/*
 * Copyright (c) 2021, Alliance for Open Media. All rights reserved
 *
 * This source code is subject to the terms of the BSD 3-Clause Clear License
 * and the Alliance for Open Media Patent License 1.0. If the BSD 3-Clause Clear
 * License was not distributed with this source code in the LICENSE file, you
 * can obtain it at aomedia.org/license/software-license/bsd-3-c-c/.  If the
 * Alliance for Open Media Patent License 1.0 was not distributed with this
 * source code in the PATENTS file, you can obtain it at
 * aomedia.org/license/patent-license/.
 */
#include <assert.h>
#include <smmintrin.h> /* SSE4.1 */

#include "config/aom_config.h"
#include "config/av1_rtcd.h"

#include "av1/common/av1_inv_txfm1d_cfg.h"
#include "av1/common/idct.h"
#include "av1/common/x86/av1_inv_txfm_ssse3.h"
#include "av1/common/x86/av1_txfm_sse2.h"
#include "av1/common/x86/av1_txfm_sse4.h"
#include "av1/common/x86/highbd_txfm_utility_sse4.h"

static INLINE __m128i highbd_clamp_epi16(__m128i u, int bd) {
  const __m128i zero = _mm_setzero_si128();
  const __m128i one = _mm_set1_epi16(1);
  const __m128i max = _mm_sub_epi16(_mm_slli_epi16(one, bd), one);
  __m128i clamped, mask;

  mask = _mm_cmpgt_epi16(u, max);
  clamped = _mm_andnot_si128(mask, u);
  mask = _mm_and_si128(mask, max);
  clamped = _mm_or_si128(mask, clamped);
  mask = _mm_cmpgt_epi16(clamped, zero);
  clamped = _mm_and_si128(clamped, mask);

  return clamped;
}

static INLINE void round_shift_4x4(__m128i *in, int shift) {
  if (shift != 0) {
    __m128i rnding = _mm_set1_epi32(1 << (shift - 1));
    in[0] = _mm_add_epi32(in[0], rnding);
    in[1] = _mm_add_epi32(in[1], rnding);
    in[2] = _mm_add_epi32(in[2], rnding);
    in[3] = _mm_add_epi32(in[3], rnding);

    in[0] = _mm_srai_epi32(in[0], shift);
    in[1] = _mm_srai_epi32(in[1], shift);
    in[2] = _mm_srai_epi32(in[2], shift);
    in[3] = _mm_srai_epi32(in[3], shift);
  }
}

static void round_shift_8x8(__m128i *in, int shift) {
  round_shift_4x4(&in[0], shift);
  round_shift_4x4(&in[4], shift);
  round_shift_4x4(&in[8], shift);
  round_shift_4x4(&in[12], shift);
}

static void highbd_clamp_epi32_sse4_1(__m128i *in, __m128i *out,
                                      const __m128i *clamp_lo,
                                      const __m128i *clamp_hi, int size) {
  __m128i a0, a1;
  for (int i = 0; i < size; i += 4) {
    a0 = _mm_max_epi32(in[i], *clamp_lo);
    out[i] = _mm_min_epi32(a0, *clamp_hi);

    a1 = _mm_max_epi32(in[i + 1], *clamp_lo);
    out[i + 1] = _mm_min_epi32(a1, *clamp_hi);

    a0 = _mm_max_epi32(in[i + 2], *clamp_lo);
    out[i + 2] = _mm_min_epi32(a0, *clamp_hi);

    a1 = _mm_max_epi32(in[i + 3], *clamp_lo);
    out[i + 3] = _mm_min_epi32(a1, *clamp_hi);
  }
}

static INLINE __m128i highbd_get_recon_8x8_sse4_1(const __m128i pred,
                                                  __m128i res0, __m128i res1,
                                                  const int bd) {
  __m128i x0 = _mm_cvtepi16_epi32(pred);
  __m128i x1 = _mm_cvtepi16_epi32(_mm_srli_si128(pred, 8));
  __m128i min_clip_val = _mm_setzero_si128();
  __m128i max_clip_val = _mm_set1_epi32((1 << bd) - 1);
  x0 = _mm_add_epi32(res0, x0);
  x1 = _mm_add_epi32(res1, x1);
  x0 = _mm_max_epi32(x0, min_clip_val);
  x0 = _mm_min_epi32(x0, max_clip_val);
  x1 = _mm_max_epi32(x1, min_clip_val);
  x1 = _mm_min_epi32(x1, max_clip_val);
  x0 = _mm_packus_epi32(x0, x1);
  return x0;
}

static INLINE __m128i highbd_get_recon_4xn_sse4_1(const __m128i pred,
                                                  __m128i res0, const int bd) {
  __m128i x0 = _mm_cvtepi16_epi32(pred);
  __m128i min_clip_val = _mm_setzero_si128();
  __m128i max_clip_val = _mm_set1_epi32((1 << bd) - 1);
  x0 = _mm_add_epi32(res0, x0);
  x0 = _mm_max_epi32(x0, min_clip_val);
  x0 = _mm_min_epi32(x0, max_clip_val);
  x0 = _mm_packus_epi32(x0, x0);
  return x0;
}

static INLINE void highbd_write_buffer_4xn_sse4_1(__m128i *in, uint16_t *output,
                                                  int stride, int flipud,
                                                  int height, const int bd) {
  int j = flipud ? (height - 1) : 0;
  const int step = flipud ? -1 : 1;
  for (int i = 0; i < height; ++i, j += step) {
    __m128i v = _mm_loadl_epi64((__m128i const *)(output + i * stride));
    __m128i u = highbd_get_recon_4xn_sse4_1(v, in[j], bd);

    _mm_storel_epi64((__m128i *)(output + i * stride), u);
  }
}

static INLINE void highbd_write_buffer_8xn_sse4_1(__m128i *in, uint16_t *output,
                                                  int stride, int flipud,
                                                  int height, const int bd) {
  int j = flipud ? (height - 1) : 0;
  const int step = flipud ? -1 : 1;
  for (int i = 0; i < height; ++i, j += step) {
    __m128i v = _mm_loadu_si128((__m128i const *)(output + i * stride));
    __m128i u = highbd_get_recon_8x8_sse4_1(v, in[j], in[j + height], bd);

    _mm_storeu_si128((__m128i *)(output + i * stride), u);
  }
}

static INLINE void load_buffer_32bit_input(const int32_t *in, int stride,
                                           __m128i *out, int out_size) {
  for (int i = 0; i < out_size; ++i) {
    out[i] = _mm_loadu_si128((const __m128i *)(in + i * stride));
  }
}

static INLINE void load_buffer_4x4(const int32_t *coeff, __m128i *in) {
  in[0] = _mm_loadu_si128((const __m128i *)(coeff + 0));
  in[1] = _mm_loadu_si128((const __m128i *)(coeff + 4));
  in[2] = _mm_loadu_si128((const __m128i *)(coeff + 8));
  in[3] = _mm_loadu_si128((const __m128i *)(coeff + 12));
}

static void addsub_sse4_1(const __m128i in0, const __m128i in1, __m128i *out0,
                          __m128i *out1, const __m128i *clamp_lo,
                          const __m128i *clamp_hi) {
  __m128i a0 = _mm_add_epi32(in0, in1);
  __m128i a1 = _mm_sub_epi32(in0, in1);

  a0 = _mm_max_epi32(a0, *clamp_lo);
  a0 = _mm_min_epi32(a0, *clamp_hi);
  a1 = _mm_max_epi32(a1, *clamp_lo);
  a1 = _mm_min_epi32(a1, *clamp_hi);

  *out0 = a0;
  *out1 = a1;
}

static void shift_and_clamp_sse4_1(__m128i *in0, __m128i *in1,
                                   const __m128i *clamp_lo,
                                   const __m128i *clamp_hi, int shift) {
  __m128i offset = _mm_set1_epi32((1 << shift) >> 1);
  __m128i in0_w_offset = _mm_add_epi32(*in0, offset);
  __m128i in1_w_offset = _mm_add_epi32(*in1, offset);

  in0_w_offset = _mm_sra_epi32(in0_w_offset, _mm_cvtsi32_si128(shift));
  in1_w_offset = _mm_sra_epi32(in1_w_offset, _mm_cvtsi32_si128(shift));

  in0_w_offset = _mm_max_epi32(in0_w_offset, *clamp_lo);
  in0_w_offset = _mm_min_epi32(in0_w_offset, *clamp_hi);
  in1_w_offset = _mm_max_epi32(in1_w_offset, *clamp_lo);
  in1_w_offset = _mm_min_epi32(in1_w_offset, *clamp_hi);

  *in0 = in0_w_offset;
  *in1 = in1_w_offset;
}

static INLINE void idct32_stage4_sse4_1(
    __m128i *bf1, const __m128i *cospim8, const __m128i *cospi56,
    const __m128i *cospi8, const __m128i *cospim56, const __m128i *cospim40,
    const __m128i *cospi24, const __m128i *cospi40, const __m128i *cospim24,
    const __m128i *rounding, int bit) {
  __m128i temp1, temp2;
  temp1 = half_btf_sse4_1(cospim8, &bf1[17], cospi56, &bf1[30], rounding, bit);
  bf1[30] = half_btf_sse4_1(cospi56, &bf1[17], cospi8, &bf1[30], rounding, bit);
  bf1[17] = temp1;

  temp2 = half_btf_sse4_1(cospim56, &bf1[18], cospim8, &bf1[29], rounding, bit);
  bf1[29] =
      half_btf_sse4_1(cospim8, &bf1[18], cospi56, &bf1[29], rounding, bit);
  bf1[18] = temp2;

  temp1 = half_btf_sse4_1(cospim40, &bf1[21], cospi24, &bf1[26], rounding, bit);
  bf1[26] =
      half_btf_sse4_1(cospi24, &bf1[21], cospi40, &bf1[26], rounding, bit);
  bf1[21] = temp1;

  temp2 =
      half_btf_sse4_1(cospim24, &bf1[22], cospim40, &bf1[25], rounding, bit);
  bf1[25] =
      half_btf_sse4_1(cospim40, &bf1[22], cospi24, &bf1[25], rounding, bit);
  bf1[22] = temp2;
}

static INLINE void idct32_stage5_sse4_1(
    __m128i *bf1, const __m128i *cospim16, const __m128i *cospi48,
    const __m128i *cospi16, const __m128i *cospim48, const __m128i *clamp_lo,
    const __m128i *clamp_hi, const __m128i *rounding, int bit) {
  __m128i temp1, temp2;
  temp1 = half_btf_sse4_1(cospim16, &bf1[9], cospi48, &bf1[14], rounding, bit);
  bf1[14] = half_btf_sse4_1(cospi48, &bf1[9], cospi16, &bf1[14], rounding, bit);
  bf1[9] = temp1;

  temp2 =
      half_btf_sse4_1(cospim48, &bf1[10], cospim16, &bf1[13], rounding, bit);
  bf1[13] =
      half_btf_sse4_1(cospim16, &bf1[10], cospi48, &bf1[13], rounding, bit);
  bf1[10] = temp2;

  addsub_sse4_1(bf1[16], bf1[19], bf1 + 16, bf1 + 19, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[17], bf1[18], bf1 + 17, bf1 + 18, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[23], bf1[20], bf1 + 23, bf1 + 20, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[22], bf1[21], bf1 + 22, bf1 + 21, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[24], bf1[27], bf1 + 24, bf1 + 27, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[25], bf1[26], bf1 + 25, bf1 + 26, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[31], bf1[28], bf1 + 31, bf1 + 28, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[30], bf1[29], bf1 + 30, bf1 + 29, clamp_lo, clamp_hi);
}

static INLINE void idct32_stage6_sse4_1(
    __m128i *bf1, const __m128i *cospim32, const __m128i *cospi32,
    const __m128i *cospim16, const __m128i *cospi48, const __m128i *cospi16,
    const __m128i *cospim48, const __m128i *clamp_lo, const __m128i *clamp_hi,
    const __m128i *rounding, int bit) {
  __m128i temp1, temp2;
  temp1 = half_btf_sse4_1(cospim32, &bf1[5], cospi32, &bf1[6], rounding, bit);
  bf1[6] = half_btf_sse4_1(cospi32, &bf1[5], cospi32, &bf1[6], rounding, bit);
  bf1[5] = temp1;

  addsub_sse4_1(bf1[8], bf1[11], bf1 + 8, bf1 + 11, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[9], bf1[10], bf1 + 9, bf1 + 10, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[15], bf1[12], bf1 + 15, bf1 + 12, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[14], bf1[13], bf1 + 14, bf1 + 13, clamp_lo, clamp_hi);

  temp1 = half_btf_sse4_1(cospim16, &bf1[18], cospi48, &bf1[29], rounding, bit);
  bf1[29] =
      half_btf_sse4_1(cospi48, &bf1[18], cospi16, &bf1[29], rounding, bit);
  bf1[18] = temp1;
  temp2 = half_btf_sse4_1(cospim16, &bf1[19], cospi48, &bf1[28], rounding, bit);
  bf1[28] =
      half_btf_sse4_1(cospi48, &bf1[19], cospi16, &bf1[28], rounding, bit);
  bf1[19] = temp2;
  temp1 =
      half_btf_sse4_1(cospim48, &bf1[20], cospim16, &bf1[27], rounding, bit);
  bf1[27] =
      half_btf_sse4_1(cospim16, &bf1[20], cospi48, &bf1[27], rounding, bit);
  bf1[20] = temp1;
  temp2 =
      half_btf_sse4_1(cospim48, &bf1[21], cospim16, &bf1[26], rounding, bit);
  bf1[26] =
      half_btf_sse4_1(cospim16, &bf1[21], cospi48, &bf1[26], rounding, bit);
  bf1[21] = temp2;
}

static INLINE void idct32_stage7_sse4_1(__m128i *bf1, const __m128i *cospim32,
                                        const __m128i *cospi32,
                                        const __m128i *clamp_lo,
                                        const __m128i *clamp_hi,
                                        const __m128i *rounding, int bit) {
  __m128i temp1, temp2;
  addsub_sse4_1(bf1[0], bf1[7], bf1 + 0, bf1 + 7, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[1], bf1[6], bf1 + 1, bf1 + 6, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[2], bf1[5], bf1 + 2, bf1 + 5, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[3], bf1[4], bf1 + 3, bf1 + 4, clamp_lo, clamp_hi);

  temp1 = half_btf_sse4_1(cospim32, &bf1[10], cospi32, &bf1[13], rounding, bit);
  bf1[13] =
      half_btf_sse4_1(cospi32, &bf1[10], cospi32, &bf1[13], rounding, bit);
  bf1[10] = temp1;
  temp2 = half_btf_sse4_1(cospim32, &bf1[11], cospi32, &bf1[12], rounding, bit);
  bf1[12] =
      half_btf_sse4_1(cospi32, &bf1[11], cospi32, &bf1[12], rounding, bit);
  bf1[11] = temp2;

  addsub_sse4_1(bf1[16], bf1[23], bf1 + 16, bf1 + 23, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[17], bf1[22], bf1 + 17, bf1 + 22, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[18], bf1[21], bf1 + 18, bf1 + 21, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[19], bf1[20], bf1 + 19, bf1 + 20, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[31], bf1[24], bf1 + 31, bf1 + 24, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[30], bf1[25], bf1 + 30, bf1 + 25, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[29], bf1[26], bf1 + 29, bf1 + 26, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[28], bf1[27], bf1 + 28, bf1 + 27, clamp_lo, clamp_hi);
}

static INLINE void idct32_stage8_sse4_1(__m128i *bf1, const __m128i *cospim32,
                                        const __m128i *cospi32,
                                        const __m128i *clamp_lo,
                                        const __m128i *clamp_hi,
                                        const __m128i *rounding, int bit) {
  __m128i temp1, temp2;
  addsub_sse4_1(bf1[0], bf1[15], bf1 + 0, bf1 + 15, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[1], bf1[14], bf1 + 1, bf1 + 14, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[2], bf1[13], bf1 + 2, bf1 + 13, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[3], bf1[12], bf1 + 3, bf1 + 12, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[4], bf1[11], bf1 + 4, bf1 + 11, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[5], bf1[10], bf1 + 5, bf1 + 10, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[6], bf1[9], bf1 + 6, bf1 + 9, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[7], bf1[8], bf1 + 7, bf1 + 8, clamp_lo, clamp_hi);

  temp1 = half_btf_sse4_1(cospim32, &bf1[20], cospi32, &bf1[27], rounding, bit);
  bf1[27] =
      half_btf_sse4_1(cospi32, &bf1[20], cospi32, &bf1[27], rounding, bit);
  bf1[20] = temp1;
  temp2 = half_btf_sse4_1(cospim32, &bf1[21], cospi32, &bf1[26], rounding, bit);
  bf1[26] =
      half_btf_sse4_1(cospi32, &bf1[21], cospi32, &bf1[26], rounding, bit);
  bf1[21] = temp2;
  temp1 = half_btf_sse4_1(cospim32, &bf1[22], cospi32, &bf1[25], rounding, bit);
  bf1[25] =
      half_btf_sse4_1(cospi32, &bf1[22], cospi32, &bf1[25], rounding, bit);
  bf1[22] = temp1;
  temp2 = half_btf_sse4_1(cospim32, &bf1[23], cospi32, &bf1[24], rounding, bit);
  bf1[24] =
      half_btf_sse4_1(cospi32, &bf1[23], cospi32, &bf1[24], rounding, bit);
  bf1[23] = temp2;
}

static INLINE void idct32_stage9_sse4_1(__m128i *bf1, __m128i *out,
                                        const int do_cols, const int bd,
                                        const int out_shift,
                                        const __m128i *clamp_lo,
                                        const __m128i *clamp_hi) {
  addsub_sse4_1(bf1[0], bf1[31], out + 0, out + 31, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[1], bf1[30], out + 1, out + 30, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[2], bf1[29], out + 2, out + 29, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[3], bf1[28], out + 3, out + 28, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[4], bf1[27], out + 4, out + 27, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[5], bf1[26], out + 5, out + 26, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[6], bf1[25], out + 6, out + 25, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[7], bf1[24], out + 7, out + 24, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[8], bf1[23], out + 8, out + 23, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[9], bf1[22], out + 9, out + 22, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[10], bf1[21], out + 10, out + 21, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[11], bf1[20], out + 11, out + 20, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[12], bf1[19], out + 12, out + 19, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[13], bf1[18], out + 13, out + 18, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[14], bf1[17], out + 14, out + 17, clamp_lo, clamp_hi);
  addsub_sse4_1(bf1[15], bf1[16], out + 15, out + 16, clamp_lo, clamp_hi);

  if (!do_cols) {
    const int log_range_out = AOMMAX(16, bd + 6);
    assert(log_range_out - 1 < 32);  // Ensure left-shift doesn't overflow.
    const __m128i clamp_lo_out = _mm_set1_epi32(-(1 << (log_range_out - 1)));
    const __m128i clamp_hi_out = _mm_set1_epi32((1 << (log_range_out - 1)) - 1);
    for (int i = 0; i < 32; i += 8) {
      round_shift_4x4(out + i, out_shift);
      round_shift_4x4(out + i + 4, out_shift);
    }
    highbd_clamp_epi32_sse4_1(out, out, &clamp_lo_out, &clamp_hi_out, 32);
  }
}

#if !USE_TUNED_ADST16
static void neg_shift_sse4_1(const __m128i in0, const __m128i in1,
                             __m128i *out0, __m128i *out1,
                             const __m128i *clamp_lo, const __m128i *clamp_hi,
                             int shift) {
  __m128i offset = _mm_set1_epi32((1 << shift) >> 1);
  __m128i a0 = _mm_add_epi32(offset, in0);
  __m128i a1 = _mm_sub_epi32(offset, in1);

  a0 = _mm_sra_epi32(a0, _mm_cvtsi32_si128(shift));
  a1 = _mm_sra_epi32(a1, _mm_cvtsi32_si128(shift));

  a0 = _mm_max_epi32(a0, *clamp_lo);
  a0 = _mm_min_epi32(a0, *clamp_hi);
  a1 = _mm_max_epi32(a1, *clamp_lo);
  a1 = _mm_min_epi32(a1, *clamp_hi);

  *out0 = a0;
  *out1 = a1;
}
#endif  // !USE_TUNED_ADST16

static void idct4x4_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                           int bd, int out_shift) {
  const int32_t *cospi = cospi_arr(bit);
  const __m128i cospi32 = _mm_set1_epi32(cospi[32]);
  const __m128i cospi48 = _mm_set1_epi32(cospi[48]);
  const __m128i cospi16 = _mm_set1_epi32(cospi[16]);
  const __m128i cospim16 = _mm_set1_epi32(-cospi[16]);
  const __m128i rnding = _mm_set1_epi32(1 << (bit - 1));
  int log_range = AOMMAX(16, bd + (do_cols ? 6 : 8));
  __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
  __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
  __m128i u0, u1, u2, u3;
  __m128i v0, v1, v2, v3, x, y;

  // Stage 0
  // Stage 1
  // Stage 2
  v0 = _mm_unpacklo_epi32(in[0], in[1]);
  v1 = _mm_unpackhi_epi32(in[0], in[1]);
  v2 = _mm_unpacklo_epi32(in[2], in[3]);
  v3 = _mm_unpackhi_epi32(in[2], in[3]);

  u0 = _mm_unpacklo_epi64(v0, v2);
  u1 = _mm_unpackhi_epi64(v0, v2);
  u2 = _mm_unpacklo_epi64(v1, v3);
  u3 = _mm_unpackhi_epi64(v1, v3);

  x = _mm_mullo_epi32(u0, cospi32);
  y = _mm_mullo_epi32(u2, cospi32);
  v0 = _mm_add_epi32(x, y);
  v0 = _mm_add_epi32(v0, rnding);
  v0 = _mm_srai_epi32(v0, bit);

  v1 = _mm_sub_epi32(x, y);
  v1 = _mm_add_epi32(v1, rnding);
  v1 = _mm_srai_epi32(v1, bit);

  x = _mm_mullo_epi32(u1, cospi48);
  y = _mm_mullo_epi32(u3, cospim16);
  v2 = _mm_add_epi32(x, y);
  v2 = _mm_add_epi32(v2, rnding);
  v2 = _mm_srai_epi32(v2, bit);

  x = _mm_mullo_epi32(u1, cospi16);
  y = _mm_mullo_epi32(u3, cospi48);
  v3 = _mm_add_epi32(x, y);
  v3 = _mm_add_epi32(v3, rnding);
  v3 = _mm_srai_epi32(v3, bit);

  // Stage 3
  addsub_sse4_1(v0, v3, out + 0, out + 3, &clamp_lo, &clamp_hi);
  addsub_sse4_1(v1, v2, out + 1, out + 2, &clamp_lo, &clamp_hi);

  if (!do_cols) {
    log_range = AOMMAX(16, bd + 6);
    assert(log_range - 1 < 32);  // Ensure left-shift doesn't overflow.
    clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
    clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);

    shift_and_clamp_sse4_1(out + 0, out + 3, &clamp_lo, &clamp_hi, out_shift);
    shift_and_clamp_sse4_1(out + 1, out + 2, &clamp_lo, &clamp_hi, out_shift);
  }
}

static void iddt4x4_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                           int bd, int out_shift) {
  (void)bit;
  __m128i tmp[4];
  TRANSPOSE_4X4(in[0], in[1], in[2], in[3], tmp[0], tmp[1], tmp[2], tmp[3]);
  iadst_matrix_mult_sse4(tmp, out, INV_DDT_BIT, do_cols, bd, out_shift,
                         ddt4_kernel[INV_TXFM], 4, 4, 1);
}

#if USE_TUNED_ADST4
static void iadst4x4_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                            int bd, int out_shift) {
  const int32_t *cospi = cospi_arr(bit);
  const __m128i rnding = _mm_set1_epi32(1 << (bit - 1));
  const __m128i cospi8 = _mm_set1_epi32((int)cospi[8]);
  const __m128i cospi24 = _mm_set1_epi32((int)cospi[24]);
  const __m128i cospi32 = _mm_set1_epi32((int)cospi[32]);
  const __m128i cospi40 = _mm_set1_epi32((int)cospi[40]);
  const __m128i cospi56 = _mm_set1_epi32((int)cospi[56]);
  int log_range = AOMMAX(16, bd + (do_cols ? 6 : 8));
  __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
  __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);

  __m128i s0, s1, s2, s3;
  __m128i x0, x1, x2, x3;
  __m128i u0, u1, u2, u3;
  __m128i v0, v1, v2, v3;

  // stage 0 transpose
  v0 = _mm_unpacklo_epi32(in[0], in[1]);
  v1 = _mm_unpackhi_epi32(in[0], in[1]);
  v2 = _mm_unpacklo_epi32(in[2], in[3]);
  v3 = _mm_unpackhi_epi32(in[2], in[3]);

  u0 = _mm_unpacklo_epi64(v0, v2);
  u1 = _mm_unpackhi_epi64(v0, v2);
  u2 = _mm_unpacklo_epi64(v1, v3);
  u3 = _mm_unpackhi_epi64(v1, v3);

  // stage 1
  x0 = u0;
  x1 = _mm_sub_epi32(_mm_setzero_si128(), u3);
  x2 = _mm_sub_epi32(_mm_setzero_si128(), u1);
  x3 = u2;

  // stage 2
  s0 = x0;
  s1 = x1;
  s2 = half_btf_sse4_1(&cospi32, &x2, &cospi32, &x3, &rnding, bit);
  s3 = half_btf_neg_sse4_1(&cospi32, &x2, &cospi32, &x3, &rnding, bit);

  // stage 3
  addsub_sse4_1(s0, s2, &x0, &x2, &clamp_lo, &clamp_hi);
  addsub_sse4_1(s1, s3, &x1, &x3, &clamp_lo, &clamp_hi);

  // stage 4
  s0 = half_btf_sse4_1(&cospi8, &x0, &cospi56, &x1, &rnding, bit);
  s1 = half_btf_neg_sse4_1(&cospi56, &x0, &cospi8, &x1, &rnding, bit);
  s2 = half_btf_sse4_1(&cospi40, &x2, &cospi24, &x3, &rnding, bit);
  s3 = half_btf_neg_sse4_1(&cospi24, &x2, &cospi40, &x3, &rnding, bit);

  // stage 5
  out[0] = s1;
  out[1] = s2;
  out[2] = s3;
  out[3] = s0;

  if (!do_cols) {
    log_range = AOMMAX(16, bd + 6);
    assert(log_range - 1 < 32);  // Ensure left-shift doesn't overflow.
    clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
    clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
    round_shift_4x4(out, out_shift);
    highbd_clamp_epi32_sse4_1(out, out, &clamp_lo, &clamp_hi, 4);
  }
}
#else
static void iadst4x4_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                            int bd, int out_shift) {
  const int32_t *sinpi = sinpi_arr(bit);
  const __m128i zero = _mm_set1_epi32(0);
  __m128i rnding = _mm_set1_epi32(1 << (bit + 4 - 1));
  rnding = _mm_unpacklo_epi32(rnding, zero);
  const __m128i mul = _mm_set1_epi32(1 << 4);
  const __m128i sinpi1 = _mm_set1_epi32((int)sinpi[1]);
  const __m128i sinpi2 = _mm_set1_epi32((int)sinpi[2]);
  const __m128i sinpi3 = _mm_set1_epi32((int)sinpi[3]);
  const __m128i sinpi4 = _mm_set1_epi32((int)sinpi[4]);
  __m128i t;
  __m128i s0, s1, s2, s3, s4, s5, s6, s7;
  __m128i x0, x1, x2, x3;
  __m128i u0, u1, u2, u3;
  __m128i v0, v1, v2, v3;
  __m128i u0_low, u1_low, u2_low, u3_low;
  __m128i u0_high, u1_high, u2_high, u3_high;

  v0 = _mm_unpacklo_epi32(in[0], in[1]);
  v1 = _mm_unpackhi_epi32(in[0], in[1]);
  v2 = _mm_unpacklo_epi32(in[2], in[3]);
  v3 = _mm_unpackhi_epi32(in[2], in[3]);

  x0 = _mm_unpacklo_epi64(v0, v2);
  x1 = _mm_unpackhi_epi64(v0, v2);
  x2 = _mm_unpacklo_epi64(v1, v3);
  x3 = _mm_unpackhi_epi64(v1, v3);

  s0 = _mm_mullo_epi32(x0, sinpi1);
  s1 = _mm_mullo_epi32(x0, sinpi2);
  s2 = _mm_mullo_epi32(x1, sinpi3);
  s3 = _mm_mullo_epi32(x2, sinpi4);
  s4 = _mm_mullo_epi32(x2, sinpi1);
  s5 = _mm_mullo_epi32(x3, sinpi2);
  s6 = _mm_mullo_epi32(x3, sinpi4);
  t = _mm_sub_epi32(x0, x2);
  s7 = _mm_add_epi32(t, x3);

  t = _mm_add_epi32(s0, s3);
  s0 = _mm_add_epi32(t, s5);
  t = _mm_sub_epi32(s1, s4);
  s1 = _mm_sub_epi32(t, s6);
  s3 = s2;
  s2 = _mm_mullo_epi32(s7, sinpi3);

  u0 = _mm_add_epi32(s0, s3);
  u1 = _mm_add_epi32(s1, s3);
  u2 = s2;
  t = _mm_add_epi32(s0, s1);
  u3 = _mm_sub_epi32(t, s3);

  // u0
  u0_low = _mm_mul_epi32(u0, mul);
  u0_low = _mm_add_epi64(u0_low, rnding);

  u0 = _mm_srli_si128(u0, 4);
  u0_high = _mm_mul_epi32(u0, mul);
  u0_high = _mm_add_epi64(u0_high, rnding);

  u0_low = _mm_srli_si128(u0_low, 2);
  u0_high = _mm_srli_si128(u0_high, 2);

  u0 = _mm_unpacklo_epi32(u0_low, u0_high);
  u0_high = _mm_unpackhi_epi32(u0_low, u0_high);
  u0 = _mm_unpacklo_epi64(u0, u0_high);

  // u1
  u1_low = _mm_mul_epi32(u1, mul);
  u1_low = _mm_add_epi64(u1_low, rnding);

  u1 = _mm_srli_si128(u1, 4);
  u1_high = _mm_mul_epi32(u1, mul);
  u1_high = _mm_add_epi64(u1_high, rnding);

  u1_low = _mm_srli_si128(u1_low, 2);
  u1_high = _mm_srli_si128(u1_high, 2);

  u1 = _mm_unpacklo_epi32(u1_low, u1_high);
  u1_high = _mm_unpackhi_epi32(u1_low, u1_high);
  u1 = _mm_unpacklo_epi64(u1, u1_high);

  // u2
  u2_low = _mm_mul_epi32(u2, mul);
  u2_low = _mm_add_epi64(u2_low, rnding);

  u2 = _mm_srli_si128(u2, 4);
  u2_high = _mm_mul_epi32(u2, mul);
  u2_high = _mm_add_epi64(u2_high, rnding);

  u2_low = _mm_srli_si128(u2_low, 2);
  u2_high = _mm_srli_si128(u2_high, 2);

  u2 = _mm_unpacklo_epi32(u2_low, u2_high);
  u2_high = _mm_unpackhi_epi32(u2_low, u2_high);
  u2 = _mm_unpacklo_epi64(u2, u2_high);

  // u3
  u3_low = _mm_mul_epi32(u3, mul);
  u3_low = _mm_add_epi64(u3_low, rnding);

  u3 = _mm_srli_si128(u3, 4);
  u3_high = _mm_mul_epi32(u3, mul);
  u3_high = _mm_add_epi64(u3_high, rnding);

  u3_low = _mm_srli_si128(u3_low, 2);
  u3_high = _mm_srli_si128(u3_high, 2);

  u3 = _mm_unpacklo_epi32(u3_low, u3_high);
  u3_high = _mm_unpackhi_epi32(u3_low, u3_high);
  u3 = _mm_unpacklo_epi64(u3, u3_high);

  out[0] = u0;
  out[1] = u1;
  out[2] = u2;
  out[3] = u3;

  if (!do_cols) {
    const int log_range = AOMMAX(16, bd + 6);
    const __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
    const __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
    round_shift_4x4(out, out_shift);
    highbd_clamp_epi32_sse4_1(out, out, &clamp_lo, &clamp_hi, 4);
  }
}
#endif  // USE_TUNED_ADST4

static void write_buffer_4x4(__m128i *in, uint16_t *output, int stride,
                             int fliplr, int flipud, int shift, int bd) {
  const __m128i zero = _mm_setzero_si128();
  __m128i u0, u1, u2, u3;
  __m128i v0, v1, v2, v3;

  round_shift_4x4(in, shift);

  v0 = _mm_loadl_epi64((__m128i const *)(output + 0 * stride));
  v1 = _mm_loadl_epi64((__m128i const *)(output + 1 * stride));
  v2 = _mm_loadl_epi64((__m128i const *)(output + 2 * stride));
  v3 = _mm_loadl_epi64((__m128i const *)(output + 3 * stride));

  v0 = _mm_unpacklo_epi16(v0, zero);
  v1 = _mm_unpacklo_epi16(v1, zero);
  v2 = _mm_unpacklo_epi16(v2, zero);
  v3 = _mm_unpacklo_epi16(v3, zero);

  if (fliplr) {
    in[0] = _mm_shuffle_epi32(in[0], 0x1B);
    in[1] = _mm_shuffle_epi32(in[1], 0x1B);
    in[2] = _mm_shuffle_epi32(in[2], 0x1B);
    in[3] = _mm_shuffle_epi32(in[3], 0x1B);
  }

  if (flipud) {
    u0 = _mm_add_epi32(in[3], v0);
    u1 = _mm_add_epi32(in[2], v1);
    u2 = _mm_add_epi32(in[1], v2);
    u3 = _mm_add_epi32(in[0], v3);
  } else {
    u0 = _mm_add_epi32(in[0], v0);
    u1 = _mm_add_epi32(in[1], v1);
    u2 = _mm_add_epi32(in[2], v2);
    u3 = _mm_add_epi32(in[3], v3);
  }

  v0 = _mm_packus_epi32(u0, u1);
  v2 = _mm_packus_epi32(u2, u3);

  u0 = highbd_clamp_epi16(v0, bd);
  u2 = highbd_clamp_epi16(v2, bd);

  v0 = _mm_unpacklo_epi64(u0, u0);
  v1 = _mm_unpackhi_epi64(u0, u0);
  v2 = _mm_unpacklo_epi64(u2, u2);
  v3 = _mm_unpackhi_epi64(u2, u2);

