2[1] - step2[14], bd);

+  output[15] = HIGHBD_WRAPLOW(step2[0] - step2[15], bd);

 }

 void vpx_highbd_idct16x16_256_add_c(const tran_low_t *input, uint8_t *dest8,

@@ -1889,22 +1889,22 @@

   s14 = x14 * cospi_29_64 + x15 * cospi_3_64;

   s15 = x14 * cospi_3_64  - x15 * cospi_29_64;

-  x0 = WRAPLOW(highbd_dct_const_round_shift(s0 + s8, bd), bd);

-  x1 = WRAPLOW(highbd_dct_const_round_shift(s1 + s9, bd), bd);

-  x2 = WRAPLOW(highbd_dct_const_round_shift(s2 + s10, bd), bd);

-  x3 = WRAPLOW(highbd_dct_const_round_shift(s3 + s11, bd), bd);

-  x4 = WRAPLOW(highbd_dct_const_round_shift(s4 + s12, bd), bd);

-  x5 = WRAPLOW(highbd_dct_const_round_shift(s5 + s13, bd), bd);

-  x6 = WRAPLOW(highbd_dct_const_round_shift(s6 + s14, bd), bd);

-  x7 = WRAPLOW(highbd_dct_const_round_shift(s7 + s15, bd), bd);

-  x8  = WRAPLOW(highbd_dct_const_round_shift(s0 - s8, bd), bd);

-  x9  = WRAPLOW(highbd_dct_const_round_shift(s1 - s9, bd), bd);

-  x10 = WRAPLOW(highbd_dct_const_round_shift(s2 - s10, bd), bd);

-  x11 = WRAPLOW(highbd_dct_const_round_shift(s3 - s11, bd), bd);

-  x12 = WRAPLOW(highbd_dct_const_round_shift(s4 - s12, bd), bd);

-  x13 = WRAPLOW(highbd_dct_const_round_shift(s5 - s13, bd), bd);

-  x14 = WRAPLOW(highbd_dct_const_round_shift(s6 - s14, bd), bd);

-  x15 = WRAPLOW(highbd_dct_const_round_shift(s7 - s15, bd), bd);

+  x0 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s0 + s8), bd);

+  x1 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s1 + s9), bd);

+  x2 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s2 + s10), bd);

+  x3 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s3 + s11), bd);

+  x4 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s4 + s12), bd);

+  x5 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s5 + s13), bd);

+  x6 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s6 + s14), bd);

+  x7 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s7 + s15), bd);

+  x8  = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s0 - s8), bd);

+  x9  = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s1 - s9), bd);

+  x10 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s2 - s10), bd);

+  x11 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s3 - s11), bd);

+  x12 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s4 - s12), bd);

+  x13 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s5 - s13), bd);

+  x14 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s6 - s14), bd);

+  x15 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s7 - s15), bd);

   // stage 2

   s0 = x0;

@@ -1924,22 +1924,22 @@

   s14 = -x14 * cospi_12_64 + x15 * cospi_20_64;

   s15 = x14 * cospi_20_64 + x15 * cospi_12_64;

-  x0 = WRAPLOW(s0 + s4, bd);

-  x1 = WRAPLOW(s1 + s5, bd);

-  x2 = WRAPLOW(s2 + s6, bd);

-  x3 = WRAPLOW(s3 + s7, bd);

-  x4 = WRAPLOW(s0 - s4, bd);

-  x5 = WRAPLOW(s1 - s5, bd);

-  x6 = WRAPLOW(s2 - s6, bd);

-  x7 = WRAPLOW(s3 - s7, bd);

-  x8 = WRAPLOW(highbd_dct_const_round_shift(s8 + s12, bd), bd);

-  x9 = WRAPLOW(highbd_dct_const_round_shift(s9 + s13, bd), bd);

-  x10 = WRAPLOW(highbd_dct_const_round_shift(s10 + s14, bd), bd);

-  x11 = WRAPLOW(highbd_dct_const_round_shift(s11 + s15, bd), bd);

-  x12 = WRAPLOW(highbd_dct_const_round_shift(s8 - s12, bd), bd);

-  x13 = WRAPLOW(highbd_dct_const_round_shift(s9 - s13, bd), bd);

-  x14 = WRAPLOW(highbd_dct_const_round_shift(s10 - s14, bd), bd);

-  x15 = WRAPLOW(highbd_dct_const_round_shift(s11 - s15, bd), bd);

+  x0 = HIGHBD_WRAPLOW(s0 + s4, bd);

+  x1 = HIGHBD_WRAPLOW(s1 + s5, bd);

+  x2 = HIGHBD_WRAPLOW(s2 + s6, bd);

+  x3 = HIGHBD_WRAPLOW(s3 + s7, bd);

+  x4 = HIGHBD_WRAPLOW(s0 - s4, bd);

+  x5 = HIGHBD_WRAPLOW(s1 - s5, bd);

+  x6 = HIGHBD_WRAPLOW(s2 - s6, bd);

+  x7 = HIGHBD_WRAPLOW(s3 - s7, bd);

+  x8 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s8 + s12), bd);

+  x9 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s9 + s13), bd);

+  x10 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s10 + s14), bd);

+  x11 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s11 + s15), bd);

+  x12 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s8 - s12), bd);

+  x13 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s9 - s13), bd);

+  x14 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s10 - s14), bd);

+  x15 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s11 - s15), bd);

   // stage 3

   s0 = x0;

@@ -1959,22 +1959,22 @@

   s14 = -x14 * cospi_24_64 + x15 * cospi_8_64;

   s15 = x14 * cospi_8_64 + x15 * cospi_24_64;

-  x0 = WRAPLOW(s0 + s2, bd);

-  x1 = WRAPLOW(s1 + s3, bd);

-  x2 = WRAPLOW(s0 - s2, bd);

-  x3 = WRAPLOW(s1 - s3, bd);

-  x4 = WRAPLOW(highbd_dct_const_round_shift(s4 + s6, bd), bd);

-  x5 = WRAPLOW(highbd_dct_const_round_shift(s5 + s7, bd), bd);

-  x6 = WRAPLOW(highbd_dct_const_round_shift(s4 - s6, bd), bd);

-  x7 = WRAPLOW(highbd_dct_const_round_shift(s5 - s7, bd), bd);

-  x8 = WRAPLOW(s8 + s10, bd);

-  x9 = WRAPLOW(s9 + s11, bd);

-  x10 = WRAPLOW(s8 - s10, bd);

-  x11 = WRAPLOW(s9 - s11, bd);

-  x12 = WRAPLOW(highbd_dct_const_round_shift(s12 + s14, bd), bd);

-  x13 = WRAPLOW(highbd_dct_const_round_shift(s13 + s15, bd), bd);

-  x14 = WRAPLOW(highbd_dct_const_round_shift(s12 - s14, bd), bd);

-  x15 = WRAPLOW(highbd_dct_const_round_shift(s13 - s15, bd), bd);

+  x0 = HIGHBD_WRAPLOW(s0 + s2, bd);

+  x1 = HIGHBD_WRAPLOW(s1 + s3, bd);

+  x2 = HIGHBD_WRAPLOW(s0 - s2, bd);

+  x3 = HIGHBD_WRAPLOW(s1 - s3, bd);

+  x4 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s4 + s6), bd);

+  x5 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s5 + s7), bd);

+  x6 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s4 - s6), bd);

+  x7 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s5 - s7), bd);

+  x8 = HIGHBD_WRAPLOW(s8 + s10, bd);

+  x9 = HIGHBD_WRAPLOW(s9 + s11, bd);

+  x10 = HIGHBD_WRAPLOW(s8 - s10, bd);

+  x11 = HIGHBD_WRAPLOW(s9 - s11, bd);

+  x12 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s12 + s14), bd);

+  x13 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s13 + s15), bd);

+  x14 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s12 - s14), bd);

+  x15 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s13 - s15), bd);

   // stage 4

   s2 = (- cospi_16_64) * (x2 + x3);

@@ -1986,31 +1986,31 @@

   s14 = (- cospi_16_64) * (x14 + x15);

   s15 = cospi_16_64 * (x14 - x15);

-  x2 = WRAPLOW(highbd_dct_const_round_shift(s2, bd), bd);

-  x3 = WRAPLOW(highbd_dct_const_round_shift(s3, bd), bd);

-  x6 = WRAPLOW(highbd_dct_const_round_shift(s6, bd), bd);

-  x7 = WRAPLOW(highbd_dct_const_round_shift(s7, bd), bd);

-  x10 = WRAPLOW(highbd_dct_const_round_shift(s10, bd), bd);

-  x11 = WRAPLOW(highbd_dct_const_round_shift(s11, bd), bd);

-  x14 = WRAPLOW(highbd_dct_const_round_shift(s14, bd), bd);

-  x15 = WRAPLOW(highbd_dct_const_round_shift(s15, bd), bd);

+  x2 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s2), bd);

+  x3 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s3), bd);

+  x6 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s6), bd);

+  x7 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s7), bd);

+  x10 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s10), bd);

+  x11 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s11), bd);

+  x14 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s14), bd);

+  x15 = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(s15), bd);

-  output[0] = WRAPLOW(x0, bd);

-  output[1] = WRAPLOW(-x8, bd);

-  output[2] = WRAPLOW(x12, bd);

-  output[3] = WRAPLOW(-x4, bd);

-  output[4] = WRAPLOW(x6, bd);

-  output[5] = WRAPLOW(x14, bd);

-  output[6] = WRAPLOW(x10, bd);

-  output[7] = WRAPLOW(x2, bd);

-  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);

+  output[0] = HIGHBD_WRAPLOW(x0, bd);

+  output[1] = HIGHBD_WRAPLOW(-x8, bd);

+  output[2] = HIGHBD_WRAPLOW(x12, bd);

+  output[3] = HIGHBD_WRAPLOW(-x4, bd);

+  output[4] = HIGHBD_WRAPLOW(x6, bd);

+  output[5] = HIGHBD_WRAPLOW(x14, bd);

+  output[6] = HIGHBD_WRAPLOW(x10, bd);

+  output[7] = HIGHBD_WRAPLOW(x2, bd);

+  output[8] = HIGHBD_WRAPLOW(x3, bd);

+  output[9] = HIGHBD_WRAPLOW(x11, bd);

+  output[10] = HIGHBD_WRAPLOW(x15, bd);

+  output[11] = HIGHBD_WRAPLOW(x7, bd);

+  output[12] = HIGHBD_WRAPLOW(x5, bd);

+  output[13] = HIGHBD_WRAPLOW(-x13, bd);

+  output[14] = HIGHBD_WRAPLOW(x9, bd);

+  output[15] = HIGHBD_WRAPLOW(-x1, bd);

 }

 void vpx_highbd_idct16x16_10_add_c(const tran_low_t *input, uint8_t *dest8,

@@ -2045,11 +2045,11 @@

                                   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);

+  tran_low_t out = HIGHBD_WRAPLOW(

+      highbd_dct_const_round_shift(input[0] * cospi_16_64), bd);

   uint16_t *dest = CONVERT_TO_SHORTPTR(dest8);

-  out = WRAPLOW(highbd_dct_const_round_shift(out * cospi_16_64, bd), bd);

+  out = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(out * cospi_16_64), bd);

   a1 = ROUND_POWER_OF_TWO(out, 6);

   for (j = 0; j < 16; ++j) {

     for (i = 0; i < 16; ++i)