  _mm_storel_epi64((__m128i *)(output + 0 * stride), v0);
  _mm_storel_epi64((__m128i *)(output + 1 * stride), v1);
  _mm_storel_epi64((__m128i *)(output + 2 * stride), v2);
  _mm_storel_epi64((__m128i *)(output + 3 * stride), v3);
}

static void iidentity4_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                              int bd, int out_shift) {
  (void)bit;
  __m128i v[4];
  __m128i zero = _mm_set1_epi32(0);
  __m128i fact = _mm_set1_epi32(NewSqrt2);
  __m128i offset = _mm_set1_epi32(1 << (NewSqrt2Bits - 1));
  __m128i a0_low, a1_low;
  __m128i a0_high, a1_high;

  offset = _mm_unpacklo_epi32(offset, zero);

  for (int i = 0; i < 4; i++) {
    a0_low = _mm_mul_epi32(in[i], fact);
    a0_low = _mm_add_epi32(a0_low, offset);
    a0_low = _mm_srli_epi64(a0_low, NewSqrt2Bits);

    a0_high = _mm_srli_si128(in[i], 4);
    a0_high = _mm_mul_epi32(a0_high, fact);
    a0_high = _mm_add_epi32(a0_high, offset);
    a0_high = _mm_srli_epi64(a0_high, NewSqrt2Bits);

    a1_low = _mm_unpacklo_epi32(a0_low, a0_high);
    a1_high = _mm_unpackhi_epi32(a0_low, a0_high);
    out[i] = _mm_unpacklo_epi64(a1_low, a1_high);
  }

  if (!do_cols) {
    const int log_range = AOMMAX(16, bd + 6);
    const __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
    const __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
    round_shift_4x4(out, out_shift);
    highbd_clamp_epi32_sse4_1(out, out, &clamp_lo, &clamp_hi, 4);
  }

  // Transpose for 4x4
  v[0] = _mm_unpacklo_epi32(out[0], out[1]);
  v[1] = _mm_unpackhi_epi32(out[0], out[1]);
  v[2] = _mm_unpacklo_epi32(out[2], out[3]);
  v[3] = _mm_unpackhi_epi32(out[2], out[3]);

  out[0] = _mm_unpacklo_epi64(v[0], v[2]);
  out[1] = _mm_unpackhi_epi64(v[0], v[2]);
  out[2] = _mm_unpacklo_epi64(v[1], v[3]);
  out[3] = _mm_unpackhi_epi64(v[1], v[3]);
}
void av1_inv_txfm2d_add_4x4_sse4_1(const int32_t *input, uint16_t *output,
                                   int stride, TX_TYPE tx_type, int use_ddt,
                                   int bd) {
  __m128i in[4];
  const int8_t *shift = av1_inv_txfm_shift_ls[TX_4X4];
  const int txw_idx = get_txw_idx(TX_4X4);
  const int txh_idx = get_txh_idx(TX_4X4);

  switch (tx_type) {
    case DCT_DCT:
      load_buffer_4x4(input, in);
      idct4x4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd, 0);
      idct4x4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);
      write_buffer_4x4(in, output, stride, 0, 0, -shift[1], bd);
      break;
    case ADST_DCT:
      load_buffer_4x4(input, in);
      idct4x4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd, 0);
      if (use_ddt && REPLACE_ADST4)
        iddt4x4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);
      else
        iadst4x4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                        0);
      write_buffer_4x4(in, output, stride, 0, 0, -shift[1], bd);
      break;
    case DCT_ADST:
      load_buffer_4x4(input, in);
      if (use_ddt && REPLACE_ADST4)
        iddt4x4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd, 0);
      else
        iadst4x4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                        0);
      idct4x4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);
      write_buffer_4x4(in, output, stride, 0, 0, -shift[1], bd);
      break;
    case ADST_ADST:
      load_buffer_4x4(input, in);
      if (use_ddt && REPLACE_ADST4) {
        iddt4x4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd, 0);
        iddt4x4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);
      } else {
        iadst4x4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                        0);
        iadst4x4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                        0);
      }
      write_buffer_4x4(in, output, stride, 0, 0, -shift[1], bd);
      break;
    case FLIPADST_DCT:
      load_buffer_4x4(input, in);
      idct4x4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd, 0);
      if (use_ddt && REPLACE_ADST4)
        iddt4x4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);
      else
        iadst4x4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                        0);
      write_buffer_4x4(in, output, stride, 0, 1, -shift[1], bd);
      break;
    case DCT_FLIPADST:
      load_buffer_4x4(input, in);
      if (use_ddt && REPLACE_ADST4)
        iddt4x4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd, 0);
      else
        iadst4x4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                        0);
      idct4x4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);
      write_buffer_4x4(in, output, stride, 1, 0, -shift[1], bd);
      break;
    case FLIPADST_FLIPADST:
      load_buffer_4x4(input, in);
      if (use_ddt && REPLACE_ADST4) {
        iddt4x4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd, 0);
        iddt4x4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);
      } else {
        iadst4x4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                        0);
        iadst4x4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                        0);
      }
      write_buffer_4x4(in, output, stride, 1, 1, -shift[1], bd);
      break;
    case ADST_FLIPADST:
      load_buffer_4x4(input, in);
      if (use_ddt && REPLACE_ADST4) {
        iddt4x4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd, 0);
        iddt4x4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);
      } else {
        iadst4x4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                        0);
        iadst4x4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                        0);
      }
      write_buffer_4x4(in, output, stride, 1, 0, -shift[1], bd);
      break;
    case FLIPADST_ADST:
      load_buffer_4x4(input, in);
      if (use_ddt && REPLACE_ADST4) {
        iddt4x4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd, 0);
        iddt4x4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);
      } else {
        iadst4x4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                        0);
        iadst4x4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                        0);
      }
      write_buffer_4x4(in, output, stride, 0, 1, -shift[1], bd);
      break;
    case IDTX:
      load_buffer_4x4(input, in);
      iidentity4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                        0);
      iidentity4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                        0);
      write_buffer_4x4(in, output, stride, 0, 0, -shift[1], bd);
      break;
    case V_DCT:
      load_buffer_4x4(input, in);
      iidentity4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                        0);
      idct4x4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);
      write_buffer_4x4(in, output, stride, 0, 0, -shift[1], bd);
      break;
    case H_DCT:
      load_buffer_4x4(input, in);
      idct4x4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd, 0);
      iidentity4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                        0);
      write_buffer_4x4(in, output, stride, 0, 0, -shift[1], bd);
      break;
    case V_ADST:
      load_buffer_4x4(input, in);
      iidentity4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                        0);
      if (use_ddt && REPLACE_ADST4)
        iddt4x4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);
      else
        iadst4x4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                        0);
      write_buffer_4x4(in, output, stride, 0, 0, -shift[1], bd);
      break;
    case H_ADST:
      load_buffer_4x4(input, in);
      if (use_ddt && REPLACE_ADST4)
        iddt4x4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd, 0);
      else
        iadst4x4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                        0);
      iidentity4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                        0);
      write_buffer_4x4(in, output, stride, 0, 0, -shift[1], bd);
      break;
    case V_FLIPADST:
      load_buffer_4x4(input, in);
      iidentity4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                        0);
      if (use_ddt && REPLACE_ADST4)
        iddt4x4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);
      else
        iadst4x4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                        0);
      write_buffer_4x4(in, output, stride, 0, 1, -shift[1], bd);
      break;
    case H_FLIPADST:
      load_buffer_4x4(input, in);
      if (use_ddt && REPLACE_ADST4)
        iddt4x4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd, 0);
      else
        iadst4x4_sse4_1(in, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                        0);
      iidentity4_sse4_1(in, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                        0);
      write_buffer_4x4(in, output, stride, 1, 0, -shift[1], bd);
      break;
    default: assert(0);
  }
}

// 8x8
static void load_buffer_8x8(const int32_t *coeff, __m128i *in) {
  in[0] = _mm_loadu_si128((const __m128i *)(coeff + 0));
  in[1] = _mm_loadu_si128((const __m128i *)(coeff + 4));
  in[2] = _mm_loadu_si128((const __m128i *)(coeff + 8));
  in[3] = _mm_loadu_si128((const __m128i *)(coeff + 12));
  in[4] = _mm_loadu_si128((const __m128i *)(coeff + 16));
  in[5] = _mm_loadu_si128((const __m128i *)(coeff + 20));
  in[6] = _mm_loadu_si128((const __m128i *)(coeff + 24));
  in[7] = _mm_loadu_si128((const __m128i *)(coeff + 28));
  in[8] = _mm_loadu_si128((const __m128i *)(coeff + 32));
  in[9] = _mm_loadu_si128((const __m128i *)(coeff + 36));
  in[10] = _mm_loadu_si128((const __m128i *)(coeff + 40));
  in[11] = _mm_loadu_si128((const __m128i *)(coeff + 44));
  in[12] = _mm_loadu_si128((const __m128i *)(coeff + 48));
  in[13] = _mm_loadu_si128((const __m128i *)(coeff + 52));
  in[14] = _mm_loadu_si128((const __m128i *)(coeff + 56));
  in[15] = _mm_loadu_si128((const __m128i *)(coeff + 60));
}

static void idct8x8_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                           int bd, int out_shift) {
  const int32_t *cospi = cospi_arr(bit);
  const __m128i cospi56 = _mm_set1_epi32(cospi[56]);
  const __m128i cospim8 = _mm_set1_epi32(-cospi[8]);
  const __m128i cospi24 = _mm_set1_epi32(cospi[24]);
  const __m128i cospim40 = _mm_set1_epi32(-cospi[40]);
  const __m128i cospi40 = _mm_set1_epi32(cospi[40]);
  const __m128i cospi8 = _mm_set1_epi32(cospi[8]);
  const __m128i cospi32 = _mm_set1_epi32(cospi[32]);
  const __m128i cospi48 = _mm_set1_epi32(cospi[48]);
  const __m128i cospim16 = _mm_set1_epi32(-cospi[16]);
  const __m128i cospi16 = _mm_set1_epi32(cospi[16]);
  const __m128i rnding = _mm_set1_epi32(1 << (bit - 1));
  const int log_range = AOMMAX(16, bd + (do_cols ? 6 : 8));
  const __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
  const __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
  __m128i u0, u1, u2, u3, u4, u5, u6, u7;
  __m128i v0, v1, v2, v3, v4, v5, v6, v7;
  __m128i x, y;
  int col;

  // Note:
  //  Even column: 0, 2, ..., 14
  //  Odd column: 1, 3, ..., 15
  //  one even column plus one odd column constructs one row (8 coeffs)
  //  total we have 8 rows (8x8).
  for (col = 0; col < 2; ++col) {
    // stage 0
    // stage 1
    // stage 2
    u0 = in[0 * 2 + col];
    u1 = in[4 * 2 + col];
    u2 = in[2 * 2 + col];
    u3 = in[6 * 2 + col];

    x = _mm_mullo_epi32(in[1 * 2 + col], cospi56);
    y = _mm_mullo_epi32(in[7 * 2 + col], cospim8);
    u4 = _mm_add_epi32(x, y);
    u4 = _mm_add_epi32(u4, rnding);
    u4 = _mm_srai_epi32(u4, bit);

    x = _mm_mullo_epi32(in[1 * 2 + col], cospi8);
    y = _mm_mullo_epi32(in[7 * 2 + col], cospi56);
    u7 = _mm_add_epi32(x, y);
    u7 = _mm_add_epi32(u7, rnding);
    u7 = _mm_srai_epi32(u7, bit);

    x = _mm_mullo_epi32(in[5 * 2 + col], cospi24);
    y = _mm_mullo_epi32(in[3 * 2 + col], cospim40);
    u5 = _mm_add_epi32(x, y);
    u5 = _mm_add_epi32(u5, rnding);
    u5 = _mm_srai_epi32(u5, bit);

    x = _mm_mullo_epi32(in[5 * 2 + col], cospi40);
    y = _mm_mullo_epi32(in[3 * 2 + col], cospi24);
    u6 = _mm_add_epi32(x, y);
    u6 = _mm_add_epi32(u6, rnding);
    u6 = _mm_srai_epi32(u6, bit);

    // stage 3
    x = _mm_mullo_epi32(u0, cospi32);
    y = _mm_mullo_epi32(u1, cospi32);
    v0 = _mm_add_epi32(x, y);
    v0 = _mm_add_epi32(v0, rnding);
    v0 = _mm_srai_epi32(v0, bit);

    v1 = _mm_sub_epi32(x, y);
    v1 = _mm_add_epi32(v1, rnding);
    v1 = _mm_srai_epi32(v1, bit);

    x = _mm_mullo_epi32(u2, cospi48);
    y = _mm_mullo_epi32(u3, cospim16);
    v2 = _mm_add_epi32(x, y);
    v2 = _mm_add_epi32(v2, rnding);
    v2 = _mm_srai_epi32(v2, bit);

    x = _mm_mullo_epi32(u2, cospi16);
    y = _mm_mullo_epi32(u3, cospi48);
    v3 = _mm_add_epi32(x, y);
    v3 = _mm_add_epi32(v3, rnding);
    v3 = _mm_srai_epi32(v3, bit);

    addsub_sse4_1(u4, u5, &v4, &v5, &clamp_lo, &clamp_hi);
    addsub_sse4_1(u7, u6, &v7, &v6, &clamp_lo, &clamp_hi);

    // stage 4
    addsub_sse4_1(v0, v3, &u0, &u3, &clamp_lo, &clamp_hi);
    addsub_sse4_1(v1, v2, &u1, &u2, &clamp_lo, &clamp_hi);
    u4 = v4;
    u7 = v7;

    x = _mm_mullo_epi32(v5, cospi32);
    y = _mm_mullo_epi32(v6, cospi32);
    u6 = _mm_add_epi32(y, x);
    u6 = _mm_add_epi32(u6, rnding);
    u6 = _mm_srai_epi32(u6, bit);

    u5 = _mm_sub_epi32(y, x);
    u5 = _mm_add_epi32(u5, rnding);
    u5 = _mm_srai_epi32(u5, bit);

    // stage 5
    addsub_sse4_1(u0, u7, out + 0 * 2 + col, out + 7 * 2 + col, &clamp_lo,
                  &clamp_hi);
    addsub_sse4_1(u1, u6, out + 1 * 2 + col, out + 6 * 2 + col, &clamp_lo,
                  &clamp_hi);
    addsub_sse4_1(u2, u5, out + 2 * 2 + col, out + 5 * 2 + col, &clamp_lo,
                  &clamp_hi);
    addsub_sse4_1(u3, u4, out + 3 * 2 + col, out + 4 * 2 + col, &clamp_lo,
                  &clamp_hi);
  }

  if (!do_cols) {
    const int log_range_out = AOMMAX(16, bd + 6);
    assert(log_range_out - 1 < 32);  // Ensure left-shift doesn't overflow.
    const __m128i clamp_lo_out = _mm_set1_epi32(-(1 << (log_range_out - 1)));
    const __m128i clamp_hi_out = _mm_set1_epi32((1 << (log_range_out - 1)) - 1);
    round_shift_8x8(out, out_shift);
    highbd_clamp_epi32_sse4_1(out, out, &clamp_lo_out, &clamp_hi_out, 16);
  }
}

static void iddt8x8_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                           int bd, int out_shift) {
  (void)bit;
  iadst_matrix_mult_sse4(in, out, INV_DDT_BIT, do_cols, bd, out_shift,
                         ddt8_kernel[INV_TXFM], 8, 8, 2);
  iadst_matrix_mult_sse4(in + 1, out + 1, INV_DDT_BIT, do_cols, bd, out_shift,
                         ddt8_kernel[INV_TXFM], 8, 8, 2);
}

#if USE_TUNED_ADST8
static void iadst8x8_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                            int bd, int out_shift) {
  (void)bit;
  iadst_matrix_mult_sse4(in, out, INV_ADST_BIT, do_cols, bd, out_shift,
                         av2_adst_kernel8[INV_TXFM], tx_size_wide[TX_8X8],
                         tx_size_wide[TX_8X8], 2);
  iadst_matrix_mult_sse4(in + 1, out + 1, INV_ADST_BIT, do_cols, bd, out_shift,
                         av2_adst_kernel8[INV_TXFM], tx_size_wide[TX_8X8],
                         tx_size_wide[TX_8X8], 2);
}
#else
static void iadst8x8_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                            int bd, int out_shift) {
  const int32_t *cospi = cospi_arr(bit);
  const __m128i cospi4 = _mm_set1_epi32(cospi[4]);
  const __m128i cospi60 = _mm_set1_epi32(cospi[60]);
  const __m128i cospi20 = _mm_set1_epi32(cospi[20]);
  const __m128i cospi44 = _mm_set1_epi32(cospi[44]);
  const __m128i cospi36 = _mm_set1_epi32(cospi[36]);
  const __m128i cospi28 = _mm_set1_epi32(cospi[28]);
  const __m128i cospi52 = _mm_set1_epi32(cospi[52]);
  const __m128i cospi12 = _mm_set1_epi32(cospi[12]);
  const __m128i cospi16 = _mm_set1_epi32(cospi[16]);
  const __m128i cospi48 = _mm_set1_epi32(cospi[48]);
  const __m128i cospim48 = _mm_set1_epi32(-cospi[48]);
  const __m128i cospi32 = _mm_set1_epi32(cospi[32]);
  const __m128i rnding = _mm_set1_epi32(1 << (bit - 1));
  const __m128i kZero = _mm_setzero_si128();
  const int log_range = AOMMAX(16, bd + (do_cols ? 6 : 8));
  const __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
  const __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
  __m128i u[8], v[8], x;

  // Even 8 points: 0, 2, ..., 14
  // stage 0
  // stage 1
  // stage 2
  // (1)
  u[0] = _mm_mullo_epi32(in[14], cospi4);
  x = _mm_mullo_epi32(in[0], cospi60);
  u[0] = _mm_add_epi32(u[0], x);
  u[0] = _mm_add_epi32(u[0], rnding);
  u[0] = _mm_srai_epi32(u[0], bit);

  u[1] = _mm_mullo_epi32(in[14], cospi60);
  x = _mm_mullo_epi32(in[0], cospi4);
  u[1] = _mm_sub_epi32(u[1], x);
  u[1] = _mm_add_epi32(u[1], rnding);
  u[1] = _mm_srai_epi32(u[1], bit);

  // (2)
  u[2] = _mm_mullo_epi32(in[10], cospi20);
  x = _mm_mullo_epi32(in[4], cospi44);
  u[2] = _mm_add_epi32(u[2], x);
  u[2] = _mm_add_epi32(u[2], rnding);
  u[2] = _mm_srai_epi32(u[2], bit);

  u[3] = _mm_mullo_epi32(in[10], cospi44);
  x = _mm_mullo_epi32(in[4], cospi20);
  u[3] = _mm_sub_epi32(u[3], x);
  u[3] = _mm_add_epi32(u[3], rnding);
  u[3] = _mm_srai_epi32(u[3], bit);

  // (3)
  u[4] = _mm_mullo_epi32(in[6], cospi36);
  x = _mm_mullo_epi32(in[8], cospi28);
  u[4] = _mm_add_epi32(u[4], x);
  u[4] = _mm_add_epi32(u[4], rnding);
  u[4] = _mm_srai_epi32(u[4], bit);

  u[5] = _mm_mullo_epi32(in[6], cospi28);
  x = _mm_mullo_epi32(in[8], cospi36);
  u[5] = _mm_sub_epi32(u[5], x);
  u[5] = _mm_add_epi32(u[5], rnding);
  u[5] = _mm_srai_epi32(u[5], bit);

  // (4)
  u[6] = _mm_mullo_epi32(in[2], cospi52);
  x = _mm_mullo_epi32(in[12], cospi12);
  u[6] = _mm_add_epi32(u[6], x);
  u[6] = _mm_add_epi32(u[6], rnding);
  u[6] = _mm_srai_epi32(u[6], bit);

  u[7] = _mm_mullo_epi32(in[2], cospi12);
  x = _mm_mullo_epi32(in[12], cospi52);
  u[7] = _mm_sub_epi32(u[7], x);
  u[7] = _mm_add_epi32(u[7], rnding);
  u[7] = _mm_srai_epi32(u[7], bit);

  // stage 3
  addsub_sse4_1(u[0], u[4], &v[0], &v[4], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[1], u[5], &v[1], &v[5], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[2], u[6], &v[2], &v[6], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[3], u[7], &v[3], &v[7], &clamp_lo, &clamp_hi);

  // stage 4
  u[0] = v[0];
  u[1] = v[1];
  u[2] = v[2];
  u[3] = v[3];

  u[4] = _mm_mullo_epi32(v[4], cospi16);
  x = _mm_mullo_epi32(v[5], cospi48);
  u[4] = _mm_add_epi32(u[4], x);
  u[4] = _mm_add_epi32(u[4], rnding);
  u[4] = _mm_srai_epi32(u[4], bit);

  u[5] = _mm_mullo_epi32(v[4], cospi48);
  x = _mm_mullo_epi32(v[5], cospi16);
  u[5] = _mm_sub_epi32(u[5], x);
  u[5] = _mm_add_epi32(u[5], rnding);
  u[5] = _mm_srai_epi32(u[5], bit);

  u[6] = _mm_mullo_epi32(v[6], cospim48);
  x = _mm_mullo_epi32(v[7], cospi16);
  u[6] = _mm_add_epi32(u[6], x);
  u[6] = _mm_add_epi32(u[6], rnding);
  u[6] = _mm_srai_epi32(u[6], bit);

  u[7] = _mm_mullo_epi32(v[6], cospi16);
  x = _mm_mullo_epi32(v[7], cospim48);
  u[7] = _mm_sub_epi32(u[7], x);
  u[7] = _mm_add_epi32(u[7], rnding);
  u[7] = _mm_srai_epi32(u[7], bit);

  // stage 5
  addsub_sse4_1(u[0], u[2], &v[0], &v[2], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[1], u[3], &v[1], &v[3], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[4], u[6], &v[4], &v[6], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[5], u[7], &v[5], &v[7], &clamp_lo, &clamp_hi);

  // stage 6
  u[0] = v[0];
  u[1] = v[1];
  u[4] = v[4];
  u[5] = v[5];

  v[0] = _mm_mullo_epi32(v[2], cospi32);
  x = _mm_mullo_epi32(v[3], cospi32);
  u[2] = _mm_add_epi32(v[0], x);
  u[2] = _mm_add_epi32(u[2], rnding);
  u[2] = _mm_srai_epi32(u[2], bit);

  u[3] = _mm_sub_epi32(v[0], x);
  u[3] = _mm_add_epi32(u[3], rnding);
  u[3] = _mm_srai_epi32(u[3], bit);

  v[0] = _mm_mullo_epi32(v[6], cospi32);
  x = _mm_mullo_epi32(v[7], cospi32);
  u[6] = _mm_add_epi32(v[0], x);
  u[6] = _mm_add_epi32(u[6], rnding);
  u[6] = _mm_srai_epi32(u[6], bit);

  u[7] = _mm_sub_epi32(v[0], x);
  u[7] = _mm_add_epi32(u[7], rnding);
  u[7] = _mm_srai_epi32(u[7], bit);

  // stage 7
  if (do_cols) {
    out[0] = u[0];
    out[2] = _mm_sub_epi32(kZero, u[4]);
    out[4] = u[6];
    out[6] = _mm_sub_epi32(kZero, u[2]);
    out[8] = u[3];
    out[10] = _mm_sub_epi32(kZero, u[7]);
    out[12] = u[5];
    out[14] = _mm_sub_epi32(kZero, u[1]);
  } else {
    const int log_range_out = AOMMAX(16, bd + 6);
    const __m128i clamp_lo_out = _mm_set1_epi32(-(1 << (log_range_out - 1)));
    const __m128i clamp_hi_out = _mm_set1_epi32((1 << (log_range_out - 1)) - 1);

    neg_shift_sse4_1(u[0], u[4], out + 0, out + 2, &clamp_lo_out, &clamp_hi_out,
                     out_shift);
    neg_shift_sse4_1(u[6], u[2], out + 4, out + 6, &clamp_lo_out, &clamp_hi_out,
                     out_shift);
    neg_shift_sse4_1(u[3], u[7], out + 8, out + 10, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
    neg_shift_sse4_1(u[5], u[1], out + 12, out + 14, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
  }

  // Odd 8 points: 1, 3, ..., 15
  // stage 0
  // stage 1
  // stage 2
  // (1)
  u[0] = _mm_mullo_epi32(in[15], cospi4);
  x = _mm_mullo_epi32(in[1], cospi60);
  u[0] = _mm_add_epi32(u[0], x);
  u[0] = _mm_add_epi32(u[0], rnding);
  u[0] = _mm_srai_epi32(u[0], bit);

  u[1] = _mm_mullo_epi32(in[15], cospi60);
  x = _mm_mullo_epi32(in[1], cospi4);
  u[1] = _mm_sub_epi32(u[1], x);
  u[1] = _mm_add_epi32(u[1], rnding);
  u[1] = _mm_srai_epi32(u[1], bit);

  // (2)
  u[2] = _mm_mullo_epi32(in[11], cospi20);
  x = _mm_mullo_epi32(in[5], cospi44);
  u[2] = _mm_add_epi32(u[2], x);
  u[2] = _mm_add_epi32(u[2], rnding);
  u[2] = _mm_srai_epi32(u[2], bit);

  u[3] = _mm_mullo_epi32(in[11], cospi44);
  x = _mm_mullo_epi32(in[5], cospi20);
  u[3] = _mm_sub_epi32(u[3], x);
  u[3] = _mm_add_epi32(u[3], rnding);
  u[3] = _mm_srai_epi32(u[3], bit);

  // (3)
  u[4] = _mm_mullo_epi32(in[7], cospi36);
  x = _mm_mullo_epi32(in[9], cospi28);
  u[4] = _mm_add_epi32(u[4], x);
  u[4] = _mm_add_epi32(u[4], rnding);
  u[4] = _mm_srai_epi32(u[4], bit);

  u[5] = _mm_mullo_epi32(in[7], cospi28);
  x = _mm_mullo_epi32(in[9], cospi36);
  u[5] = _mm_sub_epi32(u[5], x);
  u[5] = _mm_add_epi32(u[5], rnding);
  u[5] = _mm_srai_epi32(u[5], bit);

  // (4)
  u[6] = _mm_mullo_epi32(in[3], cospi52);
  x = _mm_mullo_epi32(in[13], cospi12);
  u[6] = _mm_add_epi32(u[6], x);
  u[6] = _mm_add_epi32(u[6], rnding);
  u[6] = _mm_srai_epi32(u[6], bit);

  u[7] = _mm_mullo_epi32(in[3], cospi12);
  x = _mm_mullo_epi32(in[13], cospi52);
  u[7] = _mm_sub_epi32(u[7], x);
  u[7] = _mm_add_epi32(u[7], rnding);
  u[7] = _mm_srai_epi32(u[7], bit);

  // stage 3
  addsub_sse4_1(u[0], u[4], &v[0], &v[4], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[1], u[5], &v[1], &v[5], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[2], u[6], &v[2], &v[6], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[3], u[7], &v[3], &v[7], &clamp_lo, &clamp_hi);

  // stage 4
  u[0] = v[0];
  u[1] = v[1];
  u[2] = v[2];
  u[3] = v[3];

  u[4] = _mm_mullo_epi32(v[4], cospi16);
  x = _mm_mullo_epi32(v[5], cospi48);
  u[4] = _mm_add_epi32(u[4], x);
  u[4] = _mm_add_epi32(u[4], rnding);
  u[4] = _mm_srai_epi32(u[4], bit);

  u[5] = _mm_mullo_epi32(v[4], cospi48);
  x = _mm_mullo_epi32(v[5], cospi16);
  u[5] = _mm_sub_epi32(u[5], x);
  u[5] = _mm_add_epi32(u[5], rnding);
  u[5] = _mm_srai_epi32(u[5], bit);

  u[6] = _mm_mullo_epi32(v[6], cospim48);
  x = _mm_mullo_epi32(v[7], cospi16);
  u[6] = _mm_add_epi32(u[6], x);
  u[6] = _mm_add_epi32(u[6], rnding);
  u[6] = _mm_srai_epi32(u[6], bit);

  u[7] = _mm_mullo_epi32(v[6], cospi16);
  x = _mm_mullo_epi32(v[7], cospim48);
  u[7] = _mm_sub_epi32(u[7], x);
  u[7] = _mm_add_epi32(u[7], rnding);
  u[7] = _mm_srai_epi32(u[7], bit);

  // stage 5
  addsub_sse4_1(u[0], u[2], &v[0], &v[2], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[1], u[3], &v[1], &v[3], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[4], u[6], &v[4], &v[6], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[5], u[7], &v[5], &v[7], &clamp_lo, &clamp_hi);

  // stage 6
  u[0] = v[0];
  u[1] = v[1];
  u[4] = v[4];
  u[5] = v[5];

  v[0] = _mm_mullo_epi32(v[2], cospi32);
  x = _mm_mullo_epi32(v[3], cospi32);
  u[2] = _mm_add_epi32(v[0], x);
  u[2] = _mm_add_epi32(u[2], rnding);
  u[2] = _mm_srai_epi32(u[2], bit);

  u[3] = _mm_sub_epi32(v[0], x);
  u[3] = _mm_add_epi32(u[3], rnding);
  u[3] = _mm_srai_epi32(u[3], bit);

  v[0] = _mm_mullo_epi32(v[6], cospi32);
  x = _mm_mullo_epi32(v[7], cospi32);
  u[6] = _mm_add_epi32(v[0], x);
  u[6] = _mm_add_epi32(u[6], rnding);
  u[6] = _mm_srai_epi32(u[6], bit);

  u[7] = _mm_sub_epi32(v[0], x);
  u[7] = _mm_add_epi32(u[7], rnding);
  u[7] = _mm_srai_epi32(u[7], bit);

  // stage 7
  if (do_cols) {
    out[1] = u[0];
    out[3] = _mm_sub_epi32(kZero, u[4]);
    out[5] = u[6];
    out[7] = _mm_sub_epi32(kZero, u[2]);
    out[9] = u[3];
    out[11] = _mm_sub_epi32(kZero, u[7]);
    out[13] = u[5];
    out[15] = _mm_sub_epi32(kZero, u[1]);
  } else {
    const int log_range_out = AOMMAX(16, bd + 6);
    const __m128i clamp_lo_out = _mm_set1_epi32(-(1 << (log_range_out - 1)));
    const __m128i clamp_hi_out = _mm_set1_epi32((1 << (log_range_out - 1)) - 1);

    neg_shift_sse4_1(u[0], u[4], out + 1, out + 3, &clamp_lo_out, &clamp_hi_out,
                     out_shift);
    neg_shift_sse4_1(u[6], u[2], out + 5, out + 7, &clamp_lo_out, &clamp_hi_out,
                     out_shift);
    neg_shift_sse4_1(u[3], u[7], out + 9, out + 11, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
    neg_shift_sse4_1(u[5], u[1], out + 13, out + 15, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
  }
}
#endif  // USE_TUNED_ADST8

static void iidentity8_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                              int bd, int out_shift) {
  (void)bit;
  out[0] = _mm_add_epi32(in[0], in[0]);
  out[1] = _mm_add_epi32(in[1], in[1]);
  out[2] = _mm_add_epi32(in[2], in[2]);
  out[3] = _mm_add_epi32(in[3], in[3]);
  out[4] = _mm_add_epi32(in[4], in[4]);
  out[5] = _mm_add_epi32(in[5], in[5]);
  out[6] = _mm_add_epi32(in[6], in[6]);
  out[7] = _mm_add_epi32(in[7], in[7]);

  if (!do_cols) {
    const int log_range = AOMMAX(16, bd + 6);
    const __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
    const __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
    round_shift_4x4(out, out_shift);
    round_shift_4x4(out + 4, out_shift);
    highbd_clamp_epi32_sse4_1(out, out, &clamp_lo, &clamp_hi, 8);
  }
}

static __m128i get_recon_8x8(const __m128i pred, __m128i res_lo, __m128i res_hi,
                             int fliplr, int bd) {
  __m128i x0, x1;
  const __m128i zero = _mm_setzero_si128();

  x0 = _mm_unpacklo_epi16(pred, zero);
  x1 = _mm_unpackhi_epi16(pred, zero);

  if (fliplr) {
    res_lo = _mm_shuffle_epi32(res_lo, 0x1B);
    res_hi = _mm_shuffle_epi32(res_hi, 0x1B);
    x0 = _mm_add_epi32(res_hi, x0);
    x1 = _mm_add_epi32(res_lo, x1);

  } else {
    x0 = _mm_add_epi32(res_lo, x0);
    x1 = _mm_add_epi32(res_hi, x1);
  }

  x0 = _mm_packus_epi32(x0, x1);
  return highbd_clamp_epi16(x0, bd);
}

static void write_buffer_8x8(__m128i *in, uint16_t *output, int stride,
                             int fliplr, int flipud, int shift, int bd) {
  __m128i u0, u1, u2, u3, u4, u5, u6, u7;
  __m128i v0, v1, v2, v3, v4, v5, v6, v7;

  round_shift_8x8(in, shift);

  v0 = _mm_loadu_si128((__m128i const *)(output + 0 * stride));
  v1 = _mm_loadu_si128((__m128i const *)(output + 1 * stride));
  v2 = _mm_loadu_si128((__m128i const *)(output + 2 * stride));
  v3 = _mm_loadu_si128((__m128i const *)(output + 3 * stride));
  v4 = _mm_loadu_si128((__m128i const *)(output + 4 * stride));
  v5 = _mm_loadu_si128((__m128i const *)(output + 5 * stride));
  v6 = _mm_loadu_si128((__m128i const *)(output + 6 * stride));
  v7 = _mm_loadu_si128((__m128i const *)(output + 7 * stride));

  if (flipud) {
    u0 = get_recon_8x8(v0, in[14], in[15], fliplr, bd);
    u1 = get_recon_8x8(v1, in[12], in[13], fliplr, bd);
    u2 = get_recon_8x8(v2, in[10], in[11], fliplr, bd);
    u3 = get_recon_8x8(v3, in[8], in[9], fliplr, bd);
    u4 = get_recon_8x8(v4, in[6], in[7], fliplr, bd);
    u5 = get_recon_8x8(v5, in[4], in[5], fliplr, bd);
    u6 = get_recon_8x8(v6, in[2], in[3], fliplr, bd);
    u7 = get_recon_8x8(v7, in[0], in[1], fliplr, bd);
  } else {
    u0 = get_recon_8x8(v0, in[0], in[1], fliplr, bd);
    u1 = get_recon_8x8(v1, in[2], in[3], fliplr, bd);
    u2 = get_recon_8x8(v2, in[4], in[5], fliplr, bd);
    u3 = get_recon_8x8(v3, in[6], in[7], fliplr, bd);
    u4 = get_recon_8x8(v4, in[8], in[9], fliplr, bd);
    u5 = get_recon_8x8(v5, in[10], in[11], fliplr, bd);
    u6 = get_recon_8x8(v6, in[12], in[13], fliplr, bd);
    u7 = get_recon_8x8(v7, in[14], in[15], fliplr, bd);
  }

  _mm_storeu_si128((__m128i *)(output + 0 * stride), u0);
  _mm_storeu_si128((__m128i *)(output + 1 * stride), u1);
  _mm_storeu_si128((__m128i *)(output + 2 * stride), u2);
  _mm_storeu_si128((__m128i *)(output + 3 * stride), u3);
  _mm_storeu_si128((__m128i *)(output + 4 * stride), u4);
  _mm_storeu_si128((__m128i *)(output + 5 * stride), u5);
  _mm_storeu_si128((__m128i *)(output + 6 * stride), u6);
  _mm_storeu_si128((__m128i *)(output + 7 * stride), u7);
}

void av1_inv_txfm2d_add_8x8_sse4_1(const int32_t *input, uint16_t *output,
                                   int stride, TX_TYPE tx_type, int use_ddt,
                                   int bd) {
  __m128i in[16], out[16];
  const int8_t *shift = av1_inv_txfm_shift_ls[TX_8X8];
  const int txw_idx = get_txw_idx(TX_8X8);
  const int txh_idx = get_txh_idx(TX_8X8);

  switch (tx_type) {
    case DCT_DCT:
      load_buffer_8x8(input, in);
      transpose_8x8(in, out);
      idct8x8_sse4_1(out, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                     -shift[0]);
      transpose_8x8(in, out);
      idct8x8_sse4_1(out, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);
      write_buffer_8x8(in, output, stride, 0, 0, -shift[1], bd);
      break;
    case DCT_ADST:
      load_buffer_8x8(input, in);
      transpose_8x8(in, out);
      if (use_ddt && REPLACE_ADST8)
        iddt8x8_sse4_1(out, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                       -shift[0]);
      else
        iadst8x8_sse4_1(out, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                        -shift[0]);
      transpose_8x8(in, out);
      idct8x8_sse4_1(out, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);
      write_buffer_8x8(in, output, stride, 0, 0, -shift[1], bd);
      break;
    case ADST_DCT:
      load_buffer_8x8(input, in);
      transpose_8x8(in, out);
      idct8x8_sse4_1(out, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                     -shift[0]);
      transpose_8x8(in, out);
      if (use_ddt && REPLACE_ADST8)
        iddt8x8_sse4_1(out, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                       0);
      else
        iadst8x8_sse4_1(out, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                        0);
      write_buffer_8x8(in, output, stride, 0, 0, -shift[1], bd);
      break;
    case ADST_ADST:
      load_buffer_8x8(input, in);
      transpose_8x8(in, out);
      if (use_ddt && REPLACE_ADST8)
        iddt8x8_sse4_1(out, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                       -shift[0]);
      else
        iadst8x8_sse4_1(out, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                        -shift[0]);
      transpose_8x8(in, out);
      if (use_ddt && REPLACE_ADST8)
        iddt8x8_sse4_1(out, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                       0);
      else
        iadst8x8_sse4_1(out, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                        0);
      write_buffer_8x8(in, output, stride, 0, 0, -shift[1], bd);
      break;
    case FLIPADST_DCT:
      load_buffer_8x8(input, in);
      transpose_8x8(in, out);
      idct8x8_sse4_1(out, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                     -shift[0]);
      transpose_8x8(in, out);
      if (use_ddt && REPLACE_ADST8)
        iddt8x8_sse4_1(out, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                       0);
      else
        iadst8x8_sse4_1(out, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                        0);
      write_buffer_8x8(in, output, stride, 0, 1, -shift[1], bd);
      break;
    case DCT_FLIPADST:
      load_buffer_8x8(input, in);
      transpose_8x8(in, out);
      if (use_ddt && REPLACE_ADST8)
        iddt8x8_sse4_1(out, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                       -shift[0]);
      else
        iadst8x8_sse4_1(out, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                        -shift[0]);
      transpose_8x8(in, out);
      idct8x8_sse4_1(out, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);
      write_buffer_8x8(in, output, stride, 1, 0, -shift[1], bd);
      break;
    case ADST_FLIPADST:
      load_buffer_8x8(input, in);
      transpose_8x8(in, out);
      if (use_ddt && REPLACE_ADST8)
        iddt8x8_sse4_1(out, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                       -shift[0]);
      else
        iadst8x8_sse4_1(out, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                        -shift[0]);
      transpose_8x8(in, out);
      if (use_ddt && REPLACE_ADST8)
        iddt8x8_sse4_1(out, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                       0);
      else
        iadst8x8_sse4_1(out, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                        0);
      write_buffer_8x8(in, output, stride, 1, 0, -shift[1], bd);
      break;
    case FLIPADST_FLIPADST:
      load_buffer_8x8(input, in);
      transpose_8x8(in, out);
      if (use_ddt && REPLACE_ADST8)
        iddt8x8_sse4_1(out, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                       -shift[0]);
      else
        iadst8x8_sse4_1(out, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                        -shift[0]);
      transpose_8x8(in, out);
      if (use_ddt && REPLACE_ADST8)
        iddt8x8_sse4_1(out, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                       0);
      else
        iadst8x8_sse4_1(out, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                        0);
      write_buffer_8x8(in, output, stride, 1, 1, -shift[1], bd);
      break;
    case FLIPADST_ADST:
      load_buffer_8x8(input, in);
      transpose_8x8(in, out);
      if (use_ddt && REPLACE_ADST8)
        iddt8x8_sse4_1(out, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                       -shift[0]);
      else
        iadst8x8_sse4_1(out, in, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
                        -shift[0]);
      transpose_8x8(in, out);
      if (use_ddt && REPLACE_ADST8)
        iddt8x8_sse4_1(out, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                       0);
      else
        iadst8x8_sse4_1(out, in, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd,
                        0);
      write_buffer_8x8(in, output, stride, 0, 1, -shift[1], bd);
      break;
    default: assert(0);
  }
}

static void idct8x8_low1_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                                int bd, int out_shift) {
  const int32_t *cospi = cospi_arr(bit);
  const __m128i cospi32 = _mm_set1_epi32(cospi[32]);
  const __m128i rnding = _mm_set1_epi32(1 << (bit - 1));
  const int log_range = AOMMAX(16, bd + (do_cols ? 6 : 8));
  __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
  __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
  __m128i x;

  // stage 0
  // stage 1
  // stage 2
  // stage 3
  x = _mm_mullo_epi32(in[0], cospi32);
  x = _mm_add_epi32(x, rnding);
  x = _mm_srai_epi32(x, bit);

  // stage 4
  // stage 5
  if (!do_cols) {
    const int log_range_out = AOMMAX(16, bd + 6);
    assert(log_range_out - 1 < 32);  // Ensure left-shift doesn't overflow.
    clamp_lo = _mm_set1_epi32(-(1 << (log_range_out - 1)));
    clamp_hi = _mm_set1_epi32((1 << (log_range_out - 1)) - 1);

    __m128i offset = _mm_set1_epi32((1 << out_shift) >> 1);
    x = _mm_add_epi32(x, offset);
    x = _mm_sra_epi32(x, _mm_cvtsi32_si128(out_shift));
  }

  x = _mm_max_epi32(x, clamp_lo);
  x = _mm_min_epi32(x, clamp_hi);
  out[0] = x;
  out[1] = x;
  out[2] = x;
  out[3] = x;
  out[4] = x;
  out[5] = x;
  out[6] = x;
  out[7] = x;
}

static void idct8x8_new_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                               int bd, int out_shift) {
  const int32_t *cospi = cospi_arr(bit);
  const __m128i cospi56 = _mm_set1_epi32(cospi[56]);
  const __m128i cospim8 = _mm_set1_epi32(-cospi[8]);
  const __m128i cospi24 = _mm_set1_epi32(cospi[24]);
  const __m128i cospim40 = _mm_set1_epi32(-cospi[40]);
  const __m128i cospi40 = _mm_set1_epi32(cospi[40]);
  const __m128i cospi8 = _mm_set1_epi32(cospi[8]);
  const __m128i cospi32 = _mm_set1_epi32(cospi[32]);
  const __m128i cospi48 = _mm_set1_epi32(cospi[48]);
  const __m128i cospim16 = _mm_set1_epi32(-cospi[16]);
  const __m128i cospi16 = _mm_set1_epi32(cospi[16]);
  const __m128i rnding = _mm_set1_epi32(1 << (bit - 1));
  const int log_range = AOMMAX(16, bd + (do_cols ? 6 : 8));
  const __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
  const __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
  __m128i u0, u1, u2, u3, u4, u5, u6, u7;
  __m128i v0, v1, v2, v3, v4, v5, v6, v7;
  __m128i x, y;

  // stage 0
  // stage 1
  // stage 2
  u0 = in[0];
  u1 = in[4];
  u2 = in[2];
  u3 = in[6];

  x = _mm_mullo_epi32(in[1], cospi56);
  y = _mm_mullo_epi32(in[7], cospim8);
  u4 = _mm_add_epi32(x, y);
  u4 = _mm_add_epi32(u4, rnding);
  u4 = _mm_srai_epi32(u4, bit);

  x = _mm_mullo_epi32(in[1], cospi8);
  y = _mm_mullo_epi32(in[7], cospi56);
  u7 = _mm_add_epi32(x, y);
  u7 = _mm_add_epi32(u7, rnding);
  u7 = _mm_srai_epi32(u7, bit);

  x = _mm_mullo_epi32(in[5], cospi24);
  y = _mm_mullo_epi32(in[3], cospim40);
  u5 = _mm_add_epi32(x, y);
  u5 = _mm_add_epi32(u5, rnding);
  u5 = _mm_srai_epi32(u5, bit);

  x = _mm_mullo_epi32(in[5], cospi40);
  y = _mm_mullo_epi32(in[3], cospi24);
  u6 = _mm_add_epi32(x, y);
  u6 = _mm_add_epi32(u6, rnding);
  u6 = _mm_srai_epi32(u6, bit);

  // stage 3
  x = _mm_mullo_epi32(u0, cospi32);
  y = _mm_mullo_epi32(u1, cospi32);
  v0 = _mm_add_epi32(x, y);
  v0 = _mm_add_epi32(v0, rnding);
  v0 = _mm_srai_epi32(v0, bit);

  v1 = _mm_sub_epi32(x, y);
  v1 = _mm_add_epi32(v1, rnding);
  v1 = _mm_srai_epi32(v1, bit);

  x = _mm_mullo_epi32(u2, cospi48);
  y = _mm_mullo_epi32(u3, cospim16);
  v2 = _mm_add_epi32(x, y);
  v2 = _mm_add_epi32(v2, rnding);
  v2 = _mm_srai_epi32(v2, bit);

  x = _mm_mullo_epi32(u2, cospi16);
  y = _mm_mullo_epi32(u3, cospi48);
  v3 = _mm_add_epi32(x, y);
  v3 = _mm_add_epi32(v3, rnding);
  v3 = _mm_srai_epi32(v3, bit);

  addsub_sse4_1(u4, u5, &v4, &v5, &clamp_lo, &clamp_hi);
  addsub_sse4_1(u7, u6, &v7, &v6, &clamp_lo, &clamp_hi);

  // stage 4
  addsub_sse4_1(v0, v3, &u0, &u3, &clamp_lo, &clamp_hi);
  addsub_sse4_1(v1, v2, &u1, &u2, &clamp_lo, &clamp_hi);
  u4 = v4;
  u7 = v7;

  x = _mm_mullo_epi32(v5, cospi32);
  y = _mm_mullo_epi32(v6, cospi32);
  u6 = _mm_add_epi32(y, x);
  u6 = _mm_add_epi32(u6, rnding);
  u6 = _mm_srai_epi32(u6, bit);

  u5 = _mm_sub_epi32(y, x);
  u5 = _mm_add_epi32(u5, rnding);
  u5 = _mm_srai_epi32(u5, bit);

  // stage 5
  addsub_sse4_1(u0, u7, out + 0, out + 7, &clamp_lo, &clamp_hi);
  addsub_sse4_1(u1, u6, out + 1, out + 6, &clamp_lo, &clamp_hi);
  addsub_sse4_1(u2, u5, out + 2, out + 5, &clamp_lo, &clamp_hi);
  addsub_sse4_1(u3, u4, out + 3, out + 4, &clamp_lo, &clamp_hi);

  if (!do_cols) {
    const int log_range_out = AOMMAX(16, bd + 6);
    assert(log_range_out - 1 < 32);  // Ensure left-shift doesn't overflow.
    const __m128i clamp_lo_out = _mm_set1_epi32(-(1 << (log_range_out - 1)));
    const __m128i clamp_hi_out = _mm_set1_epi32((1 << (log_range_out - 1)) - 1);

    round_shift_4x4(out, out_shift);
    round_shift_4x4(out + 4, out_shift);
    highbd_clamp_epi32_sse4_1(out, out, &clamp_lo_out, &clamp_hi_out, 8);
  }
}
static void iddt8x8_low1_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                                int bd, int out_shift) {
  (void)bit;
  iadst_matrix_mult_sse4(in, out, INV_DDT_BIT, do_cols, bd, out_shift,
                         ddt8_kernel[INV_TXFM], 8, 1, 1);
}

#if USE_TUNED_ADST8
static void iadst8x8_low1_sse4_1(__m128i *in, __m128i *out, int bit,
                                 int do_cols, int bd, int out_shift) {
  (void)bit;
  iadst_matrix_mult_sse4(in, out, INV_ADST_BIT, do_cols, bd, out_shift,
                         av2_adst_kernel8[INV_TXFM], tx_size_wide[TX_8X8], 1,
                         1);
}
#else
static void iadst8x8_low1_sse4_1(__m128i *in, __m128i *out, int bit,
                                 int do_cols, int bd, int out_shift) {
  const int32_t *cospi = cospi_arr(bit);
  const __m128i cospi4 = _mm_set1_epi32(cospi[4]);
  const __m128i cospi60 = _mm_set1_epi32(cospi[60]);
  const __m128i cospi16 = _mm_set1_epi32(cospi[16]);
  const __m128i cospi48 = _mm_set1_epi32(cospi[48]);
  const __m128i cospi32 = _mm_set1_epi32(cospi[32]);
  const __m128i rnding = _mm_set1_epi32(1 << (bit - 1));
  const __m128i kZero = _mm_setzero_si128();
  __m128i u[8], x;

  // stage 0
  // stage 1
  // stage 2

  x = _mm_mullo_epi32(in[0], cospi60);
  u[0] = _mm_add_epi32(x, rnding);
  u[0] = _mm_srai_epi32(u[0], bit);

  x = _mm_mullo_epi32(in[0], cospi4);
  u[1] = _mm_sub_epi32(kZero, x);
  u[1] = _mm_add_epi32(u[1], rnding);
  u[1] = _mm_srai_epi32(u[1], bit);

  // stage 3
  // stage 4
  __m128i temp1, temp2;
  temp1 = _mm_mullo_epi32(u[0], cospi16);
  x = _mm_mullo_epi32(u[1], cospi48);
  temp1 = _mm_add_epi32(temp1, x);
  temp1 = _mm_add_epi32(temp1, rnding);
  temp1 = _mm_srai_epi32(temp1, bit);
  u[4] = temp1;

  temp2 = _mm_mullo_epi32(u[0], cospi48);
  x = _mm_mullo_epi32(u[1], cospi16);
  u[5] = _mm_sub_epi32(temp2, x);
  u[5] = _mm_add_epi32(u[5], rnding);
  u[5] = _mm_srai_epi32(u[5], bit);

  // stage 5
  // stage 6
  temp1 = _mm_mullo_epi32(u[0], cospi32);
  x = _mm_mullo_epi32(u[1], cospi32);
  u[2] = _mm_add_epi32(temp1, x);
  u[2] = _mm_add_epi32(u[2], rnding);
  u[2] = _mm_srai_epi32(u[2], bit);

  u[3] = _mm_sub_epi32(temp1, x);
  u[3] = _mm_add_epi32(u[3], rnding);
  u[3] = _mm_srai_epi32(u[3], bit);

  temp1 = _mm_mullo_epi32(u[4], cospi32);
  x = _mm_mullo_epi32(u[5], cospi32);
  u[6] = _mm_add_epi32(temp1, x);
  u[6] = _mm_add_epi32(u[6], rnding);
  u[6] = _mm_srai_epi32(u[6], bit);

  u[7] = _mm_sub_epi32(temp1, x);
  u[7] = _mm_add_epi32(u[7], rnding);
  u[7] = _mm_srai_epi32(u[7], bit);

  // stage 7
  if (do_cols) {
    out[0] = u[0];
    out[1] = _mm_sub_epi32(kZero, u[4]);
    out[2] = u[6];
    out[3] = _mm_sub_epi32(kZero, u[2]);
    out[4] = u[3];
    out[5] = _mm_sub_epi32(kZero, u[7]);
    out[6] = u[5];
    out[7] = _mm_sub_epi32(kZero, u[1]);
  } else {
    const int log_range_out = AOMMAX(16, bd + 6);
    const __m128i clamp_lo_out = _mm_set1_epi32(-(1 << (log_range_out - 1)));
    const __m128i clamp_hi_out = _mm_set1_epi32((1 << (log_range_out - 1)) - 1);

    neg_shift_sse4_1(u[0], u[4], out + 0, out + 1, &clamp_lo_out, &clamp_hi_out,
                     out_shift);
    neg_shift_sse4_1(u[6], u[2], out + 2, out + 3, &clamp_lo_out, &clamp_hi_out,
                     out_shift);
    neg_shift_sse4_1(u[3], u[7], out + 4, out + 5, &clamp_lo_out, &clamp_hi_out,
                     out_shift);
    neg_shift_sse4_1(u[5], u[1], out + 6, out + 7, &clamp_lo_out, &clamp_hi_out,
                     out_shift);
  }
}
#endif  // USE_TUNED_ADST8

static void iddt8x8_new_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                               int bd, int out_shift) {
  (void)bit;
  iadst_matrix_mult_sse4(in, out, INV_DDT_BIT, do_cols, bd, out_shift,
                         ddt8_kernel[INV_TXFM], 8, 8, 1);
}

#if USE_TUNED_ADST8
static void iadst8x8_new_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                                int bd, int out_shift) {
  (void)bit;
  iadst_matrix_mult_sse4(in, out, INV_ADST_BIT, do_cols, bd, out_shift,
                         av2_adst_kernel8[INV_TXFM], tx_size_wide[TX_8X8],
                         tx_size_wide[TX_8X8], 1);
}
#else
static void iadst8x8_new_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                                int bd, int out_shift) {
  const int32_t *cospi = cospi_arr(bit);
  const __m128i cospi4 = _mm_set1_epi32(cospi[4]);
  const __m128i cospi60 = _mm_set1_epi32(cospi[60]);
  const __m128i cospi20 = _mm_set1_epi32(cospi[20]);
  const __m128i cospi44 = _mm_set1_epi32(cospi[44]);
  const __m128i cospi36 = _mm_set1_epi32(cospi[36]);
  const __m128i cospi28 = _mm_set1_epi32(cospi[28]);
  const __m128i cospi52 = _mm_set1_epi32(cospi[52]);
  const __m128i cospi12 = _mm_set1_epi32(cospi[12]);
  const __m128i cospi16 = _mm_set1_epi32(cospi[16]);
  const __m128i cospi48 = _mm_set1_epi32(cospi[48]);
  const __m128i cospim48 = _mm_set1_epi32(-cospi[48]);
  const __m128i cospi32 = _mm_set1_epi32(cospi[32]);
  const __m128i rnding = _mm_set1_epi32(1 << (bit - 1));
  const __m128i kZero = _mm_setzero_si128();
  const int log_range = AOMMAX(16, bd + (do_cols ? 6 : 8));
  const __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
  const __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
  __m128i u[8], v[8], x;

  // stage 0
  // stage 1
  // stage 2

  u[0] = _mm_mullo_epi32(in[7], cospi4);
  x = _mm_mullo_epi32(in[0], cospi60);
  u[0] = _mm_add_epi32(u[0], x);
  u[0] = _mm_add_epi32(u[0], rnding);
  u[0] = _mm_srai_epi32(u[0], bit);

  u[1] = _mm_mullo_epi32(in[7], cospi60);
  x = _mm_mullo_epi32(in[0], cospi4);
  u[1] = _mm_sub_epi32(u[1], x);
  u[1] = _mm_add_epi32(u[1], rnding);
  u[1] = _mm_srai_epi32(u[1], bit);

  // (2)
  u[2] = _mm_mullo_epi32(in[5], cospi20);
  x = _mm_mullo_epi32(in[2], cospi44);
  u[2] = _mm_add_epi32(u[2], x);
  u[2] = _mm_add_epi32(u[2], rnding);
  u[2] = _mm_srai_epi32(u[2], bit);

  u[3] = _mm_mullo_epi32(in[5], cospi44);
  x = _mm_mullo_epi32(in[2], cospi20);
  u[3] = _mm_sub_epi32(u[3], x);
  u[3] = _mm_add_epi32(u[3], rnding);
  u[3] = _mm_srai_epi32(u[3], bit);

  // (3)
  u[4] = _mm_mullo_epi32(in[3], cospi36);
  x = _mm_mullo_epi32(in[4], cospi28);
  u[4] = _mm_add_epi32(u[4], x);
  u[4] = _mm_add_epi32(u[4], rnding);
  u[4] = _mm_srai_epi32(u[4], bit);

  u[5] = _mm_mullo_epi32(in[3], cospi28);
  x = _mm_mullo_epi32(in[4], cospi36);
  u[5] = _mm_sub_epi32(u[5], x);
  u[5] = _mm_add_epi32(u[5], rnding);
  u[5] = _mm_srai_epi32(u[5], bit);

  // (4)
  u[6] = _mm_mullo_epi32(in[1], cospi52);
  x = _mm_mullo_epi32(in[6], cospi12);
  u[6] = _mm_add_epi32(u[6], x);
  u[6] = _mm_add_epi32(u[6], rnding);
  u[6] = _mm_srai_epi32(u[6], bit);

  u[7] = _mm_mullo_epi32(in[1], cospi12);
  x = _mm_mullo_epi32(in[6], cospi52);
  u[7] = _mm_sub_epi32(u[7], x);
  u[7] = _mm_add_epi32(u[7], rnding);
  u[7] = _mm_srai_epi32(u[7], bit);

  // stage 3
  addsub_sse4_1(u[0], u[4], &v[0], &v[4], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[1], u[5], &v[1], &v[5], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[2], u[6], &v[2], &v[6], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[3], u[7], &v[3], &v[7], &clamp_lo, &clamp_hi);

  // stage 4
  u[0] = v[0];
  u[1] = v[1];
  u[2] = v[2];
  u[3] = v[3];

  u[4] = _mm_mullo_epi32(v[4], cospi16);
  x = _mm_mullo_epi32(v[5], cospi48);
  u[4] = _mm_add_epi32(u[4], x);
  u[4] = _mm_add_epi32(u[4], rnding);
  u[4] = _mm_srai_epi32(u[4], bit);

  u[5] = _mm_mullo_epi32(v[4], cospi48);
  x = _mm_mullo_epi32(v[5], cospi16);
  u[5] = _mm_sub_epi32(u[5], x);
  u[5] = _mm_add_epi32(u[5], rnding);
  u[5] = _mm_srai_epi32(u[5], bit);

  u[6] = _mm_mullo_epi32(v[6], cospim48);
  x = _mm_mullo_epi32(v[7], cospi16);
  u[6] = _mm_add_epi32(u[6], x);
  u[6] = _mm_add_epi32(u[6], rnding);
  u[6] = _mm_srai_epi32(u[6], bit);

  u[7] = _mm_mullo_epi32(v[6], cospi16);
  x = _mm_mullo_epi32(v[7], cospim48);
  u[7] = _mm_sub_epi32(u[7], x);
  u[7] = _mm_add_epi32(u[7], rnding);
  u[7] = _mm_srai_epi32(u[7], bit);

  // stage 5
  addsub_sse4_1(u[0], u[2], &v[0], &v[2], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[1], u[3], &v[1], &v[3], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[4], u[6], &v[4], &v[6], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[5], u[7], &v[5], &v[7], &clamp_lo, &clamp_hi);

  // stage 6
  u[0] = v[0];
  u[1] = v[1];
  u[4] = v[4];
  u[5] = v[5];

  v[0] = _mm_mullo_epi32(v[2], cospi32);
  x = _mm_mullo_epi32(v[3], cospi32);
  u[2] = _mm_add_epi32(v[0], x);
  u[2] = _mm_add_epi32(u[2], rnding);
  u[2] = _mm_srai_epi32(u[2], bit);

  u[3] = _mm_sub_epi32(v[0], x);
  u[3] = _mm_add_epi32(u[3], rnding);
  u[3] = _mm_srai_epi32(u[3], bit);

  v[0] = _mm_mullo_epi32(v[6], cospi32);
  x = _mm_mullo_epi32(v[7], cospi32);
  u[6] = _mm_add_epi32(v[0], x);
  u[6] = _mm_add_epi32(u[6], rnding);
  u[6] = _mm_srai_epi32(u[6], bit);

  u[7] = _mm_sub_epi32(v[0], x);
  u[7] = _mm_add_epi32(u[7], rnding);
  u[7] = _mm_srai_epi32(u[7], bit);

  // stage 7
  if (do_cols) {
    out[0] = u[0];
    out[1] = _mm_sub_epi32(kZero, u[4]);
    out[2] = u[6];
    out[3] = _mm_sub_epi32(kZero, u[2]);
    out[4] = u[3];
    out[5] = _mm_sub_epi32(kZero, u[7]);
    out[6] = u[5];
    out[7] = _mm_sub_epi32(kZero, u[1]);
  } else {
    const int log_range_out = AOMMAX(16, bd + 6);
    const __m128i clamp_lo_out = _mm_set1_epi32(-(1 << (log_range_out - 1)));
    const __m128i clamp_hi_out = _mm_set1_epi32((1 << (log_range_out - 1)) - 1);

    neg_shift_sse4_1(u[0], u[4], out + 0, out + 1, &clamp_lo_out, &clamp_hi_out,
                     out_shift);
    neg_shift_sse4_1(u[6], u[2], out + 2, out + 3, &clamp_lo_out, &clamp_hi_out,
                     out_shift);
    neg_shift_sse4_1(u[3], u[7], out + 4, out + 5, &clamp_lo_out, &clamp_hi_out,
                     out_shift);
    neg_shift_sse4_1(u[5], u[1], out + 6, out + 7, &clamp_lo_out, &clamp_hi_out,
                     out_shift);
  }
}
#endif  // USE_TUNED_ADST8

static void idct16x16_low1_sse4_1(__m128i *in, __m128i *out, int bit,
                                  int do_cols, int bd, int out_shift) {
  const int32_t *cospi = cospi_arr(bit);
  const __m128i cospi32 = _mm_set1_epi32(cospi[32]);
  const __m128i rnding = _mm_set1_epi32(1 << (bit - 1));
  int log_range = AOMMAX(16, bd + (do_cols ? 6 : 8));
  __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
  __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
  // stage 0
  // stage 1
  // stage 2
  // stage 3
  // stage 4
  in[0] = _mm_mullo_epi32(in[0], cospi32);
  in[0] = _mm_add_epi32(in[0], rnding);
  in[0] = _mm_srai_epi32(in[0], bit);

  // stage 5
  // stage 6
  // stage 7
  if (!do_cols) {
    log_range = AOMMAX(16, bd + 6);
    assert(log_range - 1 < 32);  // Ensure left-shift doesn't overflow.
    clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
    clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
    if (out_shift != 0) {
      __m128i offset = _mm_set1_epi32((1 << out_shift) >> 1);
      in[0] = _mm_add_epi32(in[0], offset);
      in[0] = _mm_sra_epi32(in[0], _mm_cvtsi32_si128(out_shift));
    }
  }

  in[0] = _mm_max_epi32(in[0], clamp_lo);
  in[0] = _mm_min_epi32(in[0], clamp_hi);
  out[0] = in[0];
  out[1] = in[0];
  out[2] = in[0];
  out[3] = in[0];
  out[4] = in[0];
  out[5] = in[0];
  out[6] = in[0];
  out[7] = in[0];
  out[8] = in[0];
  out[9] = in[0];
  out[10] = in[0];
  out[11] = in[0];
  out[12] = in[0];
  out[13] = in[0];
  out[14] = in[0];
  out[15] = in[0];
}

static void idct16x16_low8_sse4_1(__m128i *in, __m128i *out, int bit,
                                  int do_cols, int bd, int out_shift) {
  const int32_t *cospi = cospi_arr(bit);
  const __m128i cospi60 = _mm_set1_epi32(cospi[60]);
  const __m128i cospi28 = _mm_set1_epi32(cospi[28]);
  const __m128i cospi44 = _mm_set1_epi32(cospi[44]);
  const __m128i cospi20 = _mm_set1_epi32(cospi[20]);
  const __m128i cospi12 = _mm_set1_epi32(cospi[12]);
  const __m128i cospi4 = _mm_set1_epi32(cospi[4]);
  const __m128i cospi56 = _mm_set1_epi32(cospi[56]);
  const __m128i cospi24 = _mm_set1_epi32(cospi[24]);
  const __m128i cospim40 = _mm_set1_epi32(-cospi[40]);
  const __m128i cospi8 = _mm_set1_epi32(cospi[8]);
  const __m128i cospi32 = _mm_set1_epi32(cospi[32]);
  const __m128i cospi48 = _mm_set1_epi32(cospi[48]);
  const __m128i cospi16 = _mm_set1_epi32(cospi[16]);
  const __m128i cospim16 = _mm_set1_epi32(-cospi[16]);
  const __m128i cospim48 = _mm_set1_epi32(-cospi[48]);
  const __m128i cospim36 = _mm_set1_epi32(-cospi[36]);
  const __m128i cospim52 = _mm_set1_epi32(-cospi[52]);
  const __m128i rnding = _mm_set1_epi32(1 << (bit - 1));
  const int log_range = AOMMAX(16, bd + (do_cols ? 6 : 8));
  const __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
  const __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
  __m128i u[16], x, y;
  // stage 0
  // stage 1
  u[0] = in[0];
  u[2] = in[4];
  u[4] = in[2];
  u[6] = in[6];
  u[8] = in[1];
  u[10] = in[5];
  u[12] = in[3];
  u[14] = in[7];

  // stage 2
  u[15] = half_btf_0_sse4_1(&cospi4, &u[8], &rnding, bit);
  u[8] = half_btf_0_sse4_1(&cospi60, &u[8], &rnding, bit);

  u[9] = half_btf_0_sse4_1(&cospim36, &u[14], &rnding, bit);
  u[14] = half_btf_0_sse4_1(&cospi28, &u[14], &rnding, bit);

  u[13] = half_btf_0_sse4_1(&cospi20, &u[10], &rnding, bit);
  u[10] = half_btf_0_sse4_1(&cospi44, &u[10], &rnding, bit);

  u[11] = half_btf_0_sse4_1(&cospim52, &u[12], &rnding, bit);
  u[12] = half_btf_0_sse4_1(&cospi12, &u[12], &rnding, bit);

  // stage 3
  u[7] = half_btf_0_sse4_1(&cospi8, &u[4], &rnding, bit);
  u[4] = half_btf_0_sse4_1(&cospi56, &u[4], &rnding, bit);
  u[5] = half_btf_0_sse4_1(&cospim40, &u[6], &rnding, bit);
  u[6] = half_btf_0_sse4_1(&cospi24, &u[6], &rnding, bit);

  addsub_sse4_1(u[8], u[9], &u[8], &u[9], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[11], u[10], &u[11], &u[10], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[12], u[13], &u[12], &u[13], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[15], u[14], &u[15], &u[14], &clamp_lo, &clamp_hi);