@@ -2084,43 +2084,43 @@

   temp1 = input[1] * cospi_31_64 - input[31] * cospi_1_64;

   temp2 = input[1] * cospi_1_64 + input[31] * cospi_31_64;

-  step1[16] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[31] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[16] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[31] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = input[17] * cospi_15_64 - input[15] * cospi_17_64;

   temp2 = input[17] * cospi_17_64 + input[15] * cospi_15_64;

-  step1[17] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[30] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[17] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[30] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = input[9] * cospi_23_64 - input[23] * cospi_9_64;

   temp2 = input[9] * cospi_9_64 + input[23] * cospi_23_64;

-  step1[18] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[29] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[18] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[29] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = input[25] * cospi_7_64 - input[7] * cospi_25_64;

   temp2 = input[25] * cospi_25_64 + input[7] * cospi_7_64;

-  step1[19] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[28] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[19] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[28] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = input[5] * cospi_27_64 - input[27] * cospi_5_64;

   temp2 = input[5] * cospi_5_64 + input[27] * cospi_27_64;

-  step1[20] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[27] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[20] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[27] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = input[21] * cospi_11_64 - input[11] * cospi_21_64;

   temp2 = input[21] * cospi_21_64 + input[11] * cospi_11_64;

-  step1[21] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[26] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[21] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[26] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = input[13] * cospi_19_64 - input[19] * cospi_13_64;

   temp2 = input[13] * cospi_13_64 + input[19] * cospi_19_64;

-  step1[22] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[25] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[22] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[25] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = input[29] * cospi_3_64 - input[3] * cospi_29_64;

   temp2 = input[29] * cospi_29_64 + input[3] * cospi_3_64;

-  step1[23] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[24] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[23] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[24] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   // stage 2

   step2[0] = step1[0];

@@ -2134,40 +2134,40 @@

   temp1 = step1[8] * cospi_30_64 - step1[15] * cospi_2_64;

   temp2 = step1[8] * cospi_2_64 + step1[15] * cospi_30_64;

-  step2[8] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[15] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[8] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[15] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = step1[9] * cospi_14_64 - step1[14] * cospi_18_64;

   temp2 = step1[9] * cospi_18_64 + step1[14] * cospi_14_64;

-  step2[9] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[14] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[9] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[14] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = step1[10] * cospi_22_64 - step1[13] * cospi_10_64;

   temp2 = step1[10] * cospi_10_64 + step1[13] * cospi_22_64;

-  step2[10] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[13] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[10] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[13] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = step1[11] * cospi_6_64 - step1[12] * cospi_26_64;

   temp2 = step1[11] * cospi_26_64 + step1[12] * cospi_6_64;

-  step2[11] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[12] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[11] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[12] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

-  step2[16] = WRAPLOW(step1[16] + step1[17], bd);

-  step2[17] = WRAPLOW(step1[16] - step1[17], bd);

-  step2[18] = WRAPLOW(-step1[18] + step1[19], bd);

-  step2[19] = WRAPLOW(step1[18] + step1[19], bd);

-  step2[20] = WRAPLOW(step1[20] + step1[21], bd);

-  step2[21] = WRAPLOW(step1[20] - step1[21], bd);

-  step2[22] = WRAPLOW(-step1[22] + step1[23], bd);

-  step2[23] = WRAPLOW(step1[22] + step1[23], bd);

-  step2[24] = WRAPLOW(step1[24] + step1[25], bd);

-  step2[25] = WRAPLOW(step1[24] - step1[25], bd);

-  step2[26] = WRAPLOW(-step1[26] + step1[27], bd);

-  step2[27] = WRAPLOW(step1[26] + step1[27], bd);

-  step2[28] = WRAPLOW(step1[28] + step1[29], bd);

-  step2[29] = WRAPLOW(step1[28] - step1[29], bd);

-  step2[30] = WRAPLOW(-step1[30] + step1[31], bd);

-  step2[31] = WRAPLOW(step1[30] + step1[31], bd);

+  step2[16] = HIGHBD_WRAPLOW(step1[16] + step1[17], bd);

+  step2[17] = HIGHBD_WRAPLOW(step1[16] - step1[17], bd);

+  step2[18] = HIGHBD_WRAPLOW(-step1[18] + step1[19], bd);

+  step2[19] = HIGHBD_WRAPLOW(step1[18] + step1[19], bd);

+  step2[20] = HIGHBD_WRAPLOW(step1[20] + step1[21], bd);

+  step2[21] = HIGHBD_WRAPLOW(step1[20] - step1[21], bd);

+  step2[22] = HIGHBD_WRAPLOW(-step1[22] + step1[23], bd);

+  step2[23] = HIGHBD_WRAPLOW(step1[22] + step1[23], bd);

+  step2[24] = HIGHBD_WRAPLOW(step1[24] + step1[25], bd);

+  step2[25] = HIGHBD_WRAPLOW(step1[24] - step1[25], bd);

+  step2[26] = HIGHBD_WRAPLOW(-step1[26] + step1[27], bd);

+  step2[27] = HIGHBD_WRAPLOW(step1[26] + step1[27], bd);

+  step2[28] = HIGHBD_WRAPLOW(step1[28] + step1[29], bd);

+  step2[29] = HIGHBD_WRAPLOW(step1[28] - step1[29], bd);

+  step2[30] = HIGHBD_WRAPLOW(-step1[30] + step1[31], bd);

+  step2[31] = HIGHBD_WRAPLOW(step1[30] + step1[31], bd);

   // stage 3

   step1[0] = step2[0];

@@ -2177,42 +2177,42 @@

   temp1 = step2[4] * cospi_28_64 - step2[7] * cospi_4_64;

   temp2 = step2[4] * cospi_4_64 + step2[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);