  // stage 4
  x = _mm_mullo_epi32(u[0], cospi32);
  u[0] = _mm_add_epi32(x, rnding);
  u[0] = _mm_srai_epi32(u[0], bit);
  u[1] = u[0];

  u[3] = half_btf_0_sse4_1(&cospi16, &u[2], &rnding, bit);
  u[2] = half_btf_0_sse4_1(&cospi48, &u[2], &rnding, bit);

  addsub_sse4_1(u[4], u[5], &u[4], &u[5], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[7], u[6], &u[7], &u[6], &clamp_lo, &clamp_hi);

  x = half_btf_sse4_1(&cospim16, &u[9], &cospi48, &u[14], &rnding, bit);
  u[14] = half_btf_sse4_1(&cospi48, &u[9], &cospi16, &u[14], &rnding, bit);
  u[9] = x;
  y = half_btf_sse4_1(&cospim48, &u[10], &cospim16, &u[13], &rnding, bit);
  u[13] = half_btf_sse4_1(&cospim16, &u[10], &cospi48, &u[13], &rnding, bit);
  u[10] = y;

  // stage 5
  addsub_sse4_1(u[0], u[3], &u[0], &u[3], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[1], u[2], &u[1], &u[2], &clamp_lo, &clamp_hi);

  x = _mm_mullo_epi32(u[5], cospi32);
  y = _mm_mullo_epi32(u[6], cospi32);
  u[5] = _mm_sub_epi32(y, x);
  u[5] = _mm_add_epi32(u[5], rnding);
  u[5] = _mm_srai_epi32(u[5], bit);

  u[6] = _mm_add_epi32(y, x);
  u[6] = _mm_add_epi32(u[6], rnding);
  u[6] = _mm_srai_epi32(u[6], bit);

  addsub_sse4_1(u[8], u[11], &u[8], &u[11], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[9], u[10], &u[9], &u[10], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[15], u[12], &u[15], &u[12], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[14], u[13], &u[14], &u[13], &clamp_lo, &clamp_hi);

  // stage 6
  addsub_sse4_1(u[0], u[7], &u[0], &u[7], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[1], u[6], &u[1], &u[6], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[2], u[5], &u[2], &u[5], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[3], u[4], &u[3], &u[4], &clamp_lo, &clamp_hi);

  x = _mm_mullo_epi32(u[10], cospi32);
  y = _mm_mullo_epi32(u[13], cospi32);
  u[10] = _mm_sub_epi32(y, x);
  u[10] = _mm_add_epi32(u[10], rnding);
  u[10] = _mm_srai_epi32(u[10], bit);

  u[13] = _mm_add_epi32(x, y);
  u[13] = _mm_add_epi32(u[13], rnding);
  u[13] = _mm_srai_epi32(u[13], bit);

  x = _mm_mullo_epi32(u[11], cospi32);
  y = _mm_mullo_epi32(u[12], cospi32);
  u[11] = _mm_sub_epi32(y, x);
  u[11] = _mm_add_epi32(u[11], rnding);
  u[11] = _mm_srai_epi32(u[11], bit);

  u[12] = _mm_add_epi32(x, y);
  u[12] = _mm_add_epi32(u[12], rnding);
  u[12] = _mm_srai_epi32(u[12], bit);
  // stage 7
  addsub_sse4_1(u[0], u[15], out + 0, out + 15, &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[1], u[14], out + 1, out + 14, &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[2], u[13], out + 2, out + 13, &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[3], u[12], out + 3, out + 12, &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[4], u[11], out + 4, out + 11, &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[5], u[10], out + 5, out + 10, &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[6], u[9], out + 6, out + 9, &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[7], u[8], out + 7, out + 8, &clamp_lo, &clamp_hi);

  if (!do_cols) {
    const int log_range_out = AOMMAX(16, bd + 6);
    assert(log_range_out - 1 < 32);  // Ensure left-shift doesn't overflow.
    const __m128i clamp_lo_out = _mm_set1_epi32(-(1 << (log_range_out - 1)));
    const __m128i clamp_hi_out = _mm_set1_epi32((1 << (log_range_out - 1)) - 1);
    round_shift_8x8(out, out_shift);
    highbd_clamp_epi32_sse4_1(out, out, &clamp_lo_out, &clamp_hi_out, 16);
  }
}

void iadst_matrix_mult_sse4(__m128i *in, __m128i *out, int bit, int do_cols,
                            int bd, int out_shift, const int32_t *kernel,
                            int kernel_size, int num_cols, int col_stride) {
  const __m128i zero = _mm_setzero_si128();
  const __m128i rnding = _mm_set1_epi32(1 << (bit - 1));
  int log_range = AOMMAX(16, bd + (do_cols ? 6 : 8));
  __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
  __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
  __m128i x[16];

  for (int i = 0; i < kernel_size; ++i) {
    int row_idx = i * kernel_size;
    __m128i sum = zero;
    __m128i t;
    for (int j = 0; j < num_cols; ++j) {
      const __m128i coef = _mm_set1_epi32(kernel[row_idx + j]);
      t = _mm_mullo_epi32(in[j * col_stride], coef);
      sum = _mm_add_epi32(sum, t);
    }
    sum = _mm_add_epi32(sum, rnding);
    sum = _mm_srai_epi32(sum, bit);
    sum = _mm_max_epi32(sum, clamp_lo);
    x[i] = _mm_min_epi32(sum, clamp_hi);
  }

  if (!do_cols) {
    log_range = AOMMAX(16, bd + 6);
    assert(log_range - 1 < 32);  // Ensure left-shift doesn't overflow.
    clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
    clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
    if (out_shift != 0) {
      __m128i offset = _mm_set1_epi32((1 << out_shift) >> 1);
      for (int i = 0; i < kernel_size; ++i) {
        x[i] = _mm_add_epi32(x[i], offset);
        x[i] = _mm_sra_epi32(x[i], _mm_cvtsi32_si128(out_shift));
      }
    }
  }

  for (int i = 0; i < kernel_size; ++i) {
    x[i] = _mm_max_epi32(x[i], clamp_lo);
    out[i * col_stride] = _mm_min_epi32(x[i], clamp_hi);
  }
}

static void iddt16x16_low1_sse4_1(__m128i *in, __m128i *out, int bit,
                                  int do_cols, int bd, int out_shift) {
  (void)bit;
  iadst_matrix_mult_sse4(in, out, INV_DDT_BIT, do_cols, bd, out_shift,
                         ddt16_kernel[INV_TXFM], 16, 1, 1);
}

#if USE_TUNED_ADST16
static void iadst16x16_low1_sse4_1(__m128i *in, __m128i *out, int bit,
                                   int do_cols, int bd, int out_shift) {
  (void)bit;
  iadst_matrix_mult_sse4(in, out, INV_ADST_BIT, do_cols, bd, out_shift,
                         av2_adst_kernel16[INV_TXFM], tx_size_wide[TX_16X16], 1,
                         1);
}
#else
static void iadst16x16_low1_sse4_1(__m128i *in, __m128i *out, int bit,
                                   int do_cols, int bd, int out_shift) {
  const int32_t *cospi = cospi_arr(bit);
  const __m128i cospi2 = _mm_set1_epi32(cospi[2]);
  const __m128i cospi62 = _mm_set1_epi32(cospi[62]);
  const __m128i cospi8 = _mm_set1_epi32(cospi[8]);
  const __m128i cospi56 = _mm_set1_epi32(cospi[56]);
  const __m128i cospi48 = _mm_set1_epi32(cospi[48]);
  const __m128i cospi16 = _mm_set1_epi32(cospi[16]);
  const __m128i cospi32 = _mm_set1_epi32(cospi[32]);
  const __m128i rnding = _mm_set1_epi32(1 << (bit - 1));
  const __m128i zero = _mm_setzero_si128();
  __m128i v[16], x, y, temp1, temp2;
  // stage 0
  // stage 1
  // stage 2
  x = _mm_mullo_epi32(in[0], cospi62);
  v[0] = _mm_add_epi32(x, rnding);
  v[0] = _mm_srai_epi32(v[0], bit);

  x = _mm_mullo_epi32(in[0], cospi2);
  v[1] = _mm_sub_epi32(zero, x);
  v[1] = _mm_add_epi32(v[1], rnding);
  v[1] = _mm_srai_epi32(v[1], bit);

  // stage 3
  v[8] = v[0];
  v[9] = v[1];

  // stage 4
  temp1 = _mm_mullo_epi32(v[8], cospi8);
  x = _mm_mullo_epi32(v[9], cospi56);
  temp1 = _mm_add_epi32(temp1, x);
  temp1 = _mm_add_epi32(temp1, rnding);
  temp1 = _mm_srai_epi32(temp1, bit);

  temp2 = _mm_mullo_epi32(v[8], cospi56);
  x = _mm_mullo_epi32(v[9], cospi8);
  temp2 = _mm_sub_epi32(temp2, x);
  temp2 = _mm_add_epi32(temp2, rnding);
  temp2 = _mm_srai_epi32(temp2, bit);
  v[8] = temp1;
  v[9] = temp2;

  // stage 5
  v[4] = v[0];
  v[5] = v[1];
  v[12] = v[8];
  v[13] = v[9];

  // stage 6
  temp1 = _mm_mullo_epi32(v[4], cospi16);
  x = _mm_mullo_epi32(v[5], cospi48);
  temp1 = _mm_add_epi32(temp1, x);
  temp1 = _mm_add_epi32(temp1, rnding);
  temp1 = _mm_srai_epi32(temp1, bit);

  temp2 = _mm_mullo_epi32(v[4], cospi48);
  x = _mm_mullo_epi32(v[5], cospi16);
  temp2 = _mm_sub_epi32(temp2, x);
  temp2 = _mm_add_epi32(temp2, rnding);
  temp2 = _mm_srai_epi32(temp2, bit);
  v[4] = temp1;
  v[5] = temp2;

  temp1 = _mm_mullo_epi32(v[12], cospi16);
  x = _mm_mullo_epi32(v[13], cospi48);
  temp1 = _mm_add_epi32(temp1, x);
  temp1 = _mm_add_epi32(temp1, rnding);
  temp1 = _mm_srai_epi32(temp1, bit);

  temp2 = _mm_mullo_epi32(v[12], cospi48);
  x = _mm_mullo_epi32(v[13], cospi16);
  temp2 = _mm_sub_epi32(temp2, x);
  temp2 = _mm_add_epi32(temp2, rnding);
  temp2 = _mm_srai_epi32(temp2, bit);
  v[12] = temp1;
  v[13] = temp2;

  // stage 7
  v[2] = v[0];
  v[3] = v[1];
  v[6] = v[4];
  v[7] = v[5];
  v[10] = v[8];
  v[11] = v[9];
  v[14] = v[12];
  v[15] = v[13];

  // stage 8
  y = _mm_mullo_epi32(v[2], cospi32);
  x = _mm_mullo_epi32(v[3], cospi32);
  v[2] = _mm_add_epi32(y, x);
  v[2] = _mm_add_epi32(v[2], rnding);
  v[2] = _mm_srai_epi32(v[2], bit);

  v[3] = _mm_sub_epi32(y, x);
  v[3] = _mm_add_epi32(v[3], rnding);
  v[3] = _mm_srai_epi32(v[3], bit);

  y = _mm_mullo_epi32(v[6], cospi32);
  x = _mm_mullo_epi32(v[7], cospi32);
  v[6] = _mm_add_epi32(y, x);
  v[6] = _mm_add_epi32(v[6], rnding);
  v[6] = _mm_srai_epi32(v[6], bit);

  v[7] = _mm_sub_epi32(y, x);
  v[7] = _mm_add_epi32(v[7], rnding);
  v[7] = _mm_srai_epi32(v[7], bit);

  y = _mm_mullo_epi32(v[10], cospi32);
  x = _mm_mullo_epi32(v[11], cospi32);
  v[10] = _mm_add_epi32(y, x);
  v[10] = _mm_add_epi32(v[10], rnding);
  v[10] = _mm_srai_epi32(v[10], bit);

  v[11] = _mm_sub_epi32(y, x);
  v[11] = _mm_add_epi32(v[11], rnding);
  v[11] = _mm_srai_epi32(v[11], bit);

  y = _mm_mullo_epi32(v[14], cospi32);
  x = _mm_mullo_epi32(v[15], cospi32);
  v[14] = _mm_add_epi32(y, x);
  v[14] = _mm_add_epi32(v[14], rnding);
  v[14] = _mm_srai_epi32(v[14], bit);

  v[15] = _mm_sub_epi32(y, x);
  v[15] = _mm_add_epi32(v[15], rnding);
  v[15] = _mm_srai_epi32(v[15], bit);

  // stage 9
  if (do_cols) {
    out[0] = v[0];
    out[1] = _mm_sub_epi32(zero, v[8]);
    out[2] = v[12];
    out[3] = _mm_sub_epi32(zero, v[4]);
    out[4] = v[6];
    out[5] = _mm_sub_epi32(zero, v[14]);
    out[6] = v[10];
    out[7] = _mm_sub_epi32(zero, v[2]);
    out[8] = v[3];
    out[9] = _mm_sub_epi32(zero, v[11]);
    out[10] = v[15];
    out[11] = _mm_sub_epi32(zero, v[7]);
    out[12] = v[5];
    out[13] = _mm_sub_epi32(zero, v[13]);
    out[14] = v[9];
    out[15] = _mm_sub_epi32(zero, v[1]);
  } else {
    const int log_range_out = AOMMAX(16, bd + 6);
    const __m128i clamp_lo_out = _mm_set1_epi32(-(1 << (log_range_out - 1)));
    const __m128i clamp_hi_out = _mm_set1_epi32((1 << (log_range_out - 1)) - 1);

    neg_shift_sse4_1(v[0], v[8], out + 0, out + 1, &clamp_lo_out, &clamp_hi_out,
                     out_shift);
    neg_shift_sse4_1(v[12], v[4], out + 2, out + 3, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
    neg_shift_sse4_1(v[6], v[14], out + 4, out + 5, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
    neg_shift_sse4_1(v[10], v[2], out + 6, out + 7, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
    neg_shift_sse4_1(v[3], v[11], out + 8, out + 9, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
    neg_shift_sse4_1(v[15], v[7], out + 10, out + 11, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
    neg_shift_sse4_1(v[5], v[13], out + 12, out + 13, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
    neg_shift_sse4_1(v[9], v[1], out + 14, out + 15, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
  }
}
#endif  // USE_TUNED_ADST16

static void iddt16x16_low8_sse4_1(__m128i *in, __m128i *out, int bit,
                                  int do_cols, int bd, int out_shift) {
  (void)bit;
  iadst_matrix_mult_sse4(in, out, INV_DDT_BIT, do_cols, bd, out_shift,
                         ddt16_kernel[INV_TXFM], 16, 8, 1);
}

#if USE_TUNED_ADST16
static void iadst16x16_low8_sse4_1(__m128i *in, __m128i *out, int bit,
                                   int do_cols, int bd, int out_shift) {
  (void)bit;
  iadst_matrix_mult_sse4(in, out, INV_ADST_BIT, do_cols, bd, out_shift,
                         av2_adst_kernel16[INV_TXFM], tx_size_wide[TX_16X16], 8,
                         1);
}
#else
static void iadst16x16_low8_sse4_1(__m128i *in, __m128i *out, int bit,
                                   int do_cols, int bd, int out_shift) {
  const int32_t *cospi = cospi_arr(bit);
  const __m128i cospi2 = _mm_set1_epi32(cospi[2]);
  const __m128i cospi62 = _mm_set1_epi32(cospi[62]);
  const __m128i cospi10 = _mm_set1_epi32(cospi[10]);
  const __m128i cospi54 = _mm_set1_epi32(cospi[54]);
  const __m128i cospi18 = _mm_set1_epi32(cospi[18]);
  const __m128i cospi46 = _mm_set1_epi32(cospi[46]);
  const __m128i cospi26 = _mm_set1_epi32(cospi[26]);
  const __m128i cospi38 = _mm_set1_epi32(cospi[38]);
  const __m128i cospi34 = _mm_set1_epi32(cospi[34]);
  const __m128i cospi30 = _mm_set1_epi32(cospi[30]);
  const __m128i cospi42 = _mm_set1_epi32(cospi[42]);
  const __m128i cospi22 = _mm_set1_epi32(cospi[22]);
  const __m128i cospi50 = _mm_set1_epi32(cospi[50]);
  const __m128i cospi14 = _mm_set1_epi32(cospi[14]);
  const __m128i cospi58 = _mm_set1_epi32(cospi[58]);
  const __m128i cospi6 = _mm_set1_epi32(cospi[6]);
  const __m128i cospi8 = _mm_set1_epi32(cospi[8]);
  const __m128i cospi56 = _mm_set1_epi32(cospi[56]);
  const __m128i cospi40 = _mm_set1_epi32(cospi[40]);
  const __m128i cospi24 = _mm_set1_epi32(cospi[24]);
  const __m128i cospim56 = _mm_set1_epi32(-cospi[56]);
  const __m128i cospim24 = _mm_set1_epi32(-cospi[24]);
  const __m128i cospi48 = _mm_set1_epi32(cospi[48]);
  const __m128i cospi16 = _mm_set1_epi32(cospi[16]);
  const __m128i cospim48 = _mm_set1_epi32(-cospi[48]);
  const __m128i cospi32 = _mm_set1_epi32(cospi[32]);
  const __m128i rnding = _mm_set1_epi32(1 << (bit - 1));
  const int log_range = AOMMAX(16, bd + (do_cols ? 6 : 8));
  const __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
  const __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
  __m128i zero = _mm_setzero_si128();
  __m128i u[16], x, y;

  // stage 0
  // stage 1
  // stage 2
  x = _mm_mullo_epi32(in[0], cospi62);
  u[0] = _mm_add_epi32(x, rnding);
  u[0] = _mm_srai_epi32(u[0], bit);

  x = _mm_mullo_epi32(in[0], cospi2);
  u[1] = _mm_sub_epi32(zero, x);
  u[1] = _mm_add_epi32(u[1], rnding);
  u[1] = _mm_srai_epi32(u[1], bit);

  x = _mm_mullo_epi32(in[2], cospi54);
  u[2] = _mm_add_epi32(x, rnding);
  u[2] = _mm_srai_epi32(u[2], bit);

  x = _mm_mullo_epi32(in[2], cospi10);
  u[3] = _mm_sub_epi32(zero, x);
  u[3] = _mm_add_epi32(u[3], rnding);
  u[3] = _mm_srai_epi32(u[3], bit);

  x = _mm_mullo_epi32(in[4], cospi46);
  u[4] = _mm_add_epi32(x, rnding);
  u[4] = _mm_srai_epi32(u[4], bit);

  x = _mm_mullo_epi32(in[4], cospi18);
  u[5] = _mm_sub_epi32(zero, x);
  u[5] = _mm_add_epi32(u[5], rnding);
  u[5] = _mm_srai_epi32(u[5], bit);

  x = _mm_mullo_epi32(in[6], cospi38);
  u[6] = _mm_add_epi32(x, rnding);
  u[6] = _mm_srai_epi32(u[6], bit);

  x = _mm_mullo_epi32(in[6], cospi26);
  u[7] = _mm_sub_epi32(zero, x);
  u[7] = _mm_add_epi32(u[7], rnding);
  u[7] = _mm_srai_epi32(u[7], bit);

  u[8] = _mm_mullo_epi32(in[7], cospi34);
  u[8] = _mm_add_epi32(u[8], rnding);
  u[8] = _mm_srai_epi32(u[8], bit);

  u[9] = _mm_mullo_epi32(in[7], cospi30);
  u[9] = _mm_add_epi32(u[9], rnding);
  u[9] = _mm_srai_epi32(u[9], bit);

  u[10] = _mm_mullo_epi32(in[5], cospi42);
  u[10] = _mm_add_epi32(u[10], rnding);
  u[10] = _mm_srai_epi32(u[10], bit);

  u[11] = _mm_mullo_epi32(in[5], cospi22);
  u[11] = _mm_add_epi32(u[11], rnding);
  u[11] = _mm_srai_epi32(u[11], bit);

  u[12] = _mm_mullo_epi32(in[3], cospi50);
  u[12] = _mm_add_epi32(u[12], rnding);
  u[12] = _mm_srai_epi32(u[12], bit);

  u[13] = _mm_mullo_epi32(in[3], cospi14);
  u[13] = _mm_add_epi32(u[13], rnding);
  u[13] = _mm_srai_epi32(u[13], bit);

  u[14] = _mm_mullo_epi32(in[1], cospi58);
  u[14] = _mm_add_epi32(u[14], rnding);
  u[14] = _mm_srai_epi32(u[14], bit);

  u[15] = _mm_mullo_epi32(in[1], cospi6);
  u[15] = _mm_add_epi32(u[15], rnding);
  u[15] = _mm_srai_epi32(u[15], bit);

  // stage 3
  addsub_sse4_1(u[0], u[8], &u[0], &u[8], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[1], u[9], &u[1], &u[9], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[2], u[10], &u[2], &u[10], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[3], u[11], &u[3], &u[11], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[4], u[12], &u[4], &u[12], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[5], u[13], &u[5], &u[13], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[6], u[14], &u[6], &u[14], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[7], u[15], &u[7], &u[15], &clamp_lo, &clamp_hi);

  // stage 4
  y = _mm_mullo_epi32(u[8], cospi56);
  x = _mm_mullo_epi32(u[9], cospi56);
  u[8] = _mm_mullo_epi32(u[8], cospi8);
  u[8] = _mm_add_epi32(u[8], x);
  u[8] = _mm_add_epi32(u[8], rnding);
  u[8] = _mm_srai_epi32(u[8], bit);

  x = _mm_mullo_epi32(u[9], cospi8);
  u[9] = _mm_sub_epi32(y, x);
  u[9] = _mm_add_epi32(u[9], rnding);
  u[9] = _mm_srai_epi32(u[9], bit);

  x = _mm_mullo_epi32(u[11], cospi24);
  y = _mm_mullo_epi32(u[10], cospi24);
  u[10] = _mm_mullo_epi32(u[10], cospi40);
  u[10] = _mm_add_epi32(u[10], x);
  u[10] = _mm_add_epi32(u[10], rnding);
  u[10] = _mm_srai_epi32(u[10], bit);

  x = _mm_mullo_epi32(u[11], cospi40);
  u[11] = _mm_sub_epi32(y, x);
  u[11] = _mm_add_epi32(u[11], rnding);
  u[11] = _mm_srai_epi32(u[11], bit);

  x = _mm_mullo_epi32(u[13], cospi8);
  y = _mm_mullo_epi32(u[12], cospi8);
  u[12] = _mm_mullo_epi32(u[12], cospim56);
  u[12] = _mm_add_epi32(u[12], x);
  u[12] = _mm_add_epi32(u[12], rnding);
  u[12] = _mm_srai_epi32(u[12], bit);

  x = _mm_mullo_epi32(u[13], cospim56);
  u[13] = _mm_sub_epi32(y, x);
  u[13] = _mm_add_epi32(u[13], rnding);
  u[13] = _mm_srai_epi32(u[13], bit);

  x = _mm_mullo_epi32(u[15], cospi40);
  y = _mm_mullo_epi32(u[14], cospi40);
  u[14] = _mm_mullo_epi32(u[14], cospim24);
  u[14] = _mm_add_epi32(u[14], x);
  u[14] = _mm_add_epi32(u[14], rnding);
  u[14] = _mm_srai_epi32(u[14], bit);

  x = _mm_mullo_epi32(u[15], cospim24);
  u[15] = _mm_sub_epi32(y, x);
  u[15] = _mm_add_epi32(u[15], rnding);
  u[15] = _mm_srai_epi32(u[15], bit);

  // stage 5
  addsub_sse4_1(u[0], u[4], &u[0], &u[4], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[1], u[5], &u[1], &u[5], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[2], u[6], &u[2], &u[6], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[3], u[7], &u[3], &u[7], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[8], u[12], &u[8], &u[12], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[9], u[13], &u[9], &u[13], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[10], u[14], &u[10], &u[14], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[11], u[15], &u[11], &u[15], &clamp_lo, &clamp_hi);

  // stage 6
  x = _mm_mullo_epi32(u[5], cospi48);
  y = _mm_mullo_epi32(u[4], cospi48);
  u[4] = _mm_mullo_epi32(u[4], cospi16);
  u[4] = _mm_add_epi32(u[4], x);
  u[4] = _mm_add_epi32(u[4], rnding);
  u[4] = _mm_srai_epi32(u[4], bit);

  x = _mm_mullo_epi32(u[5], cospi16);
  u[5] = _mm_sub_epi32(y, x);
  u[5] = _mm_add_epi32(u[5], rnding);
  u[5] = _mm_srai_epi32(u[5], bit);

  x = _mm_mullo_epi32(u[7], cospi16);
  y = _mm_mullo_epi32(u[6], cospi16);
  u[6] = _mm_mullo_epi32(u[6], cospim48);
  u[6] = _mm_add_epi32(u[6], x);
  u[6] = _mm_add_epi32(u[6], rnding);
  u[6] = _mm_srai_epi32(u[6], bit);

  x = _mm_mullo_epi32(u[7], cospim48);
  u[7] = _mm_sub_epi32(y, x);
  u[7] = _mm_add_epi32(u[7], rnding);
  u[7] = _mm_srai_epi32(u[7], bit);

  x = _mm_mullo_epi32(u[13], cospi48);
  y = _mm_mullo_epi32(u[12], cospi48);
  u[12] = _mm_mullo_epi32(u[12], cospi16);
  u[12] = _mm_add_epi32(u[12], x);
  u[12] = _mm_add_epi32(u[12], rnding);
  u[12] = _mm_srai_epi32(u[12], bit);

  x = _mm_mullo_epi32(u[13], cospi16);
  u[13] = _mm_sub_epi32(y, x);
  u[13] = _mm_add_epi32(u[13], rnding);
  u[13] = _mm_srai_epi32(u[13], bit);

  x = _mm_mullo_epi32(u[15], cospi16);
  y = _mm_mullo_epi32(u[14], cospi16);
  u[14] = _mm_mullo_epi32(u[14], cospim48);
  u[14] = _mm_add_epi32(u[14], x);
  u[14] = _mm_add_epi32(u[14], rnding);
  u[14] = _mm_srai_epi32(u[14], bit);

  x = _mm_mullo_epi32(u[15], cospim48);
  u[15] = _mm_sub_epi32(y, x);
  u[15] = _mm_add_epi32(u[15], rnding);
  u[15] = _mm_srai_epi32(u[15], bit);

  // stage 7
  addsub_sse4_1(u[0], u[2], &u[0], &u[2], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[1], u[3], &u[1], &u[3], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[4], u[6], &u[4], &u[6], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[5], u[7], &u[5], &u[7], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[8], u[10], &u[8], &u[10], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[9], u[11], &u[9], &u[11], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[12], u[14], &u[12], &u[14], &clamp_lo, &clamp_hi);
  addsub_sse4_1(u[13], u[15], &u[13], &u[15], &clamp_lo, &clamp_hi);

  // stage 8
  y = _mm_mullo_epi32(u[2], cospi32);
  x = _mm_mullo_epi32(u[3], cospi32);
  u[2] = _mm_add_epi32(y, x);
  u[2] = _mm_add_epi32(u[2], rnding);
  u[2] = _mm_srai_epi32(u[2], bit);

  u[3] = _mm_sub_epi32(y, x);
  u[3] = _mm_add_epi32(u[3], rnding);
  u[3] = _mm_srai_epi32(u[3], bit);
  y = _mm_mullo_epi32(u[6], cospi32);
  x = _mm_mullo_epi32(u[7], cospi32);
  u[6] = _mm_add_epi32(y, x);
  u[6] = _mm_add_epi32(u[6], rnding);
  u[6] = _mm_srai_epi32(u[6], bit);

  u[7] = _mm_sub_epi32(y, x);
  u[7] = _mm_add_epi32(u[7], rnding);
  u[7] = _mm_srai_epi32(u[7], bit);

  y = _mm_mullo_epi32(u[10], cospi32);
  x = _mm_mullo_epi32(u[11], cospi32);
  u[10] = _mm_add_epi32(y, x);
  u[10] = _mm_add_epi32(u[10], rnding);
  u[10] = _mm_srai_epi32(u[10], bit);

  u[11] = _mm_sub_epi32(y, x);
  u[11] = _mm_add_epi32(u[11], rnding);
  u[11] = _mm_srai_epi32(u[11], bit);

  y = _mm_mullo_epi32(u[14], cospi32);
  x = _mm_mullo_epi32(u[15], cospi32);
  u[14] = _mm_add_epi32(y, x);
  u[14] = _mm_add_epi32(u[14], rnding);
  u[14] = _mm_srai_epi32(u[14], bit);

  u[15] = _mm_sub_epi32(y, x);
  u[15] = _mm_add_epi32(u[15], rnding);
  u[15] = _mm_srai_epi32(u[15], bit);

  // stage 9
  if (do_cols) {
    out[0] = u[0];
    out[1] = _mm_sub_epi32(zero, u[8]);
    out[2] = u[12];
    out[3] = _mm_sub_epi32(zero, u[4]);
    out[4] = u[6];
    out[5] = _mm_sub_epi32(zero, u[14]);
    out[6] = u[10];
    out[7] = _mm_sub_epi32(zero, u[2]);
    out[8] = u[3];
    out[9] = _mm_sub_epi32(zero, u[11]);
    out[10] = u[15];
    out[11] = _mm_sub_epi32(zero, u[7]);
    out[12] = u[5];
    out[13] = _mm_sub_epi32(zero, u[13]);
    out[14] = u[9];
    out[15] = _mm_sub_epi32(zero, u[1]);
  } else {
    const int log_range_out = AOMMAX(16, bd + 6);
    const __m128i clamp_lo_out = _mm_set1_epi32(-(1 << (log_range_out - 1)));
    const __m128i clamp_hi_out = _mm_set1_epi32((1 << (log_range_out - 1)) - 1);

    neg_shift_sse4_1(u[0], u[8], out + 0, out + 1, &clamp_lo_out, &clamp_hi_out,
                     out_shift);
    neg_shift_sse4_1(u[12], u[4], out + 2, out + 3, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
    neg_shift_sse4_1(u[6], u[14], out + 4, out + 5, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
    neg_shift_sse4_1(u[10], u[2], out + 6, out + 7, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
    neg_shift_sse4_1(u[3], u[11], out + 8, out + 9, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
    neg_shift_sse4_1(u[15], u[7], out + 10, out + 11, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
    neg_shift_sse4_1(u[5], u[13], out + 12, out + 13, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
    neg_shift_sse4_1(u[9], u[1], out + 14, out + 15, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
  }
}
#endif  // USE_TUNED_ADST16

static void idct16x16_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                             int bd, int out_shift) {
  const int32_t *cospi = cospi_arr(bit);
  const __m128i cospi60 = _mm_set1_epi32(cospi[60]);
  const __m128i cospim4 = _mm_set1_epi32(-cospi[4]);
  const __m128i cospi28 = _mm_set1_epi32(cospi[28]);
  const __m128i cospim36 = _mm_set1_epi32(-cospi[36]);
  const __m128i cospi44 = _mm_set1_epi32(cospi[44]);
  const __m128i cospi20 = _mm_set1_epi32(cospi[20]);
  const __m128i cospim20 = _mm_set1_epi32(-cospi[20]);
  const __m128i cospi12 = _mm_set1_epi32(cospi[12]);
  const __m128i cospim52 = _mm_set1_epi32(-cospi[52]);
  const __m128i cospi52 = _mm_set1_epi32(cospi[52]);
  const __m128i cospi36 = _mm_set1_epi32(cospi[36]);
  const __m128i cospi4 = _mm_set1_epi32(cospi[4]);
  const __m128i cospi56 = _mm_set1_epi32(cospi[56]);
  const __m128i cospim8 = _mm_set1_epi32(-cospi[8]);
  const __m128i cospi24 = _mm_set1_epi32(cospi[24]);
  const __m128i cospim40 = _mm_set1_epi32(-cospi[40]);
  const __m128i cospi40 = _mm_set1_epi32(cospi[40]);
  const __m128i cospi8 = _mm_set1_epi32(cospi[8]);
  const __m128i cospi32 = _mm_set1_epi32(cospi[32]);
  const __m128i cospi48 = _mm_set1_epi32(cospi[48]);
  const __m128i cospi16 = _mm_set1_epi32(cospi[16]);
  const __m128i cospim16 = _mm_set1_epi32(-cospi[16]);
  const __m128i cospim48 = _mm_set1_epi32(-cospi[48]);
  const __m128i rnding = _mm_set1_epi32(1 << (bit - 1));
  const int log_range = AOMMAX(16, bd + (do_cols ? 6 : 8));
  const __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
  const __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
  __m128i u[16], v[16], x, y;

  {
    // stage 0
    // stage 1
    u[0] = in[0];
    u[1] = in[8];
    u[2] = in[4];
    u[3] = in[12];
    u[4] = in[2];
    u[5] = in[10];
    u[6] = in[6];
    u[7] = in[14];
    u[8] = in[1];
    u[9] = in[9];
    u[10] = in[5];
    u[11] = in[13];
    u[12] = in[3];
    u[13] = in[11];
    u[14] = in[7];
    u[15] = in[15];

    // stage 2
    v[0] = u[0];
    v[1] = u[1];
    v[2] = u[2];
    v[3] = u[3];
    v[4] = u[4];
    v[5] = u[5];
    v[6] = u[6];
    v[7] = u[7];

    v[8] = half_btf_sse4_1(&cospi60, &u[8], &cospim4, &u[15], &rnding, bit);
    v[9] = half_btf_sse4_1(&cospi28, &u[9], &cospim36, &u[14], &rnding, bit);
    v[10] = half_btf_sse4_1(&cospi44, &u[10], &cospim20, &u[13], &rnding, bit);
    v[11] = half_btf_sse4_1(&cospi12, &u[11], &cospim52, &u[12], &rnding, bit);
    v[12] = half_btf_sse4_1(&cospi52, &u[11], &cospi12, &u[12], &rnding, bit);
    v[13] = half_btf_sse4_1(&cospi20, &u[10], &cospi44, &u[13], &rnding, bit);
    v[14] = half_btf_sse4_1(&cospi36, &u[9], &cospi28, &u[14], &rnding, bit);
    v[15] = half_btf_sse4_1(&cospi4, &u[8], &cospi60, &u[15], &rnding, bit);