+  step1[4] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[7] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = step2[5] * cospi_12_64 - step2[6] * cospi_20_64;

   temp2 = step2[5] * cospi_20_64 + step2[6] * 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);

+  step1[5] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[6] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

-  step1[8] = WRAPLOW(step2[8] + step2[9], bd);

-  step1[9] = WRAPLOW(step2[8] - step2[9], bd);

-  step1[10] = WRAPLOW(-step2[10] + step2[11], bd);

-  step1[11] = WRAPLOW(step2[10] + step2[11], bd);

-  step1[12] = WRAPLOW(step2[12] + step2[13], bd);

-  step1[13] = WRAPLOW(step2[12] - step2[13], bd);

-  step1[14] = WRAPLOW(-step2[14] + step2[15], bd);

-  step1[15] = WRAPLOW(step2[14] + step2[15], bd);

+  step1[8] = HIGHBD_WRAPLOW(step2[8] + step2[9], bd);

+  step1[9] = HIGHBD_WRAPLOW(step2[8] - step2[9], bd);

+  step1[10] = HIGHBD_WRAPLOW(-step2[10] + step2[11], bd);

+  step1[11] = HIGHBD_WRAPLOW(step2[10] + step2[11], bd);

+  step1[12] = HIGHBD_WRAPLOW(step2[12] + step2[13], bd);

+  step1[13] = HIGHBD_WRAPLOW(step2[12] - step2[13], bd);

+  step1[14] = HIGHBD_WRAPLOW(-step2[14] + step2[15], bd);

+  step1[15] = HIGHBD_WRAPLOW(step2[14] + step2[15], bd);

   step1[16] = step2[16];

   step1[31] = step2[31];

   temp1 = -step2[17] * cospi_4_64 + step2[30] * cospi_28_64;

   temp2 = step2[17] * cospi_28_64 + step2[30] * cospi_4_64;

-  step1[17] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[30] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[17] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[30] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = -step2[18] * cospi_28_64 - step2[29] * cospi_4_64;

   temp2 = -step2[18] * cospi_4_64 + step2[29] * cospi_28_64;

-  step1[18] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[29] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[18] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[29] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   step1[19] = step2[19];

   step1[20] = step2[20];

   temp1 = -step2[21] * cospi_20_64 + step2[26] * cospi_12_64;

   temp2 = step2[21] * cospi_12_64 + step2[26] * cospi_20_64;

-  step1[21] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[26] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[21] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[26] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = -step2[22] * cospi_12_64 - step2[25] * cospi_20_64;

   temp2 = -step2[22] * cospi_20_64 + step2[25] * cospi_12_64;

-  step1[22] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[25] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[22] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[25] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   step1[23] = step2[23];

   step1[24] = step2[24];

   step1[27] = step2[27];

@@ -2221,87 +2221,87 @@

   // stage 4

   temp1 = (step1[0] + step1[1]) * cospi_16_64;

   temp2 = (step1[0] - step1[1]) * cospi_16_64;

-  step2[0] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[1] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[0] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[1] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = step1[2] * cospi_24_64 - step1[3] * cospi_8_64;

   temp2 = step1[2] * cospi_8_64 + step1[3] * cospi_24_64;

-  step2[2] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[3] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

-  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);

+  step2[2] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[3] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

+  step2[4] = HIGHBD_WRAPLOW(step1[4] + step1[5], bd);

+  step2[5] = HIGHBD_WRAPLOW(step1[4] - step1[5], bd);

+  step2[6] = HIGHBD_WRAPLOW(-step1[6] + step1[7], bd);

+  step2[7] = HIGHBD_WRAPLOW(step1[6] + step1[7], bd);

   step2[8] = step1[8];

   step2[15] = step1[15];

   temp1 = -step1[9] * cospi_8_64 + step1[14] * cospi_24_64;

   temp2 = step1[9] * cospi_24_64 + step1[14] * cospi_8_64;

-  step2[9] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[14] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[9] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[14] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = -step1[10] * cospi_24_64 - step1[13] * cospi_8_64;

   temp2 = -step1[10] * cospi_8_64 + step1[13] * cospi_24_64;

-  step2[10] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[13] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[10] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[13] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   step2[11] = step1[11];

   step2[12] = step1[12];

-  step2[16] = WRAPLOW(step1[16] + step1[19], bd);

-  step2[17] = WRAPLOW(step1[17] + step1[18], bd);

-  step2[18] = WRAPLOW(step1[17] - step1[18], bd);

-  step2[19] = WRAPLOW(step1[16] - step1[19], bd);

-  step2[20] = WRAPLOW(-step1[20] + step1[23], bd);

-  step2[21] = WRAPLOW(-step1[21] + step1[22], bd);

-  step2[22] = WRAPLOW(step1[21] + step1[22], bd);

-  step2[23] = WRAPLOW(step1[20] + step1[23], bd);

+  step2[16] = HIGHBD_WRAPLOW(step1[16] + step1[19], bd);

+  step2[17] = HIGHBD_WRAPLOW(step1[17] + step1[18], bd);

+  step2[18] = HIGHBD_WRAPLOW(step1[17] - step1[18], bd);

+  step2[19] = HIGHBD_WRAPLOW(step1[16] - step1[19], bd);

+  step2[20] = HIGHBD_WRAPLOW(-step1[20] + step1[23], bd);

+  step2[21] = HIGHBD_WRAPLOW(-step1[21] + step1[22], bd);

+  step2[22] = HIGHBD_WRAPLOW(step1[21] + step1[22], bd);

+  step2[23] = HIGHBD_WRAPLOW(step1[20] + step1[23], bd);

-  step2[24] = WRAPLOW(step1[24] + step1[27], bd);

-  step2[25] = WRAPLOW(step1[25] + step1[26], bd);

-  step2[26] = WRAPLOW(step1[25] - step1[26], bd);

-  step2[27] = WRAPLOW(step1[24] - step1[27], bd);

-  step2[28] = WRAPLOW(-step1[28] + step1[31], bd);

-  step2[29] = WRAPLOW(-step1[29] + step1[30], bd);

-  step2[30] = WRAPLOW(step1[29] + step1[30], bd);

-  step2[31] = WRAPLOW(step1[28] + step1[31], bd);

+  step2[24] = HIGHBD_WRAPLOW(step1[24] + step1[27], bd);

+  step2[25] = HIGHBD_WRAPLOW(step1[25] + step1[26], bd);

+  step2[26] = HIGHBD_WRAPLOW(step1[25] - step1[26], bd);

+  step2[27] = HIGHBD_WRAPLOW(step1[24] - step1[27], bd);

+  step2[28] = HIGHBD_WRAPLOW(-step1[28] + step1[31], bd);

+  step2[29] = HIGHBD_WRAPLOW(-step1[29] + step1[30], bd);

+  step2[30] = HIGHBD_WRAPLOW(step1[29] + step1[30], bd);

+  step2[31] = HIGHBD_WRAPLOW(step1[28] + step1[31], bd);

   // stage 5

-  step1[0] = WRAPLOW(step2[0] + step2[3], bd);

-  step1[1] = WRAPLOW(step2[1] + step2[2], bd);

-  step1[2] = WRAPLOW(step2[1] - step2[2], bd);

-  step1[3] = WRAPLOW(step2[0] - step2[3], bd);

+  step1[0] = HIGHBD_WRAPLOW(step2[0] + step2[3], bd);

+  step1[1] = HIGHBD_WRAPLOW(step2[1] + step2[2], bd);

+  step1[2] = HIGHBD_WRAPLOW(step2[1] - step2[2], bd);

+  step1[3] = HIGHBD_WRAPLOW(step2[0] - step2[3], bd);

   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[5] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[6] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   step1[7] = step2[7];

-  step1[8] = WRAPLOW(step2[8] + step2[11], bd);

-  step1[9] = WRAPLOW(step2[9] + step2[10], bd);

-  step1[10] = WRAPLOW(step2[9] - step2[10], bd);

-  step1[11] = WRAPLOW(step2[8] - step2[11], bd);

-  step1[12] = WRAPLOW(-step2[12] + step2[15], bd);

-  step1[13] = WRAPLOW(-step2[13] + step2[14], bd);

-  step1[14] = WRAPLOW(step2[13] + step2[14], bd);

-  step1[15] = WRAPLOW(step2[12] + step2[15], bd);

+  step1[8] = HIGHBD_WRAPLOW(step2[8] + step2[11], bd);

+  step1[9] = HIGHBD_WRAPLOW(step2[9] + step2[10], bd);

+  step1[10] = HIGHBD_WRAPLOW(step2[9] - step2[10], bd);

+  step1[11] = HIGHBD_WRAPLOW(step2[8] - step2[11], bd);

+  step1[12] = HIGHBD_WRAPLOW(-step2[12] + step2[15], bd);

+  step1[13] = HIGHBD_WRAPLOW(-step2[13] + step2[14], bd);

+  step1[14] = HIGHBD_WRAPLOW(step2[13] + step2[14], bd);

+  step1[15] = HIGHBD_WRAPLOW(step2[12] + step2[15], bd);

   step1[16] = step2[16];

   step1[17] = step2[17];

   temp1 = -step2[18] * cospi_8_64 + step2[29] * cospi_24_64;

   temp2 = step2[18] * cospi_24_64 + step2[29] * cospi_8_64;

-  step1[18] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[29] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[18] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[29] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = -step2[19] * cospi_8_64 + step2[28] * cospi_24_64;

   temp2 = step2[19] * cospi_24_64 + step2[28] * cospi_8_64;

-  step1[19] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[28] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[19] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[28] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = -step2[20] * cospi_24_64 - step2[27] * cospi_8_64;

   temp2 = -step2[20] * cospi_8_64 + step2[27] * cospi_24_64;

-  step1[20] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[27] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[20] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[27] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = -step2[21] * cospi_24_64 - step2[26] * cospi_8_64;

   temp2 = -step2[21] * cospi_8_64 + step2[26] * cospi_24_64;

-  step1[21] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[26] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[21] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[26] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   step1[22] = step2[22];

   step1[23] = step2[23];

   step1[24] = step2[24];

@@ -2310,62 +2310,62 @@

   step1[31] = step2[31];

   // stage 6

-  step2[0] = WRAPLOW(step1[0] + step1[7], bd);

-  step2[1] = WRAPLOW(step1[1] + step1[6], bd);

-  step2[2] = WRAPLOW(step1[2] + step1[5], bd);

-  step2[3] = WRAPLOW(step1[3] + step1[4], bd);

-  step2[4] = WRAPLOW(step1[3] - step1[4], bd);

-  step2[5] = WRAPLOW(step1[2] - step1[5], bd);

-  step2[6] = WRAPLOW(step1[1] - step1[6], bd);

-  step2[7] = WRAPLOW(step1[0] - step1[7], bd);

+  step2[0] = HIGHBD_WRAPLOW(step1[0] + step1[7], bd);

+  step2[1] = HIGHBD_WRAPLOW(step1[1] + step1[6], bd);

+  step2[2] = HIGHBD_WRAPLOW(step1[2] + step1[5], bd);

+  step2[3] = HIGHBD_WRAPLOW(step1[3] + step1[4], bd);

+  step2[4] = HIGHBD_WRAPLOW(step1[3] - step1[4], bd);

+  step2[5] = HIGHBD_WRAPLOW(step1[2] - step1[5], bd);

+  step2[6] = HIGHBD_WRAPLOW(step1[1] - step1[6], bd);