    // stage 3
    u[0] = v[0];
    u[1] = v[1];
    u[2] = v[2];
    u[3] = v[3];
    u[4] = half_btf_sse4_1(&cospi56, &v[4], &cospim8, &v[7], &rnding, bit);
    u[5] = half_btf_sse4_1(&cospi24, &v[5], &cospim40, &v[6], &rnding, bit);
    u[6] = half_btf_sse4_1(&cospi40, &v[5], &cospi24, &v[6], &rnding, bit);
    u[7] = half_btf_sse4_1(&cospi8, &v[4], &cospi56, &v[7], &rnding, bit);
    addsub_sse4_1(v[8], v[9], &u[8], &u[9], &clamp_lo, &clamp_hi);
    addsub_sse4_1(v[11], v[10], &u[11], &u[10], &clamp_lo, &clamp_hi);
    addsub_sse4_1(v[12], v[13], &u[12], &u[13], &clamp_lo, &clamp_hi);
    addsub_sse4_1(v[15], v[14], &u[15], &u[14], &clamp_lo, &clamp_hi);

    // stage 4
    x = _mm_mullo_epi32(u[0], cospi32);
    y = _mm_mullo_epi32(u[1], cospi32);
    v[0] = _mm_add_epi32(x, y);
    v[0] = _mm_add_epi32(v[0], rnding);
    v[0] = _mm_srai_epi32(v[0], bit);

    v[1] = _mm_sub_epi32(x, y);
    v[1] = _mm_add_epi32(v[1], rnding);
    v[1] = _mm_srai_epi32(v[1], bit);

    v[2] = half_btf_sse4_1(&cospi48, &u[2], &cospim16, &u[3], &rnding, bit);
    v[3] = half_btf_sse4_1(&cospi16, &u[2], &cospi48, &u[3], &rnding, bit);
    addsub_sse4_1(u[4], u[5], &v[4], &v[5], &clamp_lo, &clamp_hi);
    addsub_sse4_1(u[7], u[6], &v[7], &v[6], &clamp_lo, &clamp_hi);
    v[8] = u[8];
    v[9] = half_btf_sse4_1(&cospim16, &u[9], &cospi48, &u[14], &rnding, bit);
    v[10] = half_btf_sse4_1(&cospim48, &u[10], &cospim16, &u[13], &rnding, bit);
    v[11] = u[11];
    v[12] = u[12];
    v[13] = half_btf_sse4_1(&cospim16, &u[10], &cospi48, &u[13], &rnding, bit);
    v[14] = half_btf_sse4_1(&cospi48, &u[9], &cospi16, &u[14], &rnding, bit);
    v[15] = u[15];

    // stage 5
    addsub_sse4_1(v[0], v[3], &u[0], &u[3], &clamp_lo, &clamp_hi);
    addsub_sse4_1(v[1], v[2], &u[1], &u[2], &clamp_lo, &clamp_hi);
    u[4] = v[4];

    x = _mm_mullo_epi32(v[5], cospi32);
    y = _mm_mullo_epi32(v[6], cospi32);
    u[5] = _mm_sub_epi32(y, x);
    u[5] = _mm_add_epi32(u[5], rnding);
    u[5] = _mm_srai_epi32(u[5], bit);

    u[6] = _mm_add_epi32(y, x);
    u[6] = _mm_add_epi32(u[6], rnding);
    u[6] = _mm_srai_epi32(u[6], bit);

    u[7] = v[7];
    addsub_sse4_1(v[8], v[11], &u[8], &u[11], &clamp_lo, &clamp_hi);
    addsub_sse4_1(v[9], v[10], &u[9], &u[10], &clamp_lo, &clamp_hi);
    addsub_sse4_1(v[15], v[12], &u[15], &u[12], &clamp_lo, &clamp_hi);
    addsub_sse4_1(v[14], v[13], &u[14], &u[13], &clamp_lo, &clamp_hi);

    // stage 6
    addsub_sse4_1(u[0], u[7], &v[0], &v[7], &clamp_lo, &clamp_hi);
    addsub_sse4_1(u[1], u[6], &v[1], &v[6], &clamp_lo, &clamp_hi);
    addsub_sse4_1(u[2], u[5], &v[2], &v[5], &clamp_lo, &clamp_hi);
    addsub_sse4_1(u[3], u[4], &v[3], &v[4], &clamp_lo, &clamp_hi);
    v[8] = u[8];
    v[9] = u[9];

    x = _mm_mullo_epi32(u[10], cospi32);
    y = _mm_mullo_epi32(u[13], cospi32);
    v[10] = _mm_sub_epi32(y, x);
    v[10] = _mm_add_epi32(v[10], rnding);
    v[10] = _mm_srai_epi32(v[10], bit);

    v[13] = _mm_add_epi32(x, y);
    v[13] = _mm_add_epi32(v[13], rnding);
    v[13] = _mm_srai_epi32(v[13], bit);

    x = _mm_mullo_epi32(u[11], cospi32);
    y = _mm_mullo_epi32(u[12], cospi32);
    v[11] = _mm_sub_epi32(y, x);
    v[11] = _mm_add_epi32(v[11], rnding);
    v[11] = _mm_srai_epi32(v[11], bit);

    v[12] = _mm_add_epi32(x, y);
    v[12] = _mm_add_epi32(v[12], rnding);
    v[12] = _mm_srai_epi32(v[12], bit);

    v[14] = u[14];
    v[15] = u[15];

    // stage 7
    addsub_sse4_1(v[0], v[15], out + 0, out + 15, &clamp_lo, &clamp_hi);
    addsub_sse4_1(v[1], v[14], out + 1, out + 14, &clamp_lo, &clamp_hi);
    addsub_sse4_1(v[2], v[13], out + 2, out + 13, &clamp_lo, &clamp_hi);
    addsub_sse4_1(v[3], v[12], out + 3, out + 12, &clamp_lo, &clamp_hi);
    addsub_sse4_1(v[4], v[11], out + 4, out + 11, &clamp_lo, &clamp_hi);
    addsub_sse4_1(v[5], v[10], out + 5, out + 10, &clamp_lo, &clamp_hi);
    addsub_sse4_1(v[6], v[9], out + 6, out + 9, &clamp_lo, &clamp_hi);
    addsub_sse4_1(v[7], v[8], out + 7, out + 8, &clamp_lo, &clamp_hi);

    if (!do_cols) {
      const int log_range_out = AOMMAX(16, bd + 6);
      assert(log_range_out - 1 < 32);  // Ensure left-shift doesn't overflow.
      const __m128i clamp_lo_out = _mm_set1_epi32(-(1 << (log_range_out - 1)));
      const __m128i clamp_hi_out =
          _mm_set1_epi32((1 << (log_range_out - 1)) - 1);
      round_shift_8x8(out, out_shift);
      highbd_clamp_epi32_sse4_1(out, out, &clamp_lo_out, &clamp_hi_out, 16);
    }
  }
}

static void iddt16x16_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                             int bd, int out_shift) {
  (void)bit;
  iadst_matrix_mult_sse4(in, out, INV_DDT_BIT, do_cols, bd, out_shift,
                         ddt16_kernel[INV_TXFM], 16, 16, 1);
}

#if USE_TUNED_ADST16
static void iadst16x16_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                              int bd, int out_shift) {
  (void)bit;
  iadst_matrix_mult_sse4(in, out, INV_ADST_BIT, do_cols, bd, out_shift,
                         av2_adst_kernel16[INV_TXFM], tx_size_wide[TX_16X16],
                         tx_size_wide[TX_16X16], 1);
}
#else
static void iadst16x16_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                              int bd, int out_shift) {
  const int32_t *cospi = cospi_arr(bit);
  const __m128i cospi2 = _mm_set1_epi32(cospi[2]);
  const __m128i cospi62 = _mm_set1_epi32(cospi[62]);
  const __m128i cospi10 = _mm_set1_epi32(cospi[10]);
  const __m128i cospi54 = _mm_set1_epi32(cospi[54]);
  const __m128i cospi18 = _mm_set1_epi32(cospi[18]);
  const __m128i cospi46 = _mm_set1_epi32(cospi[46]);
  const __m128i cospi26 = _mm_set1_epi32(cospi[26]);
  const __m128i cospi38 = _mm_set1_epi32(cospi[38]);
  const __m128i cospi34 = _mm_set1_epi32(cospi[34]);
  const __m128i cospi30 = _mm_set1_epi32(cospi[30]);
  const __m128i cospi42 = _mm_set1_epi32(cospi[42]);
  const __m128i cospi22 = _mm_set1_epi32(cospi[22]);
  const __m128i cospi50 = _mm_set1_epi32(cospi[50]);
  const __m128i cospi14 = _mm_set1_epi32(cospi[14]);
  const __m128i cospi58 = _mm_set1_epi32(cospi[58]);
  const __m128i cospi6 = _mm_set1_epi32(cospi[6]);
  const __m128i cospi8 = _mm_set1_epi32(cospi[8]);
  const __m128i cospi56 = _mm_set1_epi32(cospi[56]);
  const __m128i cospi40 = _mm_set1_epi32(cospi[40]);
  const __m128i cospi24 = _mm_set1_epi32(cospi[24]);
  const __m128i cospim56 = _mm_set1_epi32(-cospi[56]);
  const __m128i cospim24 = _mm_set1_epi32(-cospi[24]);
  const __m128i cospi48 = _mm_set1_epi32(cospi[48]);
  const __m128i cospi16 = _mm_set1_epi32(cospi[16]);
  const __m128i cospim48 = _mm_set1_epi32(-cospi[48]);
  const __m128i cospi32 = _mm_set1_epi32(cospi[32]);
  const __m128i rnding = _mm_set1_epi32(1 << (bit - 1));
  const int log_range = AOMMAX(16, bd + (do_cols ? 6 : 8));
  const __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
  const __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
  const __m128i zero = _mm_setzero_si128();
  __m128i u[16], v[16], x, y;
  // Calculate the column 0, 1, 2, 3
  // stage 0
  // stage 1
  // stage 2
  v[0] = _mm_mullo_epi32(in[15], cospi2);
  x = _mm_mullo_epi32(in[0], cospi62);
  v[0] = _mm_add_epi32(v[0], x);
  v[0] = _mm_add_epi32(v[0], rnding);
  v[0] = _mm_srai_epi32(v[0], bit);

  v[1] = _mm_mullo_epi32(in[15], cospi62);
  x = _mm_mullo_epi32(in[0], cospi2);
  v[1] = _mm_sub_epi32(v[1], x);
  v[1] = _mm_add_epi32(v[1], rnding);
  v[1] = _mm_srai_epi32(v[1], bit);

  v[2] = _mm_mullo_epi32(in[13], cospi10);
  x = _mm_mullo_epi32(in[2], cospi54);
  v[2] = _mm_add_epi32(v[2], x);
  v[2] = _mm_add_epi32(v[2], rnding);
  v[2] = _mm_srai_epi32(v[2], bit);

  v[3] = _mm_mullo_epi32(in[13], cospi54);
  x = _mm_mullo_epi32(in[2], cospi10);
  v[3] = _mm_sub_epi32(v[3], x);
  v[3] = _mm_add_epi32(v[3], rnding);
  v[3] = _mm_srai_epi32(v[3], bit);

  v[4] = _mm_mullo_epi32(in[11], cospi18);
  x = _mm_mullo_epi32(in[4], cospi46);
  v[4] = _mm_add_epi32(v[4], x);
  v[4] = _mm_add_epi32(v[4], rnding);
  v[4] = _mm_srai_epi32(v[4], bit);

  v[5] = _mm_mullo_epi32(in[11], cospi46);
  x = _mm_mullo_epi32(in[4], cospi18);
  v[5] = _mm_sub_epi32(v[5], x);
  v[5] = _mm_add_epi32(v[5], rnding);
  v[5] = _mm_srai_epi32(v[5], bit);

  v[6] = _mm_mullo_epi32(in[9], cospi26);
  x = _mm_mullo_epi32(in[6], cospi38);
  v[6] = _mm_add_epi32(v[6], x);
  v[6] = _mm_add_epi32(v[6], rnding);
  v[6] = _mm_srai_epi32(v[6], bit);

  v[7] = _mm_mullo_epi32(in[9], cospi38);
  x = _mm_mullo_epi32(in[6], cospi26);
  v[7] = _mm_sub_epi32(v[7], x);
  v[7] = _mm_add_epi32(v[7], rnding);
  v[7] = _mm_srai_epi32(v[7], bit);

  v[8] = _mm_mullo_epi32(in[7], cospi34);
  x = _mm_mullo_epi32(in[8], cospi30);
  v[8] = _mm_add_epi32(v[8], x);
  v[8] = _mm_add_epi32(v[8], rnding);
  v[8] = _mm_srai_epi32(v[8], bit);

  v[9] = _mm_mullo_epi32(in[7], cospi30);
  x = _mm_mullo_epi32(in[8], cospi34);
  v[9] = _mm_sub_epi32(v[9], x);
  v[9] = _mm_add_epi32(v[9], rnding);
  v[9] = _mm_srai_epi32(v[9], bit);

  v[10] = _mm_mullo_epi32(in[5], cospi42);
  x = _mm_mullo_epi32(in[10], cospi22);
  v[10] = _mm_add_epi32(v[10], x);
  v[10] = _mm_add_epi32(v[10], rnding);
  v[10] = _mm_srai_epi32(v[10], bit);

  v[11] = _mm_mullo_epi32(in[5], cospi22);
  x = _mm_mullo_epi32(in[10], cospi42);
  v[11] = _mm_sub_epi32(v[11], x);
  v[11] = _mm_add_epi32(v[11], rnding);
  v[11] = _mm_srai_epi32(v[11], bit);

  v[12] = _mm_mullo_epi32(in[3], cospi50);
  x = _mm_mullo_epi32(in[12], cospi14);
  v[12] = _mm_add_epi32(v[12], x);
  v[12] = _mm_add_epi32(v[12], rnding);
  v[12] = _mm_srai_epi32(v[12], bit);

  v[13] = _mm_mullo_epi32(in[3], cospi14);
  x = _mm_mullo_epi32(in[12], cospi50);
  v[13] = _mm_sub_epi32(v[13], x);
  v[13] = _mm_add_epi32(v[13], rnding);
  v[13] = _mm_srai_epi32(v[13], bit);

  v[14] = _mm_mullo_epi32(in[1], cospi58);
  x = _mm_mullo_epi32(in[14], cospi6);
  v[14] = _mm_add_epi32(v[14], x);
  v[14] = _mm_add_epi32(v[14], rnding);
  v[14] = _mm_srai_epi32(v[14], bit);

  v[15] = _mm_mullo_epi32(in[1], cospi6);
  x = _mm_mullo_epi32(in[14], cospi58);
  v[15] = _mm_sub_epi32(v[15], x);
  v[15] = _mm_add_epi32(v[15], rnding);
  v[15] = _mm_srai_epi32(v[15], bit);

  // stage 3
  addsub_sse4_1(v[0], v[8], &u[0], &u[8], &clamp_lo, &clamp_hi);
  addsub_sse4_1(v[1], v[9], &u[1], &u[9], &clamp_lo, &clamp_hi);
  addsub_sse4_1(v[2], v[10], &u[2], &u[10], &clamp_lo, &clamp_hi);
  addsub_sse4_1(v[3], v[11], &u[3], &u[11], &clamp_lo, &clamp_hi);
  addsub_sse4_1(v[4], v[12], &u[4], &u[12], &clamp_lo, &clamp_hi);
  addsub_sse4_1(v[5], v[13], &u[5], &u[13], &clamp_lo, &clamp_hi);
  addsub_sse4_1(v[6], v[14], &u[6], &u[14], &clamp_lo, &clamp_hi);
  addsub_sse4_1(v[7], v[15], &u[7], &u[15], &clamp_lo, &clamp_hi);

  // stage 4
  v[0] = u[0];
  v[1] = u[1];
  v[2] = u[2];
  v[3] = u[3];
  v[4] = u[4];
  v[5] = u[5];
  v[6] = u[6];
  v[7] = u[7];

  v[8] = _mm_mullo_epi32(u[8], cospi8);
  x = _mm_mullo_epi32(u[9], cospi56);
  v[8] = _mm_add_epi32(v[8], x);
  v[8] = _mm_add_epi32(v[8], rnding);
  v[8] = _mm_srai_epi32(v[8], bit);

  v[9] = _mm_mullo_epi32(u[8], cospi56);
  x = _mm_mullo_epi32(u[9], cospi8);
  v[9] = _mm_sub_epi32(v[9], x);
  v[9] = _mm_add_epi32(v[9], rnding);
  v[9] = _mm_srai_epi32(v[9], bit);

  v[10] = _mm_mullo_epi32(u[10], cospi40);
  x = _mm_mullo_epi32(u[11], cospi24);
  v[10] = _mm_add_epi32(v[10], x);
  v[10] = _mm_add_epi32(v[10], rnding);
  v[10] = _mm_srai_epi32(v[10], bit);

  v[11] = _mm_mullo_epi32(u[10], cospi24);
  x = _mm_mullo_epi32(u[11], cospi40);
  v[11] = _mm_sub_epi32(v[11], x);
  v[11] = _mm_add_epi32(v[11], rnding);
  v[11] = _mm_srai_epi32(v[11], bit);

  v[12] = _mm_mullo_epi32(u[12], cospim56);
  x = _mm_mullo_epi32(u[13], cospi8);
  v[12] = _mm_add_epi32(v[12], x);
  v[12] = _mm_add_epi32(v[12], rnding);
  v[12] = _mm_srai_epi32(v[12], bit);

  v[13] = _mm_mullo_epi32(u[12], cospi8);
  x = _mm_mullo_epi32(u[13], cospim56);
  v[13] = _mm_sub_epi32(v[13], x);
  v[13] = _mm_add_epi32(v[13], rnding);
  v[13] = _mm_srai_epi32(v[13], bit);

  v[14] = _mm_mullo_epi32(u[14], cospim24);
  x = _mm_mullo_epi32(u[15], cospi40);
  v[14] = _mm_add_epi32(v[14], x);
  v[14] = _mm_add_epi32(v[14], rnding);
  v[14] = _mm_srai_epi32(v[14], bit);

  v[15] = _mm_mullo_epi32(u[14], cospi40);
  x = _mm_mullo_epi32(u[15], cospim24);
  v[15] = _mm_sub_epi32(v[15], x);
  v[15] = _mm_add_epi32(v[15], rnding);
  v[15] = _mm_srai_epi32(v[15], bit);

  // stage 5
  addsub_sse4_1(v[0], v[4], &u[0], &u[4], &clamp_lo, &clamp_hi);
  addsub_sse4_1(v[1], v[5], &u[1], &u[5], &clamp_lo, &clamp_hi);
  addsub_sse4_1(v[2], v[6], &u[2], &u[6], &clamp_lo, &clamp_hi);
  addsub_sse4_1(v[3], v[7], &u[3], &u[7], &clamp_lo, &clamp_hi);
  addsub_sse4_1(v[8], v[12], &u[8], &u[12], &clamp_lo, &clamp_hi);
  addsub_sse4_1(v[9], v[13], &u[9], &u[13], &clamp_lo, &clamp_hi);
  addsub_sse4_1(v[10], v[14], &u[10], &u[14], &clamp_lo, &clamp_hi);
  addsub_sse4_1(v[11], v[15], &u[11], &u[15], &clamp_lo, &clamp_hi);

  // stage 6
  v[0] = u[0];
  v[1] = u[1];
  v[2] = u[2];
  v[3] = u[3];

  v[4] = _mm_mullo_epi32(u[4], cospi16);
  x = _mm_mullo_epi32(u[5], cospi48);
  v[4] = _mm_add_epi32(v[4], x);
  v[4] = _mm_add_epi32(v[4], rnding);
  v[4] = _mm_srai_epi32(v[4], bit);

  v[5] = _mm_mullo_epi32(u[4], cospi48);
  x = _mm_mullo_epi32(u[5], cospi16);
  v[5] = _mm_sub_epi32(v[5], x);
  v[5] = _mm_add_epi32(v[5], rnding);
  v[5] = _mm_srai_epi32(v[5], bit);

  v[6] = _mm_mullo_epi32(u[6], cospim48);
  x = _mm_mullo_epi32(u[7], cospi16);
  v[6] = _mm_add_epi32(v[6], x);
  v[6] = _mm_add_epi32(v[6], rnding);
  v[6] = _mm_srai_epi32(v[6], bit);

  v[7] = _mm_mullo_epi32(u[6], cospi16);
  x = _mm_mullo_epi32(u[7], cospim48);
  v[7] = _mm_sub_epi32(v[7], x);
  v[7] = _mm_add_epi32(v[7], rnding);
  v[7] = _mm_srai_epi32(v[7], bit);

  v[8] = u[8];
  v[9] = u[9];
  v[10] = u[10];
  v[11] = u[11];

  v[12] = _mm_mullo_epi32(u[12], cospi16);
  x = _mm_mullo_epi32(u[13], cospi48);
  v[12] = _mm_add_epi32(v[12], x);
  v[12] = _mm_add_epi32(v[12], rnding);
  v[12] = _mm_srai_epi32(v[12], bit);

  v[13] = _mm_mullo_epi32(u[12], cospi48);
  x = _mm_mullo_epi32(u[13], cospi16);
  v[13] = _mm_sub_epi32(v[13], x);
  v[13] = _mm_add_epi32(v[13], rnding);
  v[13] = _mm_srai_epi32(v[13], bit);

  v[14] = _mm_mullo_epi32(u[14], cospim48);
  x = _mm_mullo_epi32(u[15], cospi16);
  v[14] = _mm_add_epi32(v[14], x);
  v[14] = _mm_add_epi32(v[14], rnding);
  v[14] = _mm_srai_epi32(v[14], bit);

  v[15] = _mm_mullo_epi32(u[14], cospi16);
  x = _mm_mullo_epi32(u[15], cospim48);
  v[15] = _mm_sub_epi32(v[15], x);
  v[15] = _mm_add_epi32(v[15], rnding);
  v[15] = _mm_srai_epi32(v[15], bit);

  // stage 7
  addsub_sse4_1(v[0], v[2], &u[0], &u[2], &clamp_lo, &clamp_hi);
  addsub_sse4_1(v[1], v[3], &u[1], &u[3], &clamp_lo, &clamp_hi);
  addsub_sse4_1(v[4], v[6], &u[4], &u[6], &clamp_lo, &clamp_hi);
  addsub_sse4_1(v[5], v[7], &u[5], &u[7], &clamp_lo, &clamp_hi);
  addsub_sse4_1(v[8], v[10], &u[8], &u[10], &clamp_lo, &clamp_hi);
  addsub_sse4_1(v[9], v[11], &u[9], &u[11], &clamp_lo, &clamp_hi);
  addsub_sse4_1(v[12], v[14], &u[12], &u[14], &clamp_lo, &clamp_hi);
  addsub_sse4_1(v[13], v[15], &u[13], &u[15], &clamp_lo, &clamp_hi);

  // stage 8
  v[0] = u[0];
  v[1] = u[1];

  y = _mm_mullo_epi32(u[2], cospi32);
  x = _mm_mullo_epi32(u[3], cospi32);
  v[2] = _mm_add_epi32(y, x);
  v[2] = _mm_add_epi32(v[2], rnding);
  v[2] = _mm_srai_epi32(v[2], bit);

  v[3] = _mm_sub_epi32(y, x);
  v[3] = _mm_add_epi32(v[3], rnding);
  v[3] = _mm_srai_epi32(v[3], bit);

  v[4] = u[4];
  v[5] = u[5];

  y = _mm_mullo_epi32(u[6], cospi32);
  x = _mm_mullo_epi32(u[7], cospi32);
  v[6] = _mm_add_epi32(y, x);
  v[6] = _mm_add_epi32(v[6], rnding);
  v[6] = _mm_srai_epi32(v[6], bit);

  v[7] = _mm_sub_epi32(y, x);
  v[7] = _mm_add_epi32(v[7], rnding);
  v[7] = _mm_srai_epi32(v[7], bit);

  v[8] = u[8];
  v[9] = u[9];

  y = _mm_mullo_epi32(u[10], cospi32);
  x = _mm_mullo_epi32(u[11], cospi32);
  v[10] = _mm_add_epi32(y, x);
  v[10] = _mm_add_epi32(v[10], rnding);
  v[10] = _mm_srai_epi32(v[10], bit);

  v[11] = _mm_sub_epi32(y, x);
  v[11] = _mm_add_epi32(v[11], rnding);
  v[11] = _mm_srai_epi32(v[11], bit);

  v[12] = u[12];
  v[13] = u[13];

  y = _mm_mullo_epi32(u[14], cospi32);
  x = _mm_mullo_epi32(u[15], cospi32);
  v[14] = _mm_add_epi32(y, x);
  v[14] = _mm_add_epi32(v[14], rnding);
  v[14] = _mm_srai_epi32(v[14], bit);

  v[15] = _mm_sub_epi32(y, x);
  v[15] = _mm_add_epi32(v[15], rnding);
  v[15] = _mm_srai_epi32(v[15], bit);

  // stage 9
  if (do_cols) {
    out[0] = v[0];
    out[1] = _mm_sub_epi32(zero, v[8]);
    out[2] = v[12];
    out[3] = _mm_sub_epi32(zero, v[4]);
    out[4] = v[6];
    out[5] = _mm_sub_epi32(zero, v[14]);
    out[6] = v[10];
    out[7] = _mm_sub_epi32(zero, v[2]);
    out[8] = v[3];
    out[9] = _mm_sub_epi32(zero, v[11]);
    out[10] = v[15];
    out[11] = _mm_sub_epi32(zero, v[7]);
    out[12] = v[5];
    out[13] = _mm_sub_epi32(zero, v[13]);
    out[14] = v[9];
    out[15] = _mm_sub_epi32(zero, v[1]);
  } else {
    const int log_range_out = AOMMAX(16, bd + 6);
    const __m128i clamp_lo_out = _mm_set1_epi32(-(1 << (log_range_out - 1)));
    const __m128i clamp_hi_out = _mm_set1_epi32((1 << (log_range_out - 1)) - 1);

    neg_shift_sse4_1(v[0], v[8], out + 0, out + 1, &clamp_lo_out, &clamp_hi_out,
                     out_shift);
    neg_shift_sse4_1(v[12], v[4], out + 2, out + 3, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
    neg_shift_sse4_1(v[6], v[14], out + 4, out + 5, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
    neg_shift_sse4_1(v[10], v[2], out + 6, out + 7, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
    neg_shift_sse4_1(v[3], v[11], out + 8, out + 9, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
    neg_shift_sse4_1(v[15], v[7], out + 10, out + 11, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
    neg_shift_sse4_1(v[5], v[13], out + 12, out + 13, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
    neg_shift_sse4_1(v[9], v[1], out + 14, out + 15, &clamp_lo_out,
                     &clamp_hi_out, out_shift);
  }
}
#endif  // USE_TUNED_ADST16

static void iidentity16_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                               int bd, int out_shift) {
  (void)bit;
  __m128i fact = _mm_set1_epi32(2 * NewSqrt2);
  __m128i offset = _mm_set1_epi32(1 << (NewSqrt2Bits - 1));
  __m128i a0_low, a0_high, a1_low, a1_high;
  __m128i zero = _mm_set1_epi32(0);
  offset = _mm_unpacklo_epi32(offset, zero);

  for (int i = 0; i < 16; i++) {
    a0_low = _mm_mul_epi32(in[i], fact);
    a0_low = _mm_add_epi32(a0_low, offset);
    a0_low = _mm_srli_epi64(a0_low, NewSqrt2Bits);

    a0_high = _mm_srli_si128(in[i], 4);
    a0_high = _mm_mul_epi32(a0_high, fact);
    a0_high = _mm_add_epi32(a0_high, offset);
    a0_high = _mm_srli_epi64(a0_high, NewSqrt2Bits);

    a1_low = _mm_unpacklo_epi32(a0_low, a0_high);
    a1_high = _mm_unpackhi_epi32(a0_low, a0_high);
    out[i] = _mm_unpacklo_epi64(a1_low, a1_high);
  }

  if (!do_cols) {
    const int log_range = AOMMAX(16, bd + 6);
    const __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
    const __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
    round_shift_8x8(out, out_shift);
    highbd_clamp_epi32_sse4_1(out, out, &clamp_lo, &clamp_hi, 16);
  }
}
static INLINE void idct64_stage8_sse4_1(
    __m128i *u, const __m128i *cospim32, const __m128i *cospi32,
    const __m128i *cospim16, const __m128i *cospi48, const __m128i *cospi16,
    const __m128i *cospim48, const __m128i *clamp_lo, const __m128i *clamp_hi,
    const __m128i *rnding, int bit) {
  int i;
  __m128i temp1, temp2, temp3, temp4;
  temp1 = half_btf_sse4_1(cospim32, &u[10], cospi32, &u[13], rnding, bit);
  u[13] = half_btf_sse4_1(cospi32, &u[10], cospi32, &u[13], rnding, bit);
  u[10] = temp1;
  temp2 = half_btf_sse4_1(cospim32, &u[11], cospi32, &u[12], rnding, bit);
  u[12] = half_btf_sse4_1(cospi32, &u[11], cospi32, &u[12], rnding, bit);
  u[11] = temp2;

  for (i = 16; i < 20; ++i) {
    addsub_sse4_1(u[i], u[i ^ 7], &u[i], &u[i ^ 7], clamp_lo, clamp_hi);
    addsub_sse4_1(u[i ^ 15], u[i ^ 8], &u[i ^ 15], &u[i ^ 8], clamp_lo,
                  clamp_hi);
  }

  temp1 = half_btf_sse4_1(cospim16, &u[36], cospi48, &u[59], rnding, bit);
  temp2 = half_btf_sse4_1(cospim16, &u[37], cospi48, &u[58], rnding, bit);
  temp3 = half_btf_sse4_1(cospim16, &u[38], cospi48, &u[57], rnding, bit);
  temp4 = half_btf_sse4_1(cospim16, &u[39], cospi48, &u[56], rnding, bit);
  u[56] = half_btf_sse4_1(cospi48, &u[39], cospi16, &u[56], rnding, bit);
  u[57] = half_btf_sse4_1(cospi48, &u[38], cospi16, &u[57], rnding, bit);
  u[58] = half_btf_sse4_1(cospi48, &u[37], cospi16, &u[58], rnding, bit);
  u[59] = half_btf_sse4_1(cospi48, &u[36], cospi16, &u[59], rnding, bit);
  u[36] = temp1;
  u[37] = temp2;
  u[38] = temp3;
  u[39] = temp4;

  temp1 = half_btf_sse4_1(cospim48, &u[40], cospim16, &u[55], rnding, bit);
  temp2 = half_btf_sse4_1(cospim48, &u[41], cospim16, &u[54], rnding, bit);
  temp3 = half_btf_sse4_1(cospim48, &u[42], cospim16, &u[53], rnding, bit);
  temp4 = half_btf_sse4_1(cospim48, &u[43], cospim16, &u[52], rnding, bit);
  u[52] = half_btf_sse4_1(cospim16, &u[43], cospi48, &u[52], rnding, bit);
  u[53] = half_btf_sse4_1(cospim16, &u[42], cospi48, &u[53], rnding, bit);
  u[54] = half_btf_sse4_1(cospim16, &u[41], cospi48, &u[54], rnding, bit);
  u[55] = half_btf_sse4_1(cospim16, &u[40], cospi48, &u[55], rnding, bit);
  u[40] = temp1;
  u[41] = temp2;
  u[42] = temp3;
  u[43] = temp4;
}

static INLINE void idct64_stage9_sse4_1(__m128i *u, const __m128i *cospim32,
                                        const __m128i *cospi32,
                                        const __m128i *clamp_lo,
                                        const __m128i *clamp_hi,
                                        const __m128i *rnding, int bit) {
  int i;
  __m128i temp1, temp2, temp3, temp4;
  for (i = 0; i < 8; ++i) {
    addsub_sse4_1(u[i], u[15 - i], &u[i], &u[15 - i], clamp_lo, clamp_hi);
  }

  temp1 = half_btf_sse4_1(cospim32, &u[20], cospi32, &u[27], rnding, bit);
  temp2 = half_btf_sse4_1(cospim32, &u[21], cospi32, &u[26], rnding, bit);
  temp3 = half_btf_sse4_1(cospim32, &u[22], cospi32, &u[25], rnding, bit);
  temp4 = half_btf_sse4_1(cospim32, &u[23], cospi32, &u[24], rnding, bit);
  u[24] = half_btf_sse4_1(cospi32, &u[23], cospi32, &u[24], rnding, bit);
  u[25] = half_btf_sse4_1(cospi32, &u[22], cospi32, &u[25], rnding, bit);
  u[26] = half_btf_sse4_1(cospi32, &u[21], cospi32, &u[26], rnding, bit);
  u[27] = half_btf_sse4_1(cospi32, &u[20], cospi32, &u[27], rnding, bit);
  u[20] = temp1;
  u[21] = temp2;
  u[22] = temp3;
  u[23] = temp4;
  for (i = 32; i < 40; i++) {
    addsub_sse4_1(u[i], u[i ^ 15], &u[i], &u[i ^ 15], clamp_lo, clamp_hi);
  }

  for (i = 48; i < 56; i++) {
    addsub_sse4_1(u[i ^ 15], u[i], &u[i ^ 15], &u[i], clamp_lo, clamp_hi);
  }
}

static INLINE void idct64_stage10_sse4_1(__m128i *u, const __m128i *cospim32,
                                         const __m128i *cospi32,
                                         const __m128i *clamp_lo,
                                         const __m128i *clamp_hi,
                                         const __m128i *rnding, int bit) {
  __m128i temp1, temp2, temp3, temp4;
  for (int i = 0; i < 16; i++) {
    addsub_sse4_1(u[i], u[31 - i], &u[i], &u[31 - i], clamp_lo, clamp_hi);
  }

  temp1 = half_btf_sse4_1(cospim32, &u[40], cospi32, &u[55], rnding, bit);
  temp2 = half_btf_sse4_1(cospim32, &u[41], cospi32, &u[54], rnding, bit);
  temp3 = half_btf_sse4_1(cospim32, &u[42], cospi32, &u[53], rnding, bit);
  temp4 = half_btf_sse4_1(cospim32, &u[43], cospi32, &u[52], rnding, bit);
  u[52] = half_btf_sse4_1(cospi32, &u[43], cospi32, &u[52], rnding, bit);
  u[53] = half_btf_sse4_1(cospi32, &u[42], cospi32, &u[53], rnding, bit);
  u[54] = half_btf_sse4_1(cospi32, &u[41], cospi32, &u[54], rnding, bit);
  u[55] = half_btf_sse4_1(cospi32, &u[40], cospi32, &u[55], rnding, bit);
  u[40] = temp1;
  u[41] = temp2;
  u[42] = temp3;
  u[43] = temp4;

  temp1 = half_btf_sse4_1(cospim32, &u[44], cospi32, &u[51], rnding, bit);
  temp2 = half_btf_sse4_1(cospim32, &u[45], cospi32, &u[50], rnding, bit);
  temp3 = half_btf_sse4_1(cospim32, &u[46], cospi32, &u[49], rnding, bit);
  temp4 = half_btf_sse4_1(cospim32, &u[47], cospi32, &u[48], rnding, bit);
  u[48] = half_btf_sse4_1(cospi32, &u[47], cospi32, &u[48], rnding, bit);
  u[49] = half_btf_sse4_1(cospi32, &u[46], cospi32, &u[49], rnding, bit);
  u[50] = half_btf_sse4_1(cospi32, &u[45], cospi32, &u[50], rnding, bit);
  u[51] = half_btf_sse4_1(cospi32, &u[44], cospi32, &u[51], rnding, bit);
  u[44] = temp1;
  u[45] = temp2;
  u[46] = temp3;
  u[47] = temp4;
}

static INLINE void idct64_stage11_sse4_1(__m128i *u, __m128i *out, int do_cols,
                                         int bd, int out_shift,
                                         const __m128i *clamp_lo,
                                         const __m128i *clamp_hi) {
  for (int i = 0; i < 32; i++) {
    addsub_sse4_1(u[i], u[63 - i], out + i, out + 63 - i, clamp_lo, clamp_hi);
  }

  if (!do_cols) {
    const int log_range_out = AOMMAX(16, bd + 6);
    const __m128i clamp_lo_out = _mm_set1_epi32(-(1 << (log_range_out - 1)));
    const __m128i clamp_hi_out = _mm_set1_epi32((1 << (log_range_out - 1)) - 1);

    for (int i = 0; i < 64; i += 4) {
      round_shift_4x4(out + i, out_shift);
      highbd_clamp_epi32_sse4_1(out + i, out + i, &clamp_lo_out, &clamp_hi_out,
                                4);
    }
  }
}

static void idct64x64_low1_sse4_1(__m128i *in, __m128i *out, int bit,
                                  int do_cols, int bd, int out_shift) {
  const int32_t *cospi = cospi_arr(bit);
  const __m128i rnding = _mm_set1_epi32(1 << (bit - 1));
  const int log_range = AOMMAX(16, bd + (do_cols ? 6 : 8));
  __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
  __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);

  const __m128i cospi32 = _mm_set1_epi32(cospi[32]);

  {
    __m128i x;

    // stage 1
    // stage 2
    // stage 3
    // stage 4
    // stage 5
    // stage 6
    x = half_btf_0_sse4_1(&cospi32, &in[0], &rnding, bit);

    // stage 8
    // stage 9
    // stage 10
    // stage 11
    if (!do_cols) {
      const int log_range_out = AOMMAX(16, bd + 6);
      assert(log_range_out - 1 < 32);  // Ensure left-shift doesn't overflow.
      clamp_lo = _mm_set1_epi32(-(1 << (log_range_out - 1)));
      clamp_hi = _mm_set1_epi32((1 << (log_range_out - 1)) - 1);
      if (out_shift != 0) {
        __m128i offset = _mm_set1_epi32((1 << out_shift) >> 1);
        x = _mm_add_epi32(x, offset);
        x = _mm_sra_epi32(x, _mm_cvtsi32_si128(out_shift));
      }
    }
    x = _mm_max_epi32(x, clamp_lo);
    x = _mm_min_epi32(x, clamp_hi);
    out[0] = x;
    out[1] = x;
    out[2] = x;
    out[3] = x;
    out[4] = x;
    out[5] = x;
    out[6] = x;
    out[7] = x;
    out[8] = x;
    out[9] = x;
    out[10] = x;
    out[11] = x;
    out[12] = x;
    out[13] = x;
    out[14] = x;
    out[15] = x;
    out[16] = x;
    out[17] = x;
    out[18] = x;
    out[19] = x;
    out[20] = x;
    out[21] = x;
    out[22] = x;
    out[23] = x;
    out[24] = x;
    out[25] = x;
    out[26] = x;
    out[27] = x;
    out[28] = x;
    out[29] = x;
    out[30] = x;
    out[31] = x;
    out[32] = x;
    out[33] = x;
    out[34] = x;
    out[35] = x;
    out[36] = x;
    out[37] = x;
    out[38] = x;
    out[39] = x;
    out[40] = x;
    out[41] = x;
    out[42] = x;
    out[43] = x;
    out[44] = x;
    out[45] = x;
    out[46] = x;
    out[47] = x;
    out[48] = x;
    out[49] = x;
    out[50] = x;
    out[51] = x;
    out[52] = x;
    out[53] = x;
    out[54] = x;
    out[55] = x;
    out[56] = x;
    out[57] = x;
    out[58] = x;
    out[59] = x;
    out[60] = x;
    out[61] = x;
    out[62] = x;
    out[63] = x;
  }
}

static void idct64x64_low8_sse4_1(__m128i *in, __m128i *out, int bit,
                                  int do_cols, int bd, int out_shift) {
  int i, j;
  const int32_t *cospi = cospi_arr(bit);
  const __m128i rnding = _mm_set1_epi32(1 << (bit - 1));
  const int log_range = AOMMAX(16, bd + (do_cols ? 6 : 8));
  const __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
  const __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);

  const __m128i cospi1 = _mm_set1_epi32(cospi[1]);
  const __m128i cospi2 = _mm_set1_epi32(cospi[2]);
  const __m128i cospi3 = _mm_set1_epi32(cospi[3]);
  const __m128i cospi4 = _mm_set1_epi32(cospi[4]);
  const __m128i cospi6 = _mm_set1_epi32(cospi[6]);
  const __m128i cospi8 = _mm_set1_epi32(cospi[8]);
  const __m128i cospi12 = _mm_set1_epi32(cospi[12]);
  const __m128i cospi16 = _mm_set1_epi32(cospi[16]);
  const __m128i cospi20 = _mm_set1_epi32(cospi[20]);
  const __m128i cospi24 = _mm_set1_epi32(cospi[24]);
  const __m128i cospi28 = _mm_set1_epi32(cospi[28]);
  const __m128i cospi32 = _mm_set1_epi32(cospi[32]);
  const __m128i cospi40 = _mm_set1_epi32(cospi[40]);
  const __m128i cospi44 = _mm_set1_epi32(cospi[44]);
  const __m128i cospi48 = _mm_set1_epi32(cospi[48]);
  const __m128i cospi56 = _mm_set1_epi32(cospi[56]);
  const __m128i cospi60 = _mm_set1_epi32(cospi[60]);
  const __m128i cospim4 = _mm_set1_epi32(-cospi[4]);
  const __m128i cospim8 = _mm_set1_epi32(-cospi[8]);
  const __m128i cospim12 = _mm_set1_epi32(-cospi[12]);
  const __m128i cospim16 = _mm_set1_epi32(-cospi[16]);
  const __m128i cospim20 = _mm_set1_epi32(-cospi[20]);
  const __m128i cospim24 = _mm_set1_epi32(-cospi[24]);
  const __m128i cospim28 = _mm_set1_epi32(-cospi[28]);
  const __m128i cospim32 = _mm_set1_epi32(-cospi[32]);
  const __m128i cospim36 = _mm_set1_epi32(-cospi[36]);
  const __m128i cospim40 = _mm_set1_epi32(-cospi[40]);
  const __m128i cospim48 = _mm_set1_epi32(-cospi[48]);
  const __m128i cospim52 = _mm_set1_epi32(-cospi[52]);
  const __m128i cospim56 = _mm_set1_epi32(-cospi[56]);
  const __m128i cospi63 = _mm_set1_epi32(cospi[63]);
  const __m128i cospim57 = _mm_set1_epi32(-cospi[57]);
  const __m128i cospi7 = _mm_set1_epi32(cospi[7]);
  const __m128i cospi5 = _mm_set1_epi32(cospi[5]);
  const __m128i cospi59 = _mm_set1_epi32(cospi[59]);
  const __m128i cospim61 = _mm_set1_epi32(-cospi[61]);
  const __m128i cospim58 = _mm_set1_epi32(-cospi[58]);
  const __m128i cospi62 = _mm_set1_epi32(cospi[62]);

  {
    __m128i u[64];

    // stage 1
    u[0] = in[0];
    u[8] = in[4];
    u[16] = in[2];
    u[24] = in[6];
    u[32] = in[1];
    u[40] = in[5];
    u[48] = in[3];
    u[56] = in[7];

    // stage 2
    u[63] = half_btf_0_sse4_1(&cospi1, &u[32], &rnding, bit);
    u[32] = half_btf_0_sse4_1(&cospi63, &u[32], &rnding, bit);
    u[39] = half_btf_0_sse4_1(&cospim57, &u[56], &rnding, bit);
    u[56] = half_btf_0_sse4_1(&cospi7, &u[56], &rnding, bit);
    u[55] = half_btf_0_sse4_1(&cospi5, &u[40], &rnding, bit);
    u[40] = half_btf_0_sse4_1(&cospi59, &u[40], &rnding, bit);
    u[47] = half_btf_0_sse4_1(&cospim61, &u[48], &rnding, bit);
    u[48] = half_btf_0_sse4_1(&cospi3, &u[48], &rnding, bit);

    // stage 3
    u[31] = half_btf_0_sse4_1(&cospi2, &u[16], &rnding, bit);
    u[16] = half_btf_0_sse4_1(&cospi62, &u[16], &rnding, bit);
    u[23] = half_btf_0_sse4_1(&cospim58, &u[24], &rnding, bit);
    u[24] = half_btf_0_sse4_1(&cospi6, &u[24], &rnding, bit);
    u[33] = u[32];
    u[38] = u[39];
    u[41] = u[40];
    u[46] = u[47];
    u[49] = u[48];
    u[54] = u[55];
    u[57] = u[56];
    u[62] = u[63];

    // stage 4
    __m128i temp1, temp2;
    u[15] = half_btf_0_sse4_1(&cospi4, &u[8], &rnding, bit);
    u[8] = half_btf_0_sse4_1(&cospi60, &u[8], &rnding, bit);
    u[17] = u[16];
    u[22] = u[23];
    u[25] = u[24];
    u[30] = u[31];

    temp1 = half_btf_sse4_1(&cospim4, &u[33], &cospi60, &u[62], &rnding, bit);
    u[62] = half_btf_sse4_1(&cospi60, &u[33], &cospi4, &u[62], &rnding, bit);
    u[33] = temp1;

    temp2 = half_btf_sse4_1(&cospim36, &u[38], &cospi28, &u[57], &rnding, bit);
    u[38] = half_btf_sse4_1(&cospim28, &u[38], &cospim36, &u[57], &rnding, bit);
    u[57] = temp2;

    temp1 = half_btf_sse4_1(&cospim20, &u[41], &cospi44, &u[54], &rnding, bit);
    u[54] = half_btf_sse4_1(&cospi44, &u[41], &cospi20, &u[54], &rnding, bit);
    u[41] = temp1;

    temp2 = half_btf_sse4_1(&cospim12, &u[46], &cospim52, &u[49], &rnding, bit);
    u[49] = half_btf_sse4_1(&cospim52, &u[46], &cospi12, &u[49], &rnding, bit);
    u[46] = temp2;

    // stage 5
    u[9] = u[8];
    u[14] = u[15];

    temp1 = half_btf_sse4_1(&cospim8, &u[17], &cospi56, &u[30], &rnding, bit);
    u[30] = half_btf_sse4_1(&cospi56, &u[17], &cospi8, &u[30], &rnding, bit);
    u[17] = temp1;

    temp2 = half_btf_sse4_1(&cospim24, &u[22], &cospim40, &u[25], &rnding, bit);
    u[25] = half_btf_sse4_1(&cospim40, &u[22], &cospi24, &u[25], &rnding, bit);
    u[22] = temp2;

    u[35] = u[32];
    u[34] = u[33];
    u[36] = u[39];
    u[37] = u[38];
    u[43] = u[40];
    u[42] = u[41];
    u[44] = u[47];
    u[45] = u[46];
    u[51] = u[48];
    u[50] = u[49];
    u[52] = u[55];
    u[53] = u[54];
    u[59] = u[56];
    u[58] = u[57];
    u[60] = u[63];
    u[61] = u[62];

    // stage 6
    temp1 = half_btf_0_sse4_1(&cospi32, &u[0], &rnding, bit);
    u[1] = half_btf_0_sse4_1(&cospi32, &u[0], &rnding, bit);
    u[0] = temp1;

    temp2 = half_btf_sse4_1(&cospim16, &u[9], &cospi48, &u[14], &rnding, bit);
    u[14] = half_btf_sse4_1(&cospi48, &u[9], &cospi16, &u[14], &rnding, bit);
    u[9] = temp2;
    u[19] = u[16];
    u[18] = u[17];
    u[20] = u[23];
    u[21] = u[22];
    u[27] = u[24];
    u[26] = u[25];
    u[28] = u[31];
    u[29] = u[30];

    temp1 = half_btf_sse4_1(&cospim8, &u[34], &cospi56, &u[61], &rnding, bit);
    u[61] = half_btf_sse4_1(&cospi56, &u[34], &cospi8, &u[61], &rnding, bit);
    u[34] = temp1;
    temp2 = half_btf_sse4_1(&cospim8, &u[35], &cospi56, &u[60], &rnding, bit);
    u[60] = half_btf_sse4_1(&cospi56, &u[35], &cospi8, &u[60], &rnding, bit);
    u[35] = temp2;
    temp1 = half_btf_sse4_1(&cospim56, &u[36], &cospim8, &u[59], &rnding, bit);
    u[59] = half_btf_sse4_1(&cospim8, &u[36], &cospi56, &u[59], &rnding, bit);
    u[36] = temp1;
    temp2 = half_btf_sse4_1(&cospim56, &u[37], &cospim8, &u[58], &rnding, bit);
    u[58] = half_btf_sse4_1(&cospim8, &u[37], &cospi56, &u[58], &rnding, bit);
    u[37] = temp2;
    temp1 = half_btf_sse4_1(&cospim40, &u[42], &cospi24, &u[53], &rnding, bit);
    u[53] = half_btf_sse4_1(&cospi24, &u[42], &cospi40, &u[53], &rnding, bit);
    u[42] = temp1;
    temp2 = half_btf_sse4_1(&cospim40, &u[43], &cospi24, &u[52], &rnding, bit);
    u[52] = half_btf_sse4_1(&cospi24, &u[43], &cospi40, &u[52], &rnding, bit);
    u[43] = temp2;
    temp1 = half_btf_sse4_1(&cospim24, &u[44], &cospim40, &u[51], &rnding, bit);
    u[51] = half_btf_sse4_1(&cospim40, &u[44], &cospi24, &u[51], &rnding, bit);
    u[44] = temp1;
    temp2 = half_btf_sse4_1(&cospim24, &u[45], &cospim40, &u[50], &rnding, bit);
    u[50] = half_btf_sse4_1(&cospim40, &u[45], &cospi24, &u[50], &rnding, bit);
    u[45] = temp2;

    // stage 7
    u[3] = u[0];
    u[2] = u[1];
    u[11] = u[8];
    u[10] = u[9];
    u[12] = u[15];
    u[13] = u[14];

    temp1 = half_btf_sse4_1(&cospim16, &u[18], &cospi48, &u[29], &rnding, bit);
    u[29] = half_btf_sse4_1(&cospi48, &u[18], &cospi16, &u[29], &rnding, bit);
    u[18] = temp1;
    temp2 = half_btf_sse4_1(&cospim16, &u[19], &cospi48, &u[28], &rnding, bit);
    u[28] = half_btf_sse4_1(&cospi48, &u[19], &cospi16, &u[28], &rnding, bit);
    u[19] = temp2;
    temp1 = half_btf_sse4_1(&cospim48, &u[20], &cospim16, &u[27], &rnding, bit);
    u[27] = half_btf_sse4_1(&cospim16, &u[20], &cospi48, &u[27], &rnding, bit);
    u[20] = temp1;
    temp2 = half_btf_sse4_1(&cospim48, &u[21], &cospim16, &u[26], &rnding, bit);
    u[26] = half_btf_sse4_1(&cospim16, &u[21], &cospi48, &u[26], &rnding, bit);
    u[21] = temp2;
    for (i = 32; i < 64; i += 16) {
      for (j = i; j < i + 4; j++) {
        addsub_sse4_1(u[j], u[j ^ 7], &u[j], &u[j ^ 7], &clamp_lo, &clamp_hi);
        addsub_sse4_1(u[j ^ 15], u[j ^ 8], &u[j ^ 15], &u[j ^ 8], &clamp_lo,
                      &clamp_hi);
      }
    }

    // stage 8
    u[7] = u[0];
    u[6] = u[1];
    u[5] = u[2];
    u[4] = u[3];

    idct64_stage8_sse4_1(u, &cospim32, &cospi32, &cospim16, &cospi48, &cospi16,
                         &cospim48, &clamp_lo, &clamp_hi, &rnding, bit);

    // stage 9
    idct64_stage9_sse4_1(u, &cospim32, &cospi32, &clamp_lo, &clamp_hi, &rnding,
                         bit);

    // stage 10
    idct64_stage10_sse4_1(u, &cospim32, &cospi32, &clamp_lo, &clamp_hi, &rnding,
                          bit);

    // stage 11
    idct64_stage11_sse4_1(u, out, do_cols, bd, out_shift, &clamp_lo, &clamp_hi);
  }
}

static void idct64x64_low16_sse4_1(__m128i *in, __m128i *out, int bit,
                                   int do_cols, int bd, int out_shift) {
  int i, j;
  const int32_t *cospi = cospi_arr(bit);
  const __m128i rnding = _mm_set1_epi32(1 << (bit - 1));
  const int log_range = AOMMAX(16, bd + (do_cols ? 6 : 8));
  const __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
  const __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);

  const __m128i cospi1 = _mm_set1_epi32(cospi[1]);
  const __m128i cospi2 = _mm_set1_epi32(cospi[2]);
  const __m128i cospi3 = _mm_set1_epi32(cospi[3]);
  const __m128i cospi4 = _mm_set1_epi32(cospi[4]);
  const __m128i cospi5 = _mm_set1_epi32(cospi[5]);
  const __m128i cospi6 = _mm_set1_epi32(cospi[6]);
  const __m128i cospi7 = _mm_set1_epi32(cospi[7]);
  const __m128i cospi8 = _mm_set1_epi32(cospi[8]);
  const __m128i cospi9 = _mm_set1_epi32(cospi[9]);
  const __m128i cospi10 = _mm_set1_epi32(cospi[10]);
  const __m128i cospi11 = _mm_set1_epi32(cospi[11]);
  const __m128i cospi12 = _mm_set1_epi32(cospi[12]);
  const __m128i cospi13 = _mm_set1_epi32(cospi[13]);
  const __m128i cospi14 = _mm_set1_epi32(cospi[14]);
  const __m128i cospi15 = _mm_set1_epi32(cospi[15]);
  const __m128i cospi16 = _mm_set1_epi32(cospi[16]);
  const __m128i cospi20 = _mm_set1_epi32(cospi[20]);
  const __m128i cospi24 = _mm_set1_epi32(cospi[24]);
  const __m128i cospi28 = _mm_set1_epi32(cospi[28]);
  const __m128i cospi32 = _mm_set1_epi32(cospi[32]);
  const __m128i cospi36 = _mm_set1_epi32(cospi[36]);
  const __m128i cospi40 = _mm_set1_epi32(cospi[40]);
  const __m128i cospi44 = _mm_set1_epi32(cospi[44]);
  const __m128i cospi48 = _mm_set1_epi32(cospi[48]);
  const __m128i cospi51 = _mm_set1_epi32(cospi[51]);
  const __m128i cospi52 = _mm_set1_epi32(cospi[52]);
  const __m128i cospi54 = _mm_set1_epi32(cospi[54]);
  const __m128i cospi55 = _mm_set1_epi32(cospi[55]);
  const __m128i cospi56 = _mm_set1_epi32(cospi[56]);
  const __m128i cospi59 = _mm_set1_epi32(cospi[59]);
  const __m128i cospi60 = _mm_set1_epi32(cospi[60]);
  const __m128i cospi62 = _mm_set1_epi32(cospi[62]);
  const __m128i cospi63 = _mm_set1_epi32(cospi[63]);

  const __m128i cospim4 = _mm_set1_epi32(-cospi[4]);
  const __m128i cospim8 = _mm_set1_epi32(-cospi[8]);
  const __m128i cospim12 = _mm_set1_epi32(-cospi[12]);
  const __m128i cospim16 = _mm_set1_epi32(-cospi[16]);
  const __m128i cospim20 = _mm_set1_epi32(-cospi[20]);
  const __m128i cospim24 = _mm_set1_epi32(-cospi[24]);
  const __m128i cospim28 = _mm_set1_epi32(-cospi[28]);
  const __m128i cospim32 = _mm_set1_epi32(-cospi[32]);
  const __m128i cospim36 = _mm_set1_epi32(-cospi[36]);
  const __m128i cospim40 = _mm_set1_epi32(-cospi[40]);
  const __m128i cospim44 = _mm_set1_epi32(-cospi[44]);
  const __m128i cospim48 = _mm_set1_epi32(-cospi[48]);
  const __m128i cospim49 = _mm_set1_epi32(-cospi[49]);
  const __m128i cospim50 = _mm_set1_epi32(-cospi[50]);
  const __m128i cospim52 = _mm_set1_epi32(-cospi[52]);
  const __m128i cospim53 = _mm_set1_epi32(-cospi[53]);
  const __m128i cospim56 = _mm_set1_epi32(-cospi[56]);
  const __m128i cospim57 = _mm_set1_epi32(-cospi[57]);
  const __m128i cospim58 = _mm_set1_epi32(-cospi[58]);
  const __m128i cospim60 = _mm_set1_epi32(-cospi[60]);
  const __m128i cospim61 = _mm_set1_epi32(-cospi[61]);

  {
    __m128i u[64];
    __m128i tmp1, tmp2, tmp3, tmp4;
    // stage 1
    u[0] = in[0];
    u[32] = in[1];
    u[36] = in[9];
    u[40] = in[5];
    u[44] = in[13];
    u[48] = in[3];
    u[52] = in[11];
    u[56] = in[7];
    u[60] = in[15];
    u[16] = in[2];
    u[20] = in[10];
    u[24] = in[6];
    u[28] = in[14];
    u[4] = in[8];
    u[8] = in[4];
    u[12] = in[12];

    // stage 2
    u[63] = half_btf_0_sse4_1(&cospi1, &u[32], &rnding, bit);
    u[32] = half_btf_0_sse4_1(&cospi63, &u[32], &rnding, bit);
    u[35] = half_btf_0_sse4_1(&cospim49, &u[60], &rnding, bit);
    u[60] = half_btf_0_sse4_1(&cospi15, &u[60], &rnding, bit);
    u[59] = half_btf_0_sse4_1(&cospi9, &u[36], &rnding, bit);
    u[36] = half_btf_0_sse4_1(&cospi55, &u[36], &rnding, bit);
    u[39] = half_btf_0_sse4_1(&cospim57, &u[56], &rnding, bit);
    u[56] = half_btf_0_sse4_1(&cospi7, &u[56], &rnding, bit);
    u[55] = half_btf_0_sse4_1(&cospi5, &u[40], &rnding, bit);
    u[40] = half_btf_0_sse4_1(&cospi59, &u[40], &rnding, bit);
    u[43] = half_btf_0_sse4_1(&cospim53, &u[52], &rnding, bit);
    u[52] = half_btf_0_sse4_1(&cospi11, &u[52], &rnding, bit);
    u[47] = half_btf_0_sse4_1(&cospim61, &u[48], &rnding, bit);
    u[48] = half_btf_0_sse4_1(&cospi3, &u[48], &rnding, bit);
    u[51] = half_btf_0_sse4_1(&cospi13, &u[44], &rnding, bit);
    u[44] = half_btf_0_sse4_1(&cospi51, &u[44], &rnding, bit);

    // stage 3
    u[31] = half_btf_0_sse4_1(&cospi2, &u[16], &rnding, bit);
    u[16] = half_btf_0_sse4_1(&cospi62, &u[16], &rnding, bit);
    u[19] = half_btf_0_sse4_1(&cospim50, &u[28], &rnding, bit);
    u[28] = half_btf_0_sse4_1(&cospi14, &u[28], &rnding, bit);
    u[27] = half_btf_0_sse4_1(&cospi10, &u[20], &rnding, bit);
    u[20] = half_btf_0_sse4_1(&cospi54, &u[20], &rnding, bit);
    u[23] = half_btf_0_sse4_1(&cospim58, &u[24], &rnding, bit);
    u[24] = half_btf_0_sse4_1(&cospi6, &u[24], &rnding, bit);
    u[33] = u[32];
    u[34] = u[35];
    u[37] = u[36];
    u[38] = u[39];
    u[41] = u[40];
    u[42] = u[43];
    u[45] = u[44];
    u[46] = u[47];
    u[49] = u[48];
    u[50] = u[51];
    u[53] = u[52];
    u[54] = u[55];
    u[57] = u[56];
    u[58] = u[59];
    u[61] = u[60];
    u[62] = u[63];

    // stage 4
    u[15] = half_btf_0_sse4_1(&cospi4, &u[8], &rnding, bit);
    u[8] = half_btf_0_sse4_1(&cospi60, &u[8], &rnding, bit);
    u[11] = half_btf_0_sse4_1(&cospim52, &u[12], &rnding, bit);
    u[12] = half_btf_0_sse4_1(&cospi12, &u[12], &rnding, bit);

    u[17] = u[16];
    u[18] = u[19];
    u[21] = u[20];
    u[22] = u[23];
    u[25] = u[24];
    u[26] = u[27];
    u[29] = u[28];
    u[30] = u[31];

    tmp1 = half_btf_sse4_1(&cospim4, &u[33], &cospi60, &u[62], &rnding, bit);
    tmp2 = half_btf_sse4_1(&cospim60, &u[34], &cospim4, &u[61], &rnding, bit);
    tmp3 = half_btf_sse4_1(&cospim36, &u[37], &cospi28, &u[58], &rnding, bit);
    tmp4 = half_btf_sse4_1(&cospim28, &u[38], &cospim36, &u[57], &rnding, bit);
    u[57] = half_btf_sse4_1(&cospim36, &u[38], &cospi28, &u[57], &rnding, bit);
    u[58] = half_btf_sse4_1(&cospi28, &u[37], &cospi36, &u[58], &rnding, bit);
    u[61] = half_btf_sse4_1(&cospim4, &u[34], &cospi60, &u[61], &rnding, bit);
    u[62] = half_btf_sse4_1(&cospi60, &u[33], &cospi4, &u[62], &rnding, bit);
    u[33] = tmp1;
    u[34] = tmp2;
    u[37] = tmp3;
    u[38] = tmp4;

    tmp1 = half_btf_sse4_1(&cospim20, &u[41], &cospi44, &u[54], &rnding, bit);
    tmp2 = half_btf_sse4_1(&cospim44, &u[42], &cospim20, &u[53], &rnding, bit);
    tmp3 = half_btf_sse4_1(&cospim52, &u[45], &cospi12, &u[50], &rnding, bit);
    tmp4 = half_btf_sse4_1(&cospim12, &u[46], &cospim52, &u[49], &rnding, bit);
    u[49] = half_btf_sse4_1(&cospim52, &u[46], &cospi12, &u[49], &rnding, bit);
    u[50] = half_btf_sse4_1(&cospi12, &u[45], &cospi52, &u[50], &rnding, bit);
    u[53] = half_btf_sse4_1(&cospim20, &u[42], &cospi44, &u[53], &rnding, bit);
    u[54] = half_btf_sse4_1(&cospi44, &u[41], &cospi20, &u[54], &rnding, bit);
    u[41] = tmp1;
    u[42] = tmp2;
    u[45] = tmp3;
    u[46] = tmp4;

    // stage 5
    u[7] = half_btf_0_sse4_1(&cospi8, &u[4], &rnding, bit);
    u[4] = half_btf_0_sse4_1(&cospi56, &u[4], &rnding, bit);

    u[9] = u[8];
    u[10] = u[11];
    u[13] = u[12];
    u[14] = u[15];

    tmp1 = half_btf_sse4_1(&cospim8, &u[17], &cospi56, &u[30], &rnding, bit);
    tmp2 = half_btf_sse4_1(&cospim56, &u[18], &cospim8, &u[29], &rnding, bit);
    tmp3 = half_btf_sse4_1(&cospim40, &u[21], &cospi24, &u[26], &rnding, bit);
    tmp4 = half_btf_sse4_1(&cospim24, &u[22], &cospim40, &u[25], &rnding, bit);
    u[25] = half_btf_sse4_1(&cospim40, &u[22], &cospi24, &u[25], &rnding, bit);
    u[26] = half_btf_sse4_1(&cospi24, &u[21], &cospi40, &u[26], &rnding, bit);
    u[29] = half_btf_sse4_1(&cospim8, &u[18], &cospi56, &u[29], &rnding, bit);
    u[30] = half_btf_sse4_1(&cospi56, &u[17], &cospi8, &u[30], &rnding, bit);
    u[17] = tmp1;
    u[18] = tmp2;
    u[21] = tmp3;
    u[22] = tmp4;

    for (i = 32; i < 64; i += 8) {
      addsub_sse4_1(u[i + 0], u[i + 3], &u[i + 0], &u[i + 3], &clamp_lo,
                    &clamp_hi);
      addsub_sse4_1(u[i + 1], u[i + 2], &u[i + 1], &u[i + 2], &clamp_lo,
                    &clamp_hi);

      addsub_sse4_1(u[i + 7], u[i + 4], &u[i + 7], &u[i + 4], &clamp_lo,
                    &clamp_hi);
      addsub_sse4_1(u[i + 6], u[i + 5], &u[i + 6], &u[i + 5], &clamp_lo,
                    &clamp_hi);
    }

    // stage 6
    tmp1 = half_btf_0_sse4_1(&cospi32, &u[0], &rnding, bit);
    u[1] = half_btf_0_sse4_1(&cospi32, &u[0], &rnding, bit);
    u[0] = tmp1;
    u[5] = u[4];
    u[6] = u[7];

    tmp1 = half_btf_sse4_1(&cospim16, &u[9], &cospi48, &u[14], &rnding, bit);
    u[14] = half_btf_sse4_1(&cospi48, &u[9], &cospi16, &u[14], &rnding, bit);
    u[9] = tmp1;
    tmp2 = half_btf_sse4_1(&cospim48, &u[10], &cospim16, &u[13], &rnding, bit);
    u[13] = half_btf_sse4_1(&cospim16, &u[10], &cospi48, &u[13], &rnding, bit);
    u[10] = tmp2;

    for (i = 16; i < 32; i += 8) {
      addsub_sse4_1(u[i + 0], u[i + 3], &u[i + 0], &u[i + 3], &clamp_lo,
                    &clamp_hi);
      addsub_sse4_1(u[i + 1], u[i + 2], &u[i + 1], &u[i + 2], &clamp_lo,
                    &clamp_hi);

      addsub_sse4_1(u[i + 7], u[i + 4], &u[i + 7], &u[i + 4], &clamp_lo,
                    &clamp_hi);
      addsub_sse4_1(u[i + 6], u[i + 5], &u[i + 6], &u[i + 5], &clamp_lo,
                    &clamp_hi);
    }

    tmp1 = half_btf_sse4_1(&cospim8, &u[34], &cospi56, &u[61], &rnding, bit);
    tmp2 = half_btf_sse4_1(&cospim8, &u[35], &cospi56, &u[60], &rnding, bit);
    tmp3 = half_btf_sse4_1(&cospim56, &u[36], &cospim8, &u[59], &rnding, bit);
    tmp4 = half_btf_sse4_1(&cospim56, &u[37], &cospim8, &u[58], &rnding, bit);
    u[58] = half_btf_sse4_1(&cospim8, &u[37], &cospi56, &u[58], &rnding, bit);
    u[59] = half_btf_sse4_1(&cospim8, &u[36], &cospi56, &u[59], &rnding, bit);
    u[60] = half_btf_sse4_1(&cospi56, &u[35], &cospi8, &u[60], &rnding, bit);
    u[61] = half_btf_sse4_1(&cospi56, &u[34], &cospi8, &u[61], &rnding, bit);
    u[34] = tmp1;
    u[35] = tmp2;
    u[36] = tmp3;
    u[37] = tmp4;

    tmp1 = half_btf_sse4_1(&cospim40, &u[42], &cospi24, &u[53], &rnding, bit);
    tmp2 = half_btf_sse4_1(&cospim40, &u[43], &cospi24, &u[52], &rnding, bit);
    tmp3 = half_btf_sse4_1(&cospim24, &u[44], &cospim40, &u[51], &rnding, bit);
    tmp4 = half_btf_sse4_1(&cospim24, &u[45], &cospim40, &u[50], &rnding, bit);
    u[50] = half_btf_sse4_1(&cospim40, &u[45], &cospi24, &u[50], &rnding, bit);
    u[51] = half_btf_sse4_1(&cospim40, &u[44], &cospi24, &u[51], &rnding, bit);
    u[52] = half_btf_sse4_1(&cospi24, &u[43], &cospi40, &u[52], &rnding, bit);
    u[53] = half_btf_sse4_1(&cospi24, &u[42], &cospi40, &u[53], &rnding, bit);
    u[42] = tmp1;
    u[43] = tmp2;
    u[44] = tmp3;
    u[45] = tmp4;