+  step2[7] = HIGHBD_WRAPLOW(step1[0] - step1[7], bd);

   step2[8] = step1[8];

   step2[9] = step1[9];

   temp1 = (-step1[10] + step1[13]) * cospi_16_64;

   temp2 = (step1[10] + step1[13]) * cospi_16_64;

-  step2[10] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[13] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[10] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[13] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = (-step1[11] + step1[12]) * cospi_16_64;

   temp2 = (step1[11] + step1[12]) * cospi_16_64;

-  step2[11] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step2[12] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step2[11] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step2[12] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   step2[14] = step1[14];

   step2[15] = step1[15];

-  step2[16] = WRAPLOW(step1[16] + step1[23], bd);

-  step2[17] = WRAPLOW(step1[17] + step1[22], bd);

-  step2[18] = WRAPLOW(step1[18] + step1[21], bd);

-  step2[19] = WRAPLOW(step1[19] + step1[20], bd);

-  step2[20] = WRAPLOW(step1[19] - step1[20], bd);

-  step2[21] = WRAPLOW(step1[18] - step1[21], bd);

-  step2[22] = WRAPLOW(step1[17] - step1[22], bd);

-  step2[23] = WRAPLOW(step1[16] - step1[23], bd);

+  step2[16] = HIGHBD_WRAPLOW(step1[16] + step1[23], bd);

+  step2[17] = HIGHBD_WRAPLOW(step1[17] + step1[22], bd);

+  step2[18] = HIGHBD_WRAPLOW(step1[18] + step1[21], bd);

+  step2[19] = HIGHBD_WRAPLOW(step1[19] + step1[20], bd);

+  step2[20] = HIGHBD_WRAPLOW(step1[19] - step1[20], bd);

+  step2[21] = HIGHBD_WRAPLOW(step1[18] - step1[21], bd);

+  step2[22] = HIGHBD_WRAPLOW(step1[17] - step1[22], bd);

+  step2[23] = HIGHBD_WRAPLOW(step1[16] - step1[23], bd);

-  step2[24] = WRAPLOW(-step1[24] + step1[31], bd);

-  step2[25] = WRAPLOW(-step1[25] + step1[30], bd);

-  step2[26] = WRAPLOW(-step1[26] + step1[29], bd);

-  step2[27] = WRAPLOW(-step1[27] + step1[28], bd);

-  step2[28] = WRAPLOW(step1[27] + step1[28], bd);

-  step2[29] = WRAPLOW(step1[26] + step1[29], bd);

-  step2[30] = WRAPLOW(step1[25] + step1[30], bd);

-  step2[31] = WRAPLOW(step1[24] + step1[31], bd);

+  step2[24] = HIGHBD_WRAPLOW(-step1[24] + step1[31], bd);

+  step2[25] = HIGHBD_WRAPLOW(-step1[25] + step1[30], bd);

+  step2[26] = HIGHBD_WRAPLOW(-step1[26] + step1[29], bd);

+  step2[27] = HIGHBD_WRAPLOW(-step1[27] + step1[28], bd);

+  step2[28] = HIGHBD_WRAPLOW(step1[27] + step1[28], bd);

+  step2[29] = HIGHBD_WRAPLOW(step1[26] + step1[29], bd);

+  step2[30] = HIGHBD_WRAPLOW(step1[25] + step1[30], bd);

+  step2[31] = HIGHBD_WRAPLOW(step1[24] + step1[31], bd);

   // stage 7

-  step1[0] = WRAPLOW(step2[0] + step2[15], bd);

-  step1[1] = WRAPLOW(step2[1] + step2[14], bd);

-  step1[2] = WRAPLOW(step2[2] + step2[13], bd);

-  step1[3] = WRAPLOW(step2[3] + step2[12], bd);

-  step1[4] = WRAPLOW(step2[4] + step2[11], bd);

-  step1[5] = WRAPLOW(step2[5] + step2[10], bd);

-  step1[6] = WRAPLOW(step2[6] + step2[9], bd);

-  step1[7] = WRAPLOW(step2[7] + step2[8], bd);

-  step1[8] = WRAPLOW(step2[7] - step2[8], bd);

-  step1[9] = WRAPLOW(step2[6] - step2[9], bd);

-  step1[10] = WRAPLOW(step2[5] - step2[10], bd);

-  step1[11] = WRAPLOW(step2[4] - step2[11], bd);

-  step1[12] = WRAPLOW(step2[3] - step2[12], bd);

-  step1[13] = WRAPLOW(step2[2] - step2[13], bd);

-  step1[14] = WRAPLOW(step2[1] - step2[14], bd);

-  step1[15] = WRAPLOW(step2[0] - step2[15], bd);

+  step1[0] = HIGHBD_WRAPLOW(step2[0] + step2[15], bd);

+  step1[1] = HIGHBD_WRAPLOW(step2[1] + step2[14], bd);

+  step1[2] = HIGHBD_WRAPLOW(step2[2] + step2[13], bd);

+  step1[3] = HIGHBD_WRAPLOW(step2[3] + step2[12], bd);

+  step1[4] = HIGHBD_WRAPLOW(step2[4] + step2[11], bd);

+  step1[5] = HIGHBD_WRAPLOW(step2[5] + step2[10], bd);

+  step1[6] = HIGHBD_WRAPLOW(step2[6] + step2[9], bd);

+  step1[7] = HIGHBD_WRAPLOW(step2[7] + step2[8], bd);

+  step1[8] = HIGHBD_WRAPLOW(step2[7] - step2[8], bd);

+  step1[9] = HIGHBD_WRAPLOW(step2[6] - step2[9], bd);

+  step1[10] = HIGHBD_WRAPLOW(step2[5] - step2[10], bd);

+  step1[11] = HIGHBD_WRAPLOW(step2[4] - step2[11], bd);

+  step1[12] = HIGHBD_WRAPLOW(step2[3] - step2[12], bd);

+  step1[13] = HIGHBD_WRAPLOW(step2[2] - step2[13], bd);

+  step1[14] = HIGHBD_WRAPLOW(step2[1] - step2[14], bd);

+  step1[15] = HIGHBD_WRAPLOW(step2[0] - step2[15], bd);

   step1[16] = step2[16];

   step1[17] = step2[17];

@@ -2373,20 +2373,20 @@

   step1[19] = step2[19];

   temp1 = (-step2[20] + step2[27]) * cospi_16_64;

   temp2 = (step2[20] + step2[27]) * cospi_16_64;

-  step1[20] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[27] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[20] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[27] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = (-step2[21] + step2[26]) * cospi_16_64;

   temp2 = (step2[21] + step2[26]) * cospi_16_64;

-  step1[21] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[26] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[21] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[26] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = (-step2[22] + step2[25]) * cospi_16_64;

   temp2 = (step2[22] + step2[25]) * cospi_16_64;

-  step1[22] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[25] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[22] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[25] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   temp1 = (-step2[23] + step2[24]) * cospi_16_64;

   temp2 = (step2[23] + step2[24]) * cospi_16_64;

-  step1[23] = WRAPLOW(highbd_dct_const_round_shift(temp1, bd), bd);

-  step1[24] = WRAPLOW(highbd_dct_const_round_shift(temp2, bd), bd);

+  step1[23] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp1), bd);

+  step1[24] = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(temp2), bd);

   step1[28] = step2[28];

   step1[29] = step2[29];

   step1[30] = step2[30];

@@ -2393,38 +2393,38 @@

   step1[31] = step2[31];

   // final stage

-  output[0] = WRAPLOW(step1[0] + step1[31], bd);

-  output[1] = WRAPLOW(step1[1] + step1[30], bd);

-  output[2] = WRAPLOW(step1[2] + step1[29], bd);

-  output[3] = WRAPLOW(step1[3] + step1[28], bd);

-  output[4] = WRAPLOW(step1[4] + step1[27], bd);

-  output[5] = WRAPLOW(step1[5] + step1[26], bd);

-  output[6] = WRAPLOW(step1[6] + step1[25], bd);

-  output[7] = WRAPLOW(step1[7] + step1[24], bd);

-  output[8] = WRAPLOW(step1[8] + step1[23], bd);

-  output[9] = WRAPLOW(step1[9] + step1[22], bd);

-  output[10] = WRAPLOW(step1[10] + step1[21], bd);

-  output[11] = WRAPLOW(step1[11] + step1[20], bd);

-  output[12] = WRAPLOW(step1[12] + step1[19], bd);

-  output[13] = WRAPLOW(step1[13] + step1[18], bd);

-  output[14] = WRAPLOW(step1[14] + step1[17], bd);

-  output[15] = WRAPLOW(step1[15] + step1[16], bd);

-  output[16] = WRAPLOW(step1[15] - step1[16], bd);

-  output[17] = WRAPLOW(step1[14] - step1[17], bd);

-  output[18] = WRAPLOW(step1[13] - step1[18], bd);

-  output[19] = WRAPLOW(step1[12] - step1[19], bd);

-  output[20] = WRAPLOW(step1[11] - step1[20], bd);

-  output[21] = WRAPLOW(step1[10] - step1[21], bd);

-  output[22] = WRAPLOW(step1[9] - step1[22], bd);

-  output[23] = WRAPLOW(step1[8] - step1[23], bd);

-  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);

+  output[0] = HIGHBD_WRAPLOW(step1[0] + step1[31], bd);

+  output[1] = HIGHBD_WRAPLOW(step1[1] + step1[30], bd);

+  output[2] = HIGHBD_WRAPLOW(step1[2] + step1[29], bd);

+  output[3] = HIGHBD_WRAPLOW(step1[3] + step1[28], bd);

+  output[4] = HIGHBD_WRAPLOW(step1[4] + step1[27], bd);

+  output[5] = HIGHBD_WRAPLOW(step1[5] + step1[26], bd);

+  output[6] = HIGHBD_WRAPLOW(step1[6] + step1[25], bd);

+  output[7] = HIGHBD_WRAPLOW(step1[7] + step1[24], bd);

+  output[8] = HIGHBD_WRAPLOW(step1[8] + step1[23], bd);

+  output[9] = HIGHBD_WRAPLOW(step1[9] + step1[22], bd);

+  output[10] = HIGHBD_WRAPLOW(step1[10] + step1[21], bd);

+  output[11] = HIGHBD_WRAPLOW(step1[11] + step1[20], bd);

+  output[12] = HIGHBD_WRAPLOW(step1[12] + step1[19], bd);

+  output[13] = HIGHBD_WRAPLOW(step1[13] + step1[18], bd);

+  output[14] = HIGHBD_WRAPLOW(step1[14] + step1[17], bd);

+  output[15] = HIGHBD_WRAPLOW(step1[15] + step1[16], bd);

+  output[16] = HIGHBD_WRAPLOW(step1[15] - step1[16], bd);

+  output[17] = HIGHBD_WRAPLOW(step1[14] - step1[17], bd);

+  output[18] = HIGHBD_WRAPLOW(step1[13] - step1[18], bd);

+  output[19] = HIGHBD_WRAPLOW(step1[12] - step1[19], bd);

+  output[20] = HIGHBD_WRAPLOW(step1[11] - step1[20], bd);

+  output[21] = HIGHBD_WRAPLOW(step1[10] - step1[21], bd);

+  output[22] = HIGHBD_WRAPLOW(step1[9] - step1[22], bd);

+  output[23] = HIGHBD_WRAPLOW(step1[8] - step1[23], bd);