    // stage 7
    u[3] = u[0];
    u[2] = u[1];
    tmp1 = half_btf_sse4_1(&cospim32, &u[5], &cospi32, &u[6], &rnding, bit);
    u[6] = half_btf_sse4_1(&cospi32, &u[5], &cospi32, &u[6], &rnding, bit);
    u[5] = tmp1;
    addsub_sse4_1(u[8], u[11], &u[8], &u[11], &clamp_lo, &clamp_hi);
    addsub_sse4_1(u[9], u[10], &u[9], &u[10], &clamp_lo, &clamp_hi);
    addsub_sse4_1(u[15], u[12], &u[15], &u[12], &clamp_lo, &clamp_hi);
    addsub_sse4_1(u[14], u[13], &u[14], &u[13], &clamp_lo, &clamp_hi);

    tmp1 = half_btf_sse4_1(&cospim16, &u[18], &cospi48, &u[29], &rnding, bit);
    tmp2 = half_btf_sse4_1(&cospim16, &u[19], &cospi48, &u[28], &rnding, bit);
    tmp3 = half_btf_sse4_1(&cospim48, &u[20], &cospim16, &u[27], &rnding, bit);
    tmp4 = half_btf_sse4_1(&cospim48, &u[21], &cospim16, &u[26], &rnding, bit);
    u[26] = half_btf_sse4_1(&cospim16, &u[21], &cospi48, &u[26], &rnding, bit);
    u[27] = half_btf_sse4_1(&cospim16, &u[20], &cospi48, &u[27], &rnding, bit);
    u[28] = half_btf_sse4_1(&cospi48, &u[19], &cospi16, &u[28], &rnding, bit);
    u[29] = half_btf_sse4_1(&cospi48, &u[18], &cospi16, &u[29], &rnding, bit);
    u[18] = tmp1;
    u[19] = tmp2;
    u[20] = tmp3;
    u[21] = tmp4;

    for (i = 32; i < 64; i += 16) {
      for (j = i; j < i + 4; j++) {
        addsub_sse4_1(u[j], u[j ^ 7], &u[j], &u[j ^ 7], &clamp_lo, &clamp_hi);
        addsub_sse4_1(u[j ^ 15], u[j ^ 8], &u[j ^ 15], &u[j ^ 8], &clamp_lo,
                      &clamp_hi);
      }
    }

    // stage 8
    for (i = 0; i < 4; ++i) {
      addsub_sse4_1(u[i], u[7 - i], &u[i], &u[7 - i], &clamp_lo, &clamp_hi);
    }

    idct64_stage8_sse4_1(u, &cospim32, &cospi32, &cospim16, &cospi48, &cospi16,
                         &cospim48, &clamp_lo, &clamp_hi, &rnding, bit);

    // stage 9
    idct64_stage9_sse4_1(u, &cospim32, &cospi32, &clamp_lo, &clamp_hi, &rnding,
                         bit);

    // stage 10
    idct64_stage10_sse4_1(u, &cospim32, &cospi32, &clamp_lo, &clamp_hi, &rnding,
                          bit);

    // stage 11
    idct64_stage11_sse4_1(u, out, do_cols, bd, out_shift, &clamp_lo, &clamp_hi);
  }
}

static void idct64x64_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                             int bd, int out_shift) {
  int i, j;
  const int32_t *cospi = cospi_arr(bit);
  const __m128i rnding = _mm_set1_epi32(1 << (bit - 1));
  const int log_range = AOMMAX(16, bd + (do_cols ? 6 : 8));
  const __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
  const __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);

  const __m128i cospi1 = _mm_set1_epi32(cospi[1]);
  const __m128i cospi2 = _mm_set1_epi32(cospi[2]);
  const __m128i cospi3 = _mm_set1_epi32(cospi[3]);
  const __m128i cospi4 = _mm_set1_epi32(cospi[4]);
  const __m128i cospi5 = _mm_set1_epi32(cospi[5]);
  const __m128i cospi6 = _mm_set1_epi32(cospi[6]);
  const __m128i cospi7 = _mm_set1_epi32(cospi[7]);
  const __m128i cospi8 = _mm_set1_epi32(cospi[8]);
  const __m128i cospi9 = _mm_set1_epi32(cospi[9]);
  const __m128i cospi10 = _mm_set1_epi32(cospi[10]);
  const __m128i cospi11 = _mm_set1_epi32(cospi[11]);
  const __m128i cospi12 = _mm_set1_epi32(cospi[12]);
  const __m128i cospi13 = _mm_set1_epi32(cospi[13]);
  const __m128i cospi14 = _mm_set1_epi32(cospi[14]);
  const __m128i cospi15 = _mm_set1_epi32(cospi[15]);
  const __m128i cospi16 = _mm_set1_epi32(cospi[16]);
  const __m128i cospi17 = _mm_set1_epi32(cospi[17]);
  const __m128i cospi18 = _mm_set1_epi32(cospi[18]);
  const __m128i cospi19 = _mm_set1_epi32(cospi[19]);
  const __m128i cospi20 = _mm_set1_epi32(cospi[20]);
  const __m128i cospi21 = _mm_set1_epi32(cospi[21]);
  const __m128i cospi22 = _mm_set1_epi32(cospi[22]);
  const __m128i cospi23 = _mm_set1_epi32(cospi[23]);
  const __m128i cospi24 = _mm_set1_epi32(cospi[24]);
  const __m128i cospi25 = _mm_set1_epi32(cospi[25]);
  const __m128i cospi26 = _mm_set1_epi32(cospi[26]);
  const __m128i cospi27 = _mm_set1_epi32(cospi[27]);
  const __m128i cospi28 = _mm_set1_epi32(cospi[28]);
  const __m128i cospi29 = _mm_set1_epi32(cospi[29]);
  const __m128i cospi30 = _mm_set1_epi32(cospi[30]);
  const __m128i cospi31 = _mm_set1_epi32(cospi[31]);
  const __m128i cospi32 = _mm_set1_epi32(cospi[32]);
  const __m128i cospi35 = _mm_set1_epi32(cospi[35]);
  const __m128i cospi36 = _mm_set1_epi32(cospi[36]);
  const __m128i cospi38 = _mm_set1_epi32(cospi[38]);
  const __m128i cospi39 = _mm_set1_epi32(cospi[39]);
  const __m128i cospi40 = _mm_set1_epi32(cospi[40]);
  const __m128i cospi43 = _mm_set1_epi32(cospi[43]);
  const __m128i cospi44 = _mm_set1_epi32(cospi[44]);
  const __m128i cospi46 = _mm_set1_epi32(cospi[46]);
  const __m128i cospi47 = _mm_set1_epi32(cospi[47]);
  const __m128i cospi48 = _mm_set1_epi32(cospi[48]);
  const __m128i cospi51 = _mm_set1_epi32(cospi[51]);
  const __m128i cospi52 = _mm_set1_epi32(cospi[52]);
  const __m128i cospi54 = _mm_set1_epi32(cospi[54]);
  const __m128i cospi55 = _mm_set1_epi32(cospi[55]);
  const __m128i cospi56 = _mm_set1_epi32(cospi[56]);
  const __m128i cospi59 = _mm_set1_epi32(cospi[59]);
  const __m128i cospi60 = _mm_set1_epi32(cospi[60]);
  const __m128i cospi62 = _mm_set1_epi32(cospi[62]);
  const __m128i cospi63 = _mm_set1_epi32(cospi[63]);

  const __m128i cospim4 = _mm_set1_epi32(-cospi[4]);
  const __m128i cospim8 = _mm_set1_epi32(-cospi[8]);
  const __m128i cospim12 = _mm_set1_epi32(-cospi[12]);
  const __m128i cospim16 = _mm_set1_epi32(-cospi[16]);
  const __m128i cospim20 = _mm_set1_epi32(-cospi[20]);
  const __m128i cospim24 = _mm_set1_epi32(-cospi[24]);
  const __m128i cospim28 = _mm_set1_epi32(-cospi[28]);
  const __m128i cospim32 = _mm_set1_epi32(-cospi[32]);
  const __m128i cospim33 = _mm_set1_epi32(-cospi[33]);
  const __m128i cospim34 = _mm_set1_epi32(-cospi[34]);
  const __m128i cospim36 = _mm_set1_epi32(-cospi[36]);
  const __m128i cospim37 = _mm_set1_epi32(-cospi[37]);
  const __m128i cospim40 = _mm_set1_epi32(-cospi[40]);
  const __m128i cospim41 = _mm_set1_epi32(-cospi[41]);
  const __m128i cospim42 = _mm_set1_epi32(-cospi[42]);
  const __m128i cospim44 = _mm_set1_epi32(-cospi[44]);
  const __m128i cospim45 = _mm_set1_epi32(-cospi[45]);
  const __m128i cospim48 = _mm_set1_epi32(-cospi[48]);
  const __m128i cospim49 = _mm_set1_epi32(-cospi[49]);
  const __m128i cospim50 = _mm_set1_epi32(-cospi[50]);
  const __m128i cospim52 = _mm_set1_epi32(-cospi[52]);
  const __m128i cospim53 = _mm_set1_epi32(-cospi[53]);
  const __m128i cospim56 = _mm_set1_epi32(-cospi[56]);
  const __m128i cospim57 = _mm_set1_epi32(-cospi[57]);
  const __m128i cospim58 = _mm_set1_epi32(-cospi[58]);
  const __m128i cospim60 = _mm_set1_epi32(-cospi[60]);
  const __m128i cospim61 = _mm_set1_epi32(-cospi[61]);

  {
    __m128i u[64], v[64];

    // stage 1
    u[32] = in[1];
    u[34] = in[17];
    u[36] = in[9];
    u[38] = in[25];
    u[40] = in[5];
    u[42] = in[21];
    u[44] = in[13];
    u[46] = in[29];
    u[48] = in[3];
    u[50] = in[19];
    u[52] = in[11];
    u[54] = in[27];
    u[56] = in[7];
    u[58] = in[23];
    u[60] = in[15];
    u[62] = in[31];

    v[16] = in[2];
    v[18] = in[18];
    v[20] = in[10];
    v[22] = in[26];
    v[24] = in[6];
    v[26] = in[22];
    v[28] = in[14];
    v[30] = in[30];

    u[8] = in[4];
    u[10] = in[20];
    u[12] = in[12];
    u[14] = in[28];

    v[4] = in[8];
    v[6] = in[24];

    u[0] = in[0];
    u[2] = in[16];

    // stage 2
    v[32] = half_btf_0_sse4_1(&cospi63, &u[32], &rnding, bit);
    v[33] = half_btf_0_sse4_1(&cospim33, &u[62], &rnding, bit);
    v[34] = half_btf_0_sse4_1(&cospi47, &u[34], &rnding, bit);
    v[35] = half_btf_0_sse4_1(&cospim49, &u[60], &rnding, bit);
    v[36] = half_btf_0_sse4_1(&cospi55, &u[36], &rnding, bit);
    v[37] = half_btf_0_sse4_1(&cospim41, &u[58], &rnding, bit);
    v[38] = half_btf_0_sse4_1(&cospi39, &u[38], &rnding, bit);
    v[39] = half_btf_0_sse4_1(&cospim57, &u[56], &rnding, bit);
    v[40] = half_btf_0_sse4_1(&cospi59, &u[40], &rnding, bit);
    v[41] = half_btf_0_sse4_1(&cospim37, &u[54], &rnding, bit);
    v[42] = half_btf_0_sse4_1(&cospi43, &u[42], &rnding, bit);
    v[43] = half_btf_0_sse4_1(&cospim53, &u[52], &rnding, bit);
    v[44] = half_btf_0_sse4_1(&cospi51, &u[44], &rnding, bit);
    v[45] = half_btf_0_sse4_1(&cospim45, &u[50], &rnding, bit);
    v[46] = half_btf_0_sse4_1(&cospi35, &u[46], &rnding, bit);
    v[47] = half_btf_0_sse4_1(&cospim61, &u[48], &rnding, bit);
    v[48] = half_btf_0_sse4_1(&cospi3, &u[48], &rnding, bit);
    v[49] = half_btf_0_sse4_1(&cospi29, &u[46], &rnding, bit);
    v[50] = half_btf_0_sse4_1(&cospi19, &u[50], &rnding, bit);
    v[51] = half_btf_0_sse4_1(&cospi13, &u[44], &rnding, bit);
    v[52] = half_btf_0_sse4_1(&cospi11, &u[52], &rnding, bit);
    v[53] = half_btf_0_sse4_1(&cospi21, &u[42], &rnding, bit);
    v[54] = half_btf_0_sse4_1(&cospi27, &u[54], &rnding, bit);
    v[55] = half_btf_0_sse4_1(&cospi5, &u[40], &rnding, bit);
    v[56] = half_btf_0_sse4_1(&cospi7, &u[56], &rnding, bit);
    v[57] = half_btf_0_sse4_1(&cospi25, &u[38], &rnding, bit);
    v[58] = half_btf_0_sse4_1(&cospi23, &u[58], &rnding, bit);
    v[59] = half_btf_0_sse4_1(&cospi9, &u[36], &rnding, bit);
    v[60] = half_btf_0_sse4_1(&cospi15, &u[60], &rnding, bit);
    v[61] = half_btf_0_sse4_1(&cospi17, &u[34], &rnding, bit);
    v[62] = half_btf_0_sse4_1(&cospi31, &u[62], &rnding, bit);
    v[63] = half_btf_0_sse4_1(&cospi1, &u[32], &rnding, bit);

    // stage 3
    u[16] = half_btf_0_sse4_1(&cospi62, &v[16], &rnding, bit);
    u[17] = half_btf_0_sse4_1(&cospim34, &v[30], &rnding, bit);
    u[18] = half_btf_0_sse4_1(&cospi46, &v[18], &rnding, bit);
    u[19] = half_btf_0_sse4_1(&cospim50, &v[28], &rnding, bit);
    u[20] = half_btf_0_sse4_1(&cospi54, &v[20], &rnding, bit);
    u[21] = half_btf_0_sse4_1(&cospim42, &v[26], &rnding, bit);
    u[22] = half_btf_0_sse4_1(&cospi38, &v[22], &rnding, bit);
    u[23] = half_btf_0_sse4_1(&cospim58, &v[24], &rnding, bit);
    u[24] = half_btf_0_sse4_1(&cospi6, &v[24], &rnding, bit);
    u[25] = half_btf_0_sse4_1(&cospi26, &v[22], &rnding, bit);
    u[26] = half_btf_0_sse4_1(&cospi22, &v[26], &rnding, bit);
    u[27] = half_btf_0_sse4_1(&cospi10, &v[20], &rnding, bit);
    u[28] = half_btf_0_sse4_1(&cospi14, &v[28], &rnding, bit);
    u[29] = half_btf_0_sse4_1(&cospi18, &v[18], &rnding, bit);
    u[30] = half_btf_0_sse4_1(&cospi30, &v[30], &rnding, bit);
    u[31] = half_btf_0_sse4_1(&cospi2, &v[16], &rnding, bit);

    for (i = 32; i < 64; i += 4) {
      addsub_sse4_1(v[i + 0], v[i + 1], &u[i + 0], &u[i + 1], &clamp_lo,
                    &clamp_hi);
      addsub_sse4_1(v[i + 3], v[i + 2], &u[i + 3], &u[i + 2], &clamp_lo,
                    &clamp_hi);
    }

    // stage 4
    v[8] = half_btf_0_sse4_1(&cospi60, &u[8], &rnding, bit);
    v[9] = half_btf_0_sse4_1(&cospim36, &u[14], &rnding, bit);
    v[10] = half_btf_0_sse4_1(&cospi44, &u[10], &rnding, bit);
    v[11] = half_btf_0_sse4_1(&cospim52, &u[12], &rnding, bit);
    v[12] = half_btf_0_sse4_1(&cospi12, &u[12], &rnding, bit);
    v[13] = half_btf_0_sse4_1(&cospi20, &u[10], &rnding, bit);
    v[14] = half_btf_0_sse4_1(&cospi28, &u[14], &rnding, bit);
    v[15] = half_btf_0_sse4_1(&cospi4, &u[8], &rnding, bit);

    for (i = 16; i < 32; i += 4) {
      addsub_sse4_1(u[i + 0], u[i + 1], &v[i + 0], &v[i + 1], &clamp_lo,
                    &clamp_hi);
      addsub_sse4_1(u[i + 3], u[i + 2], &v[i + 3], &v[i + 2], &clamp_lo,
                    &clamp_hi);
    }

    for (i = 32; i < 64; i += 4) {
      v[i + 0] = u[i + 0];
      v[i + 3] = u[i + 3];
    }

    v[33] = half_btf_sse4_1(&cospim4, &u[33], &cospi60, &u[62], &rnding, bit);
    v[34] = half_btf_sse4_1(&cospim60, &u[34], &cospim4, &u[61], &rnding, bit);
    v[37] = half_btf_sse4_1(&cospim36, &u[37], &cospi28, &u[58], &rnding, bit);
    v[38] = half_btf_sse4_1(&cospim28, &u[38], &cospim36, &u[57], &rnding, bit);
    v[41] = half_btf_sse4_1(&cospim20, &u[41], &cospi44, &u[54], &rnding, bit);
    v[42] = half_btf_sse4_1(&cospim44, &u[42], &cospim20, &u[53], &rnding, bit);
    v[45] = half_btf_sse4_1(&cospim52, &u[45], &cospi12, &u[50], &rnding, bit);
    v[46] = half_btf_sse4_1(&cospim12, &u[46], &cospim52, &u[49], &rnding, bit);
    v[49] = half_btf_sse4_1(&cospim52, &u[46], &cospi12, &u[49], &rnding, bit);
    v[50] = half_btf_sse4_1(&cospi12, &u[45], &cospi52, &u[50], &rnding, bit);
    v[53] = half_btf_sse4_1(&cospim20, &u[42], &cospi44, &u[53], &rnding, bit);
    v[54] = half_btf_sse4_1(&cospi44, &u[41], &cospi20, &u[54], &rnding, bit);
    v[57] = half_btf_sse4_1(&cospim36, &u[38], &cospi28, &u[57], &rnding, bit);
    v[58] = half_btf_sse4_1(&cospi28, &u[37], &cospi36, &u[58], &rnding, bit);
    v[61] = half_btf_sse4_1(&cospim4, &u[34], &cospi60, &u[61], &rnding, bit);
    v[62] = half_btf_sse4_1(&cospi60, &u[33], &cospi4, &u[62], &rnding, bit);

    // stage 5
    u[4] = half_btf_0_sse4_1(&cospi56, &v[4], &rnding, bit);
    u[5] = half_btf_0_sse4_1(&cospim40, &v[6], &rnding, bit);
    u[6] = half_btf_0_sse4_1(&cospi24, &v[6], &rnding, bit);
    u[7] = half_btf_0_sse4_1(&cospi8, &v[4], &rnding, bit);

    for (i = 8; i < 16; i += 4) {
      addsub_sse4_1(v[i + 0], v[i + 1], &u[i + 0], &u[i + 1], &clamp_lo,
                    &clamp_hi);
      addsub_sse4_1(v[i + 3], v[i + 2], &u[i + 3], &u[i + 2], &clamp_lo,
                    &clamp_hi);
    }

    for (i = 16; i < 32; i += 4) {
      u[i + 0] = v[i + 0];
      u[i + 3] = v[i + 3];
    }

    u[17] = half_btf_sse4_1(&cospim8, &v[17], &cospi56, &v[30], &rnding, bit);
    u[18] = half_btf_sse4_1(&cospim56, &v[18], &cospim8, &v[29], &rnding, bit);
    u[21] = half_btf_sse4_1(&cospim40, &v[21], &cospi24, &v[26], &rnding, bit);
    u[22] = half_btf_sse4_1(&cospim24, &v[22], &cospim40, &v[25], &rnding, bit);
    u[25] = half_btf_sse4_1(&cospim40, &v[22], &cospi24, &v[25], &rnding, bit);
    u[26] = half_btf_sse4_1(&cospi24, &v[21], &cospi40, &v[26], &rnding, bit);
    u[29] = half_btf_sse4_1(&cospim8, &v[18], &cospi56, &v[29], &rnding, bit);
    u[30] = half_btf_sse4_1(&cospi56, &v[17], &cospi8, &v[30], &rnding, bit);

    for (i = 32; i < 64; i += 8) {
      addsub_sse4_1(v[i + 0], v[i + 3], &u[i + 0], &u[i + 3], &clamp_lo,
                    &clamp_hi);
      addsub_sse4_1(v[i + 1], v[i + 2], &u[i + 1], &u[i + 2], &clamp_lo,
                    &clamp_hi);

      addsub_sse4_1(v[i + 7], v[i + 4], &u[i + 7], &u[i + 4], &clamp_lo,
                    &clamp_hi);
      addsub_sse4_1(v[i + 6], v[i + 5], &u[i + 6], &u[i + 5], &clamp_lo,
                    &clamp_hi);
    }

    // stage 6
    v[0] = half_btf_0_sse4_1(&cospi32, &u[0], &rnding, bit);
    v[1] = half_btf_0_sse4_1(&cospi32, &u[0], &rnding, bit);
    v[2] = half_btf_0_sse4_1(&cospi48, &u[2], &rnding, bit);
    v[3] = half_btf_0_sse4_1(&cospi16, &u[2], &rnding, bit);

    addsub_sse4_1(u[4], u[5], &v[4], &v[5], &clamp_lo, &clamp_hi);
    addsub_sse4_1(u[7], u[6], &v[7], &v[6], &clamp_lo, &clamp_hi);

    for (i = 8; i < 16; i += 4) {
      v[i + 0] = u[i + 0];
      v[i + 3] = u[i + 3];
    }

    v[9] = half_btf_sse4_1(&cospim16, &u[9], &cospi48, &u[14], &rnding, bit);
    v[10] = half_btf_sse4_1(&cospim48, &u[10], &cospim16, &u[13], &rnding, bit);
    v[13] = half_btf_sse4_1(&cospim16, &u[10], &cospi48, &u[13], &rnding, bit);
    v[14] = half_btf_sse4_1(&cospi48, &u[9], &cospi16, &u[14], &rnding, bit);

    for (i = 16; i < 32; i += 8) {
      addsub_sse4_1(u[i + 0], u[i + 3], &v[i + 0], &v[i + 3], &clamp_lo,
                    &clamp_hi);
      addsub_sse4_1(u[i + 1], u[i + 2], &v[i + 1], &v[i + 2], &clamp_lo,
                    &clamp_hi);

      addsub_sse4_1(u[i + 7], u[i + 4], &v[i + 7], &v[i + 4], &clamp_lo,
                    &clamp_hi);
      addsub_sse4_1(u[i + 6], u[i + 5], &v[i + 6], &v[i + 5], &clamp_lo,
                    &clamp_hi);
    }

    for (i = 32; i < 64; i += 8) {
      v[i + 0] = u[i + 0];
      v[i + 1] = u[i + 1];
      v[i + 6] = u[i + 6];
      v[i + 7] = u[i + 7];
    }

    v[34] = half_btf_sse4_1(&cospim8, &u[34], &cospi56, &u[61], &rnding, bit);
    v[35] = half_btf_sse4_1(&cospim8, &u[35], &cospi56, &u[60], &rnding, bit);
    v[36] = half_btf_sse4_1(&cospim56, &u[36], &cospim8, &u[59], &rnding, bit);
    v[37] = half_btf_sse4_1(&cospim56, &u[37], &cospim8, &u[58], &rnding, bit);
    v[42] = half_btf_sse4_1(&cospim40, &u[42], &cospi24, &u[53], &rnding, bit);
    v[43] = half_btf_sse4_1(&cospim40, &u[43], &cospi24, &u[52], &rnding, bit);
    v[44] = half_btf_sse4_1(&cospim24, &u[44], &cospim40, &u[51], &rnding, bit);
    v[45] = half_btf_sse4_1(&cospim24, &u[45], &cospim40, &u[50], &rnding, bit);
    v[50] = half_btf_sse4_1(&cospim40, &u[45], &cospi24, &u[50], &rnding, bit);
    v[51] = half_btf_sse4_1(&cospim40, &u[44], &cospi24, &u[51], &rnding, bit);
    v[52] = half_btf_sse4_1(&cospi24, &u[43], &cospi40, &u[52], &rnding, bit);
    v[53] = half_btf_sse4_1(&cospi24, &u[42], &cospi40, &u[53], &rnding, bit);
    v[58] = half_btf_sse4_1(&cospim8, &u[37], &cospi56, &u[58], &rnding, bit);
    v[59] = half_btf_sse4_1(&cospim8, &u[36], &cospi56, &u[59], &rnding, bit);
    v[60] = half_btf_sse4_1(&cospi56, &u[35], &cospi8, &u[60], &rnding, bit);
    v[61] = half_btf_sse4_1(&cospi56, &u[34], &cospi8, &u[61], &rnding, bit);

    // stage 7
    addsub_sse4_1(v[0], v[3], &u[0], &u[3], &clamp_lo, &clamp_hi);
    addsub_sse4_1(v[1], v[2], &u[1], &u[2], &clamp_lo, &clamp_hi);

    u[4] = v[4];
    u[7] = v[7];
    u[5] = half_btf_sse4_1(&cospim32, &v[5], &cospi32, &v[6], &rnding, bit);
    u[6] = half_btf_sse4_1(&cospi32, &v[5], &cospi32, &v[6], &rnding, bit);

    addsub_sse4_1(v[8], v[11], &u[8], &u[11], &clamp_lo, &clamp_hi);
    addsub_sse4_1(v[9], v[10], &u[9], &u[10], &clamp_lo, &clamp_hi);
    addsub_sse4_1(v[15], v[12], &u[15], &u[12], &clamp_lo, &clamp_hi);
    addsub_sse4_1(v[14], v[13], &u[14], &u[13], &clamp_lo, &clamp_hi);

    for (i = 16; i < 32; i += 8) {
      u[i + 0] = v[i + 0];
      u[i + 1] = v[i + 1];
      u[i + 6] = v[i + 6];
      u[i + 7] = v[i + 7];
    }

    u[18] = half_btf_sse4_1(&cospim16, &v[18], &cospi48, &v[29], &rnding, bit);
    u[19] = half_btf_sse4_1(&cospim16, &v[19], &cospi48, &v[28], &rnding, bit);
    u[20] = half_btf_sse4_1(&cospim48, &v[20], &cospim16, &v[27], &rnding, bit);
    u[21] = half_btf_sse4_1(&cospim48, &v[21], &cospim16, &v[26], &rnding, bit);
    u[26] = half_btf_sse4_1(&cospim16, &v[21], &cospi48, &v[26], &rnding, bit);
    u[27] = half_btf_sse4_1(&cospim16, &v[20], &cospi48, &v[27], &rnding, bit);
    u[28] = half_btf_sse4_1(&cospi48, &v[19], &cospi16, &v[28], &rnding, bit);
    u[29] = half_btf_sse4_1(&cospi48, &v[18], &cospi16, &v[29], &rnding, bit);

    for (i = 32; i < 64; i += 16) {
      for (j = i; j < i + 4; j++) {
        addsub_sse4_1(v[j], v[j ^ 7], &u[j], &u[j ^ 7], &clamp_lo, &clamp_hi);
        addsub_sse4_1(v[j ^ 15], v[j ^ 8], &u[j ^ 15], &u[j ^ 8], &clamp_lo,
                      &clamp_hi);
      }
    }

    // stage 8
    for (i = 0; i < 4; ++i) {
      addsub_sse4_1(u[i], u[7 - i], &v[i], &v[7 - i], &clamp_lo, &clamp_hi);
    }

    v[8] = u[8];
    v[9] = u[9];
    v[14] = u[14];
    v[15] = u[15];

    v[10] = half_btf_sse4_1(&cospim32, &u[10], &cospi32, &u[13], &rnding, bit);
    v[11] = half_btf_sse4_1(&cospim32, &u[11], &cospi32, &u[12], &rnding, bit);
    v[12] = half_btf_sse4_1(&cospi32, &u[11], &cospi32, &u[12], &rnding, bit);
    v[13] = half_btf_sse4_1(&cospi32, &u[10], &cospi32, &u[13], &rnding, bit);

    for (i = 16; i < 20; ++i) {
      addsub_sse4_1(u[i], u[i ^ 7], &v[i], &v[i ^ 7], &clamp_lo, &clamp_hi);
      addsub_sse4_1(u[i ^ 15], u[i ^ 8], &v[i ^ 15], &v[i ^ 8], &clamp_lo,
                    &clamp_hi);
    }

    for (i = 32; i < 36; ++i) {
      v[i] = u[i];
      v[i + 12] = u[i + 12];
      v[i + 16] = u[i + 16];
      v[i + 28] = u[i + 28];
    }

    v[36] = half_btf_sse4_1(&cospim16, &u[36], &cospi48, &u[59], &rnding, bit);
    v[37] = half_btf_sse4_1(&cospim16, &u[37], &cospi48, &u[58], &rnding, bit);
    v[38] = half_btf_sse4_1(&cospim16, &u[38], &cospi48, &u[57], &rnding, bit);
    v[39] = half_btf_sse4_1(&cospim16, &u[39], &cospi48, &u[56], &rnding, bit);
    v[40] = half_btf_sse4_1(&cospim48, &u[40], &cospim16, &u[55], &rnding, bit);
    v[41] = half_btf_sse4_1(&cospim48, &u[41], &cospim16, &u[54], &rnding, bit);
    v[42] = half_btf_sse4_1(&cospim48, &u[42], &cospim16, &u[53], &rnding, bit);
    v[43] = half_btf_sse4_1(&cospim48, &u[43], &cospim16, &u[52], &rnding, bit);
    v[52] = half_btf_sse4_1(&cospim16, &u[43], &cospi48, &u[52], &rnding, bit);
    v[53] = half_btf_sse4_1(&cospim16, &u[42], &cospi48, &u[53], &rnding, bit);
    v[54] = half_btf_sse4_1(&cospim16, &u[41], &cospi48, &u[54], &rnding, bit);
    v[55] = half_btf_sse4_1(&cospim16, &u[40], &cospi48, &u[55], &rnding, bit);
    v[56] = half_btf_sse4_1(&cospi48, &u[39], &cospi16, &u[56], &rnding, bit);
    v[57] = half_btf_sse4_1(&cospi48, &u[38], &cospi16, &u[57], &rnding, bit);
    v[58] = half_btf_sse4_1(&cospi48, &u[37], &cospi16, &u[58], &rnding, bit);
    v[59] = half_btf_sse4_1(&cospi48, &u[36], &cospi16, &u[59], &rnding, bit);

    // stage 9
    for (i = 0; i < 8; ++i) {
      addsub_sse4_1(v[i], v[15 - i], &u[i], &u[15 - i], &clamp_lo, &clamp_hi);
    }

    for (i = 16; i < 20; ++i) {
      u[i] = v[i];
      u[i + 12] = v[i + 12];
    }

    u[20] = half_btf_sse4_1(&cospim32, &v[20], &cospi32, &v[27], &rnding, bit);
    u[21] = half_btf_sse4_1(&cospim32, &v[21], &cospi32, &v[26], &rnding, bit);
    u[22] = half_btf_sse4_1(&cospim32, &v[22], &cospi32, &v[25], &rnding, bit);
    u[23] = half_btf_sse4_1(&cospim32, &v[23], &cospi32, &v[24], &rnding, bit);
    u[24] = half_btf_sse4_1(&cospi32, &v[23], &cospi32, &v[24], &rnding, bit);
    u[25] = half_btf_sse4_1(&cospi32, &v[22], &cospi32, &v[25], &rnding, bit);
    u[26] = half_btf_sse4_1(&cospi32, &v[21], &cospi32, &v[26], &rnding, bit);
    u[27] = half_btf_sse4_1(&cospi32, &v[20], &cospi32, &v[27], &rnding, bit);

    for (i = 32; i < 40; i++) {
      addsub_sse4_1(v[i], v[i ^ 15], &u[i], &u[i ^ 15], &clamp_lo, &clamp_hi);
    }

    for (i = 48; i < 56; i++) {
      addsub_sse4_1(v[i ^ 15], v[i], &u[i ^ 15], &u[i], &clamp_lo, &clamp_hi);
    }