+  output[24] = HIGHBD_WRAPLOW(step1[7] - step1[24], bd);

+  output[25] = HIGHBD_WRAPLOW(step1[6] - step1[25], bd);

+  output[26] = HIGHBD_WRAPLOW(step1[5] - step1[26], bd);

+  output[27] = HIGHBD_WRAPLOW(step1[4] - step1[27], bd);

+  output[28] = HIGHBD_WRAPLOW(step1[3] - step1[28], bd);

+  output[29] = HIGHBD_WRAPLOW(step1[2] - step1[29], bd);

+  output[30] = HIGHBD_WRAPLOW(step1[1] - step1[30], bd);

+  output[31] = HIGHBD_WRAPLOW(step1[0] - step1[31], bd);

 }

 void vpx_highbd_idct32x32_1024_add_c(const tran_low_t *input, uint8_t *dest8,

@@ -2500,9 +2500,9 @@

   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);

+  tran_low_t out = HIGHBD_WRAPLOW(

+      highbd_dct_const_round_shift(input[0] * cospi_16_64), bd);

+  out = HIGHBD_WRAPLOW(highbd_dct_const_round_shift(out * cospi_16_64), bd);

   a1 = ROUND_POWER_OF_TWO(out, 6);

   for (j = 0; j < 32; ++j) {

--- a/vpx_dsp/inv_txfm.h

+++ b/vpx_dsp/inv_txfm.h

@@ -21,7 +21,7 @@

 extern "C" {

 #endif

-static INLINE tran_low_t check_range(tran_high_t input) {

+static INLINE tran_high_t check_range(tran_high_t input) {

 #if CONFIG_COEFFICIENT_RANGE_CHECKING

   // For valid VP9 input streams, intermediate stage coefficients should always

   // stay within the range of a signed 16 bit integer. Coefficients can go out

@@ -32,17 +32,17 @@

   assert(INT16_MIN <= input);

   assert(input <= INT16_MAX);

 #endif  // CONFIG_COEFFICIENT_RANGE_CHECKING

-  return (tran_low_t)input;

+  return input;

 }

-static INLINE tran_low_t dct_const_round_shift(tran_high_t input) {

+static INLINE tran_high_t dct_const_round_shift(tran_high_t input) {

   tran_high_t rv = ROUND_POWER_OF_TWO(input, DCT_CONST_BITS);

-  return check_range(rv);

+  return (tran_high_t)rv;

 }

 #if CONFIG_VP9_HIGHBITDEPTH

-static INLINE tran_low_t highbd_check_range(tran_high_t input,

-                                            int bd) {

+static INLINE tran_high_t highbd_check_range(tran_high_t input,

+                                             int bd) {

 #if CONFIG_COEFFICIENT_RANGE_CHECKING

   // For valid highbitdepth VP9 streams, intermediate stage coefficients will

   // stay within the ranges:

@@ -56,13 +56,12 @@

   (void) int_min;

 #endif  // CONFIG_COEFFICIENT_RANGE_CHECKING

   (void) bd;

-  return (tran_low_t)input;

+  return input;

 }

-static INLINE tran_low_t highbd_dct_const_round_shift(tran_high_t input,

-                                                      int bd) {

+static INLINE tran_high_t highbd_dct_const_round_shift(tran_high_t input) {

   tran_high_t rv = ROUND_POWER_OF_TWO(input, DCT_CONST_BITS);

-  return highbd_check_range(rv, bd);

+  return (tran_high_t)rv;

 }

 #endif  // CONFIG_VP9_HIGHBITDEPTH

@@ -83,9 +82,20 @@

 // bd of 10 uses trans_low with 18bits, need to remove 14bits

 // bd of 12 uses trans_low with 20bits, need to remove 12bits

 // bd of x uses trans_low with 8+x bits, need to remove 24-x bits

-#define WRAPLOW(x, bd) ((((int32_t)(x)) << (24 - bd)) >> (24 - bd))

-#else

-#define WRAPLOW(x, bd) ((int32_t)(x))

+

+#define WRAPLOW(x) ((((int32_t)check_range(x)) << 16) >> 16)

+#if CONFIG_VP9_HIGHBITDEPTH

+#define HIGHBD_WRAPLOW(x, bd) \

+    ((((int32_t)highbd_check_range((x), bd)) << (24 - bd)) >> (24 - bd))

+#endif  // CONFIG_VP9_HIGHBITDEPTH

+

+#else   // CONFIG_EMULATE_HARDWARE

+

+#define WRAPLOW(x) ((int32_t)check_range(x))

+#if CONFIG_VP9_HIGHBITDEPTH

+#define HIGHBD_WRAPLOW(x, bd) \

+    ((int32_t)highbd_check_range((x), bd))

+#endif  // CONFIG_VP9_HIGHBITDEPTH

 #endif  // CONFIG_EMULATE_HARDWARE

 void idct4_c(const tran_low_t *input, tran_low_t *output);

@@ -107,14 +117,14 @@

 static INLINE uint16_t highbd_clip_pixel_add(uint16_t dest, tran_high_t trans,

                                              int bd) {

-  trans = WRAPLOW(trans, bd);

-  return clip_pixel_highbd(WRAPLOW(dest + trans, bd), bd);

+  trans = HIGHBD_WRAPLOW(trans, bd);

+  return clip_pixel_highbd(dest + trans, bd);

 }

 #endif

 static INLINE uint8_t clip_pixel_add(uint8_t dest, tran_high_t trans) {

-  trans = WRAPLOW(trans, 8);

-  return clip_pixel(WRAPLOW(dest + trans, 8));

+  trans = WRAPLOW(trans);

+  return clip_pixel(dest + trans);

 }

 #ifdef __cplusplus

 }  // extern "C"