    // stage 10
    for (i = 0; i < 16; i++) {
      addsub_sse4_1(u[i], u[31 - i], &v[i], &v[31 - i], &clamp_lo, &clamp_hi);
    }

    for (i = 32; i < 40; i++) v[i] = u[i];

    v[40] = half_btf_sse4_1(&cospim32, &u[40], &cospi32, &u[55], &rnding, bit);
    v[41] = half_btf_sse4_1(&cospim32, &u[41], &cospi32, &u[54], &rnding, bit);
    v[42] = half_btf_sse4_1(&cospim32, &u[42], &cospi32, &u[53], &rnding, bit);
    v[43] = half_btf_sse4_1(&cospim32, &u[43], &cospi32, &u[52], &rnding, bit);
    v[44] = half_btf_sse4_1(&cospim32, &u[44], &cospi32, &u[51], &rnding, bit);
    v[45] = half_btf_sse4_1(&cospim32, &u[45], &cospi32, &u[50], &rnding, bit);
    v[46] = half_btf_sse4_1(&cospim32, &u[46], &cospi32, &u[49], &rnding, bit);
    v[47] = half_btf_sse4_1(&cospim32, &u[47], &cospi32, &u[48], &rnding, bit);
    v[48] = half_btf_sse4_1(&cospi32, &u[47], &cospi32, &u[48], &rnding, bit);
    v[49] = half_btf_sse4_1(&cospi32, &u[46], &cospi32, &u[49], &rnding, bit);
    v[50] = half_btf_sse4_1(&cospi32, &u[45], &cospi32, &u[50], &rnding, bit);
    v[51] = half_btf_sse4_1(&cospi32, &u[44], &cospi32, &u[51], &rnding, bit);
    v[52] = half_btf_sse4_1(&cospi32, &u[43], &cospi32, &u[52], &rnding, bit);
    v[53] = half_btf_sse4_1(&cospi32, &u[42], &cospi32, &u[53], &rnding, bit);
    v[54] = half_btf_sse4_1(&cospi32, &u[41], &cospi32, &u[54], &rnding, bit);
    v[55] = half_btf_sse4_1(&cospi32, &u[40], &cospi32, &u[55], &rnding, bit);

    for (i = 56; i < 64; i++) v[i] = u[i];

    // stage 11
    for (i = 0; i < 32; i++) {
      addsub_sse4_1(v[i], v[63 - i], &out[(i)], &out[(63 - i)], &clamp_lo,
                    &clamp_hi);
    }

    if (!do_cols) {
      const int log_range_out = AOMMAX(16, bd + 6);
      const __m128i clamp_lo_out = _mm_set1_epi32(-(1 << (log_range_out - 1)));
      const __m128i clamp_hi_out =
          _mm_set1_epi32((1 << (log_range_out - 1)) - 1);
      for (i = 0; i < 64; i += 4) {
        round_shift_4x4(out + i, out_shift);
        highbd_clamp_epi32_sse4_1(out + i, out + i, &clamp_lo_out,
                                  &clamp_hi_out, 4);
      }
    }
  }
}

static void idct32x32_low1_sse4_1(__m128i *in, __m128i *out, int bit,
                                  int do_cols, int bd, int out_shift) {
  const int32_t *cospi = cospi_arr(bit);
  const __m128i cospi32 = _mm_set1_epi32(cospi[32]);
  const __m128i rounding = _mm_set1_epi32(1 << (bit - 1));
  const int log_range = AOMMAX(16, bd + (do_cols ? 6 : 8));
  __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
  __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
  __m128i bf1;

  // stage 0
  // stage 1
  bf1 = in[0];

  // stage 2
  // stage 3
  // stage 4
  // stage 5
  bf1 = half_btf_0_sse4_1(&cospi32, &bf1, &rounding, bit);

  // stage 6
  // stage 7
  // stage 8
  // stage 9
  if (do_cols) {
    bf1 = _mm_max_epi32(bf1, clamp_lo);
    bf1 = _mm_min_epi32(bf1, clamp_hi);
  } else {
    const int log_range_out = AOMMAX(16, bd + 6);
    assert(log_range_out - 1 < 32);  // Ensure left-shift doesn't overflow.
    clamp_lo = _mm_set1_epi32(-(1 << (log_range_out - 1)));
    clamp_hi = _mm_set1_epi32((1 << (log_range_out - 1)) - 1);
    if (out_shift != 0) {
      __m128i offset = _mm_set1_epi32((1 << out_shift) >> 1);
      bf1 = _mm_add_epi32(bf1, offset);
      bf1 = _mm_sra_epi32(bf1, _mm_cvtsi32_si128(out_shift));
    }
  }

  bf1 = _mm_max_epi32(bf1, clamp_lo);
  bf1 = _mm_min_epi32(bf1, clamp_hi);
  out[0] = bf1;
  out[1] = bf1;
  out[2] = bf1;
  out[3] = bf1;
  out[4] = bf1;
  out[5] = bf1;
  out[6] = bf1;
  out[7] = bf1;
  out[8] = bf1;
  out[9] = bf1;
  out[10] = bf1;
  out[11] = bf1;
  out[12] = bf1;
  out[13] = bf1;
  out[14] = bf1;
  out[15] = bf1;
  out[16] = bf1;
  out[17] = bf1;
  out[18] = bf1;
  out[19] = bf1;
  out[20] = bf1;
  out[21] = bf1;
  out[22] = bf1;
  out[23] = bf1;
  out[24] = bf1;
  out[25] = bf1;
  out[26] = bf1;
  out[27] = bf1;
  out[28] = bf1;
  out[29] = bf1;
  out[30] = bf1;
  out[31] = bf1;
}

static void idct32x32_low8_sse4_1(__m128i *in, __m128i *out, int bit,
                                  int do_cols, int bd, int out_shift) {
  const int32_t *cospi = cospi_arr(bit);
  const __m128i cospi62 = _mm_set1_epi32(cospi[62]);
  const __m128i cospi14 = _mm_set1_epi32(cospi[14]);
  const __m128i cospi54 = _mm_set1_epi32(cospi[54]);
  const __m128i cospi6 = _mm_set1_epi32(cospi[6]);
  const __m128i cospi10 = _mm_set1_epi32(cospi[10]);
  const __m128i cospi2 = _mm_set1_epi32(cospi[2]);
  const __m128i cospim58 = _mm_set1_epi32(-cospi[58]);
  const __m128i cospim50 = _mm_set1_epi32(-cospi[50]);
  const __m128i cospi60 = _mm_set1_epi32(cospi[60]);
  const __m128i cospi12 = _mm_set1_epi32(cospi[12]);
  const __m128i cospi4 = _mm_set1_epi32(cospi[4]);
  const __m128i cospim52 = _mm_set1_epi32(-cospi[52]);
  const __m128i cospi56 = _mm_set1_epi32(cospi[56]);
  const __m128i cospi24 = _mm_set1_epi32(cospi[24]);
  const __m128i cospi40 = _mm_set1_epi32(cospi[40]);
  const __m128i cospi8 = _mm_set1_epi32(cospi[8]);
  const __m128i cospim40 = _mm_set1_epi32(-cospi[40]);
  const __m128i cospim8 = _mm_set1_epi32(-cospi[8]);
  const __m128i cospim56 = _mm_set1_epi32(-cospi[56]);
  const __m128i cospim24 = _mm_set1_epi32(-cospi[24]);
  const __m128i cospi32 = _mm_set1_epi32(cospi[32]);
  const __m128i cospim32 = _mm_set1_epi32(-cospi[32]);
  const __m128i cospi48 = _mm_set1_epi32(cospi[48]);
  const __m128i cospim48 = _mm_set1_epi32(-cospi[48]);
  const __m128i cospi16 = _mm_set1_epi32(cospi[16]);
  const __m128i cospim16 = _mm_set1_epi32(-cospi[16]);
  const __m128i rounding = _mm_set1_epi32(1 << (bit - 1));
  const int log_range = AOMMAX(16, bd + (do_cols ? 6 : 8));
  const __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
  const __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
  __m128i bf1[32];

  // stage 0
  // stage 1
  bf1[0] = in[0];
  bf1[4] = in[4];
  bf1[8] = in[2];
  bf1[12] = in[6];
  bf1[16] = in[1];
  bf1[20] = in[5];
  bf1[24] = in[3];
  bf1[28] = in[7];

  // stage 2
  bf1[31] = half_btf_0_sse4_1(&cospi2, &bf1[16], &rounding, bit);
  bf1[16] = half_btf_0_sse4_1(&cospi62, &bf1[16], &rounding, bit);
  bf1[19] = half_btf_0_sse4_1(&cospim50, &bf1[28], &rounding, bit);
  bf1[28] = half_btf_0_sse4_1(&cospi14, &bf1[28], &rounding, bit);
  bf1[27] = half_btf_0_sse4_1(&cospi10, &bf1[20], &rounding, bit);
  bf1[20] = half_btf_0_sse4_1(&cospi54, &bf1[20], &rounding, bit);
  bf1[23] = half_btf_0_sse4_1(&cospim58, &bf1[24], &rounding, bit);
  bf1[24] = half_btf_0_sse4_1(&cospi6, &bf1[24], &rounding, bit);

  // stage 3
  bf1[15] = half_btf_0_sse4_1(&cospi4, &bf1[8], &rounding, bit);
  bf1[8] = half_btf_0_sse4_1(&cospi60, &bf1[8], &rounding, bit);

  bf1[11] = half_btf_0_sse4_1(&cospim52, &bf1[12], &rounding, bit);
  bf1[12] = half_btf_0_sse4_1(&cospi12, &bf1[12], &rounding, bit);
  bf1[17] = bf1[16];
  bf1[18] = bf1[19];
  bf1[21] = bf1[20];
  bf1[22] = bf1[23];
  bf1[25] = bf1[24];
  bf1[26] = bf1[27];
  bf1[29] = bf1[28];
  bf1[30] = bf1[31];

  // stage 4 :
  bf1[7] = half_btf_0_sse4_1(&cospi8, &bf1[4], &rounding, bit);
  bf1[4] = half_btf_0_sse4_1(&cospi56, &bf1[4], &rounding, bit);

  bf1[9] = bf1[8];
  bf1[10] = bf1[11];
  bf1[13] = bf1[12];
  bf1[14] = bf1[15];

  idct32_stage4_sse4_1(bf1, &cospim8, &cospi56, &cospi8, &cospim56, &cospim40,
                       &cospi24, &cospi40, &cospim24, &rounding, bit);

  // stage 5
  bf1[0] = half_btf_0_sse4_1(&cospi32, &bf1[0], &rounding, bit);
  bf1[1] = bf1[0];
  bf1[5] = bf1[4];
  bf1[6] = bf1[7];

  idct32_stage5_sse4_1(bf1, &cospim16, &cospi48, &cospi16, &cospim48, &clamp_lo,
                       &clamp_hi, &rounding, bit);

  // stage 6
  bf1[3] = bf1[0];
  bf1[2] = bf1[1];

  idct32_stage6_sse4_1(bf1, &cospim32, &cospi32, &cospim16, &cospi48, &cospi16,
                       &cospim48, &clamp_lo, &clamp_hi, &rounding, bit);

  // stage 7
  idct32_stage7_sse4_1(bf1, &cospim32, &cospi32, &clamp_lo, &clamp_hi,
                       &rounding, bit);

  // stage 8
  idct32_stage8_sse4_1(bf1, &cospim32, &cospi32, &clamp_lo, &clamp_hi,
                       &rounding, bit);

  // stage 9
  idct32_stage9_sse4_1(bf1, out, do_cols, bd, out_shift, &clamp_lo, &clamp_hi);
}

static void idct32x32_low16_sse4_1(__m128i *in, __m128i *out, int bit,
                                   int do_cols, int bd, int out_shift) {
  const int32_t *cospi = cospi_arr(bit);
  const __m128i cospi62 = _mm_set1_epi32(cospi[62]);
  const __m128i cospi30 = _mm_set1_epi32(cospi[30]);
  const __m128i cospi46 = _mm_set1_epi32(cospi[46]);
  const __m128i cospi14 = _mm_set1_epi32(cospi[14]);
  const __m128i cospi54 = _mm_set1_epi32(cospi[54]);
  const __m128i cospi22 = _mm_set1_epi32(cospi[22]);
  const __m128i cospi38 = _mm_set1_epi32(cospi[38]);
  const __m128i cospi6 = _mm_set1_epi32(cospi[6]);
  const __m128i cospi26 = _mm_set1_epi32(cospi[26]);
  const __m128i cospi10 = _mm_set1_epi32(cospi[10]);
  const __m128i cospi18 = _mm_set1_epi32(cospi[18]);
  const __m128i cospi2 = _mm_set1_epi32(cospi[2]);
  const __m128i cospim58 = _mm_set1_epi32(-cospi[58]);
  const __m128i cospim42 = _mm_set1_epi32(-cospi[42]);
  const __m128i cospim50 = _mm_set1_epi32(-cospi[50]);
  const __m128i cospim34 = _mm_set1_epi32(-cospi[34]);
  const __m128i cospi60 = _mm_set1_epi32(cospi[60]);
  const __m128i cospi28 = _mm_set1_epi32(cospi[28]);
  const __m128i cospi44 = _mm_set1_epi32(cospi[44]);
  const __m128i cospi12 = _mm_set1_epi32(cospi[12]);
  const __m128i cospi20 = _mm_set1_epi32(cospi[20]);
  const __m128i cospi4 = _mm_set1_epi32(cospi[4]);
  const __m128i cospim52 = _mm_set1_epi32(-cospi[52]);
  const __m128i cospim36 = _mm_set1_epi32(-cospi[36]);
  const __m128i cospi56 = _mm_set1_epi32(cospi[56]);
  const __m128i cospi24 = _mm_set1_epi32(cospi[24]);
  const __m128i cospi40 = _mm_set1_epi32(cospi[40]);
  const __m128i cospi8 = _mm_set1_epi32(cospi[8]);
  const __m128i cospim40 = _mm_set1_epi32(-cospi[40]);
  const __m128i cospim8 = _mm_set1_epi32(-cospi[8]);
  const __m128i cospim56 = _mm_set1_epi32(-cospi[56]);
  const __m128i cospim24 = _mm_set1_epi32(-cospi[24]);
  const __m128i cospi32 = _mm_set1_epi32(cospi[32]);
  const __m128i cospim32 = _mm_set1_epi32(-cospi[32]);
  const __m128i cospi48 = _mm_set1_epi32(cospi[48]);
  const __m128i cospim48 = _mm_set1_epi32(-cospi[48]);
  const __m128i cospi16 = _mm_set1_epi32(cospi[16]);
  const __m128i cospim16 = _mm_set1_epi32(-cospi[16]);
  const __m128i rounding = _mm_set1_epi32(1 << (bit - 1));
  const int log_range = AOMMAX(16, bd + (do_cols ? 6 : 8));
  const __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
  const __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
  __m128i bf1[32];

  // stage 0
  // stage 1

  bf1[0] = in[0];
  bf1[2] = in[8];
  bf1[4] = in[4];
  bf1[6] = in[12];
  bf1[8] = in[2];
  bf1[10] = in[10];
  bf1[12] = in[6];
  bf1[14] = in[14];
  bf1[16] = in[1];
  bf1[18] = in[9];
  bf1[20] = in[5];
  bf1[22] = in[13];
  bf1[24] = in[3];
  bf1[26] = in[11];
  bf1[28] = in[7];
  bf1[30] = in[15];

  // stage 2
  bf1[31] = half_btf_0_sse4_1(&cospi2, &bf1[16], &rounding, bit);
  bf1[16] = half_btf_0_sse4_1(&cospi62, &bf1[16], &rounding, bit);
  bf1[17] = half_btf_0_sse4_1(&cospim34, &bf1[30], &rounding, bit);
  bf1[30] = half_btf_0_sse4_1(&cospi30, &bf1[30], &rounding, bit);
  bf1[29] = half_btf_0_sse4_1(&cospi18, &bf1[18], &rounding, bit);
  bf1[18] = half_btf_0_sse4_1(&cospi46, &bf1[18], &rounding, bit);
  bf1[19] = half_btf_0_sse4_1(&cospim50, &bf1[28], &rounding, bit);
  bf1[28] = half_btf_0_sse4_1(&cospi14, &bf1[28], &rounding, bit);
  bf1[27] = half_btf_0_sse4_1(&cospi10, &bf1[20], &rounding, bit);
  bf1[20] = half_btf_0_sse4_1(&cospi54, &bf1[20], &rounding, bit);
  bf1[21] = half_btf_0_sse4_1(&cospim42, &bf1[26], &rounding, bit);
  bf1[26] = half_btf_0_sse4_1(&cospi22, &bf1[26], &rounding, bit);
  bf1[25] = half_btf_0_sse4_1(&cospi26, &bf1[22], &rounding, bit);
  bf1[22] = half_btf_0_sse4_1(&cospi38, &bf1[22], &rounding, bit);
  bf1[23] = half_btf_0_sse4_1(&cospim58, &bf1[24], &rounding, bit);
  bf1[24] = half_btf_0_sse4_1(&cospi6, &bf1[24], &rounding, bit);

  // stage 3
  bf1[15] = half_btf_0_sse4_1(&cospi4, &bf1[8], &rounding, bit);
  bf1[8] = half_btf_0_sse4_1(&cospi60, &bf1[8], &rounding, bit);
  bf1[9] = half_btf_0_sse4_1(&cospim36, &bf1[14], &rounding, bit);
  bf1[14] = half_btf_0_sse4_1(&cospi28, &bf1[14], &rounding, bit);
  bf1[13] = half_btf_0_sse4_1(&cospi20, &bf1[10], &rounding, bit);
  bf1[10] = half_btf_0_sse4_1(&cospi44, &bf1[10], &rounding, bit);
  bf1[11] = half_btf_0_sse4_1(&cospim52, &bf1[12], &rounding, bit);
  bf1[12] = half_btf_0_sse4_1(&cospi12, &bf1[12], &rounding, bit);

  addsub_sse4_1(bf1[16], bf1[17], bf1 + 16, bf1 + 17, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[19], bf1[18], bf1 + 19, bf1 + 18, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[20], bf1[21], bf1 + 20, bf1 + 21, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[23], bf1[22], bf1 + 23, bf1 + 22, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[24], bf1[25], bf1 + 24, bf1 + 25, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[27], bf1[26], bf1 + 27, bf1 + 26, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[28], bf1[29], bf1 + 28, bf1 + 29, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[31], bf1[30], bf1 + 31, bf1 + 30, &clamp_lo, &clamp_hi);
  // stage 4
  bf1[7] = half_btf_0_sse4_1(&cospi8, &bf1[4], &rounding, bit);
  bf1[4] = half_btf_0_sse4_1(&cospi56, &bf1[4], &rounding, bit);
  bf1[5] = half_btf_0_sse4_1(&cospim40, &bf1[6], &rounding, bit);
  bf1[6] = half_btf_0_sse4_1(&cospi24, &bf1[6], &rounding, bit);

  addsub_sse4_1(bf1[8], bf1[9], bf1 + 8, bf1 + 9, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[11], bf1[10], bf1 + 11, bf1 + 10, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[12], bf1[13], bf1 + 12, bf1 + 13, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[15], bf1[14], bf1 + 15, bf1 + 14, &clamp_lo, &clamp_hi);

  idct32_stage4_sse4_1(bf1, &cospim8, &cospi56, &cospi8, &cospim56, &cospim40,
                       &cospi24, &cospi40, &cospim24, &rounding, bit);

  // stage 5
  bf1[0] = half_btf_0_sse4_1(&cospi32, &bf1[0], &rounding, bit);
  bf1[1] = bf1[0];
  bf1[3] = half_btf_0_sse4_1(&cospi16, &bf1[2], &rounding, bit);
  bf1[2] = half_btf_0_sse4_1(&cospi48, &bf1[2], &rounding, bit);

  addsub_sse4_1(bf1[4], bf1[5], bf1 + 4, bf1 + 5, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[7], bf1[6], bf1 + 7, bf1 + 6, &clamp_lo, &clamp_hi);

  idct32_stage5_sse4_1(bf1, &cospim16, &cospi48, &cospi16, &cospim48, &clamp_lo,
                       &clamp_hi, &rounding, bit);

  // stage 6
  addsub_sse4_1(bf1[0], bf1[3], bf1 + 0, bf1 + 3, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[1], bf1[2], bf1 + 1, bf1 + 2, &clamp_lo, &clamp_hi);

  idct32_stage6_sse4_1(bf1, &cospim32, &cospi32, &cospim16, &cospi48, &cospi16,
                       &cospim48, &clamp_lo, &clamp_hi, &rounding, bit);

  // stage 7
  idct32_stage7_sse4_1(bf1, &cospim32, &cospi32, &clamp_lo, &clamp_hi,
                       &rounding, bit);

  // stage 8
  idct32_stage8_sse4_1(bf1, &cospim32, &cospi32, &clamp_lo, &clamp_hi,
                       &rounding, bit);
  // stage 9
  idct32_stage9_sse4_1(bf1, out, do_cols, bd, out_shift, &clamp_lo, &clamp_hi);
}

static void idct32x32_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                             int bd, int out_shift) {
  const int32_t *cospi = cospi_arr(bit);
  const __m128i cospi62 = _mm_set1_epi32(cospi[62]);
  const __m128i cospi30 = _mm_set1_epi32(cospi[30]);
  const __m128i cospi46 = _mm_set1_epi32(cospi[46]);
  const __m128i cospi14 = _mm_set1_epi32(cospi[14]);
  const __m128i cospi54 = _mm_set1_epi32(cospi[54]);
  const __m128i cospi22 = _mm_set1_epi32(cospi[22]);
  const __m128i cospi38 = _mm_set1_epi32(cospi[38]);
  const __m128i cospi6 = _mm_set1_epi32(cospi[6]);
  const __m128i cospi58 = _mm_set1_epi32(cospi[58]);
  const __m128i cospi26 = _mm_set1_epi32(cospi[26]);
  const __m128i cospi42 = _mm_set1_epi32(cospi[42]);
  const __m128i cospi10 = _mm_set1_epi32(cospi[10]);
  const __m128i cospi50 = _mm_set1_epi32(cospi[50]);
  const __m128i cospi18 = _mm_set1_epi32(cospi[18]);
  const __m128i cospi34 = _mm_set1_epi32(cospi[34]);
  const __m128i cospi2 = _mm_set1_epi32(cospi[2]);
  const __m128i cospim58 = _mm_set1_epi32(-cospi[58]);
  const __m128i cospim26 = _mm_set1_epi32(-cospi[26]);
  const __m128i cospim42 = _mm_set1_epi32(-cospi[42]);
  const __m128i cospim10 = _mm_set1_epi32(-cospi[10]);
  const __m128i cospim50 = _mm_set1_epi32(-cospi[50]);
  const __m128i cospim18 = _mm_set1_epi32(-cospi[18]);
  const __m128i cospim34 = _mm_set1_epi32(-cospi[34]);
  const __m128i cospim2 = _mm_set1_epi32(-cospi[2]);
  const __m128i cospi60 = _mm_set1_epi32(cospi[60]);
  const __m128i cospi28 = _mm_set1_epi32(cospi[28]);
  const __m128i cospi44 = _mm_set1_epi32(cospi[44]);
  const __m128i cospi12 = _mm_set1_epi32(cospi[12]);
  const __m128i cospi52 = _mm_set1_epi32(cospi[52]);
  const __m128i cospi20 = _mm_set1_epi32(cospi[20]);
  const __m128i cospi36 = _mm_set1_epi32(cospi[36]);
  const __m128i cospi4 = _mm_set1_epi32(cospi[4]);
  const __m128i cospim52 = _mm_set1_epi32(-cospi[52]);
  const __m128i cospim20 = _mm_set1_epi32(-cospi[20]);
  const __m128i cospim36 = _mm_set1_epi32(-cospi[36]);
  const __m128i cospim4 = _mm_set1_epi32(-cospi[4]);
  const __m128i cospi56 = _mm_set1_epi32(cospi[56]);
  const __m128i cospi24 = _mm_set1_epi32(cospi[24]);
  const __m128i cospi40 = _mm_set1_epi32(cospi[40]);
  const __m128i cospi8 = _mm_set1_epi32(cospi[8]);
  const __m128i cospim40 = _mm_set1_epi32(-cospi[40]);
  const __m128i cospim8 = _mm_set1_epi32(-cospi[8]);
  const __m128i cospim56 = _mm_set1_epi32(-cospi[56]);
  const __m128i cospim24 = _mm_set1_epi32(-cospi[24]);
  const __m128i cospi32 = _mm_set1_epi32(cospi[32]);
  const __m128i cospim32 = _mm_set1_epi32(-cospi[32]);
  const __m128i cospi48 = _mm_set1_epi32(cospi[48]);
  const __m128i cospim48 = _mm_set1_epi32(-cospi[48]);
  const __m128i cospi16 = _mm_set1_epi32(cospi[16]);
  const __m128i cospim16 = _mm_set1_epi32(-cospi[16]);
  const __m128i rounding = _mm_set1_epi32(1 << (bit - 1));
  const int log_range = AOMMAX(16, bd + (do_cols ? 6 : 8));
  const __m128i clamp_lo = _mm_set1_epi32(-(1 << (log_range - 1)));
  const __m128i clamp_hi = _mm_set1_epi32((1 << (log_range - 1)) - 1);
  __m128i bf1[32], bf0[32];

  // stage 0
  // stage 1
  bf1[0] = in[0];
  bf1[1] = in[16];
  bf1[2] = in[8];
  bf1[3] = in[24];
  bf1[4] = in[4];
  bf1[5] = in[20];
  bf1[6] = in[12];
  bf1[7] = in[28];
  bf1[8] = in[2];
  bf1[9] = in[18];
  bf1[10] = in[10];
  bf1[11] = in[26];
  bf1[12] = in[6];
  bf1[13] = in[22];
  bf1[14] = in[14];
  bf1[15] = in[30];
  bf1[16] = in[1];
  bf1[17] = in[17];
  bf1[18] = in[9];
  bf1[19] = in[25];
  bf1[20] = in[5];
  bf1[21] = in[21];
  bf1[22] = in[13];
  bf1[23] = in[29];
  bf1[24] = in[3];
  bf1[25] = in[19];
  bf1[26] = in[11];
  bf1[27] = in[27];
  bf1[28] = in[7];
  bf1[29] = in[23];
  bf1[30] = in[15];
  bf1[31] = in[31];

  // stage 2
  bf0[0] = bf1[0];
  bf0[1] = bf1[1];
  bf0[2] = bf1[2];
  bf0[3] = bf1[3];
  bf0[4] = bf1[4];
  bf0[5] = bf1[5];
  bf0[6] = bf1[6];
  bf0[7] = bf1[7];
  bf0[8] = bf1[8];
  bf0[9] = bf1[9];
  bf0[10] = bf1[10];
  bf0[11] = bf1[11];
  bf0[12] = bf1[12];
  bf0[13] = bf1[13];
  bf0[14] = bf1[14];
  bf0[15] = bf1[15];
  bf0[16] =
      half_btf_sse4_1(&cospi62, &bf1[16], &cospim2, &bf1[31], &rounding, bit);
  bf0[17] =
      half_btf_sse4_1(&cospi30, &bf1[17], &cospim34, &bf1[30], &rounding, bit);
  bf0[18] =
      half_btf_sse4_1(&cospi46, &bf1[18], &cospim18, &bf1[29], &rounding, bit);
  bf0[19] =
      half_btf_sse4_1(&cospi14, &bf1[19], &cospim50, &bf1[28], &rounding, bit);
  bf0[20] =
      half_btf_sse4_1(&cospi54, &bf1[20], &cospim10, &bf1[27], &rounding, bit);
  bf0[21] =
      half_btf_sse4_1(&cospi22, &bf1[21], &cospim42, &bf1[26], &rounding, bit);
  bf0[22] =
      half_btf_sse4_1(&cospi38, &bf1[22], &cospim26, &bf1[25], &rounding, bit);
  bf0[23] =
      half_btf_sse4_1(&cospi6, &bf1[23], &cospim58, &bf1[24], &rounding, bit);
  bf0[24] =
      half_btf_sse4_1(&cospi58, &bf1[23], &cospi6, &bf1[24], &rounding, bit);
  bf0[25] =
      half_btf_sse4_1(&cospi26, &bf1[22], &cospi38, &bf1[25], &rounding, bit);
  bf0[26] =
      half_btf_sse4_1(&cospi42, &bf1[21], &cospi22, &bf1[26], &rounding, bit);
  bf0[27] =
      half_btf_sse4_1(&cospi10, &bf1[20], &cospi54, &bf1[27], &rounding, bit);
  bf0[28] =
      half_btf_sse4_1(&cospi50, &bf1[19], &cospi14, &bf1[28], &rounding, bit);
  bf0[29] =
      half_btf_sse4_1(&cospi18, &bf1[18], &cospi46, &bf1[29], &rounding, bit);
  bf0[30] =
      half_btf_sse4_1(&cospi34, &bf1[17], &cospi30, &bf1[30], &rounding, bit);
  bf0[31] =
      half_btf_sse4_1(&cospi2, &bf1[16], &cospi62, &bf1[31], &rounding, bit);

  // stage 3
  bf1[0] = bf0[0];
  bf1[1] = bf0[1];
  bf1[2] = bf0[2];
  bf1[3] = bf0[3];
  bf1[4] = bf0[4];
  bf1[5] = bf0[5];
  bf1[6] = bf0[6];
  bf1[7] = bf0[7];
  bf1[8] =
      half_btf_sse4_1(&cospi60, &bf0[8], &cospim4, &bf0[15], &rounding, bit);
  bf1[9] =
      half_btf_sse4_1(&cospi28, &bf0[9], &cospim36, &bf0[14], &rounding, bit);
  bf1[10] =
      half_btf_sse4_1(&cospi44, &bf0[10], &cospim20, &bf0[13], &rounding, bit);
  bf1[11] =
      half_btf_sse4_1(&cospi12, &bf0[11], &cospim52, &bf0[12], &rounding, bit);
  bf1[12] =
      half_btf_sse4_1(&cospi52, &bf0[11], &cospi12, &bf0[12], &rounding, bit);
  bf1[13] =
      half_btf_sse4_1(&cospi20, &bf0[10], &cospi44, &bf0[13], &rounding, bit);
  bf1[14] =
      half_btf_sse4_1(&cospi36, &bf0[9], &cospi28, &bf0[14], &rounding, bit);
  bf1[15] =
      half_btf_sse4_1(&cospi4, &bf0[8], &cospi60, &bf0[15], &rounding, bit);

  addsub_sse4_1(bf0[16], bf0[17], bf1 + 16, bf1 + 17, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[19], bf0[18], bf1 + 19, bf1 + 18, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[20], bf0[21], bf1 + 20, bf1 + 21, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[23], bf0[22], bf1 + 23, bf1 + 22, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[24], bf0[25], bf1 + 24, bf1 + 25, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[27], bf0[26], bf1 + 27, bf1 + 26, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[28], bf0[29], bf1 + 28, bf1 + 29, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[31], bf0[30], bf1 + 31, bf1 + 30, &clamp_lo, &clamp_hi);

  // stage 4
  bf0[0] = bf1[0];
  bf0[1] = bf1[1];
  bf0[2] = bf1[2];
  bf0[3] = bf1[3];
  bf0[4] =
      half_btf_sse4_1(&cospi56, &bf1[4], &cospim8, &bf1[7], &rounding, bit);
  bf0[5] =
      half_btf_sse4_1(&cospi24, &bf1[5], &cospim40, &bf1[6], &rounding, bit);
  bf0[6] =
      half_btf_sse4_1(&cospi40, &bf1[5], &cospi24, &bf1[6], &rounding, bit);
  bf0[7] = half_btf_sse4_1(&cospi8, &bf1[4], &cospi56, &bf1[7], &rounding, bit);

  addsub_sse4_1(bf1[8], bf1[9], bf0 + 8, bf0 + 9, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[11], bf1[10], bf0 + 11, bf0 + 10, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[12], bf1[13], bf0 + 12, bf0 + 13, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[15], bf1[14], bf0 + 15, bf0 + 14, &clamp_lo, &clamp_hi);

  bf0[16] = bf1[16];
  bf0[17] =
      half_btf_sse4_1(&cospim8, &bf1[17], &cospi56, &bf1[30], &rounding, bit);
  bf0[18] =
      half_btf_sse4_1(&cospim56, &bf1[18], &cospim8, &bf1[29], &rounding, bit);
  bf0[19] = bf1[19];
  bf0[20] = bf1[20];
  bf0[21] =
      half_btf_sse4_1(&cospim40, &bf1[21], &cospi24, &bf1[26], &rounding, bit);
  bf0[22] =
      half_btf_sse4_1(&cospim24, &bf1[22], &cospim40, &bf1[25], &rounding, bit);
  bf0[23] = bf1[23];
  bf0[24] = bf1[24];
  bf0[25] =
      half_btf_sse4_1(&cospim40, &bf1[22], &cospi24, &bf1[25], &rounding, bit);
  bf0[26] =
      half_btf_sse4_1(&cospi24, &bf1[21], &cospi40, &bf1[26], &rounding, bit);
  bf0[27] = bf1[27];
  bf0[28] = bf1[28];
  bf0[29] =
      half_btf_sse4_1(&cospim8, &bf1[18], &cospi56, &bf1[29], &rounding, bit);
  bf0[30] =
      half_btf_sse4_1(&cospi56, &bf1[17], &cospi8, &bf1[30], &rounding, bit);
  bf0[31] = bf1[31];

  // stage 5
  bf1[0] =
      half_btf_sse4_1(&cospi32, &bf0[0], &cospi32, &bf0[1], &rounding, bit);
  bf1[1] =
      half_btf_sse4_1(&cospi32, &bf0[0], &cospim32, &bf0[1], &rounding, bit);
  bf1[2] =
      half_btf_sse4_1(&cospi48, &bf0[2], &cospim16, &bf0[3], &rounding, bit);
  bf1[3] =
      half_btf_sse4_1(&cospi16, &bf0[2], &cospi48, &bf0[3], &rounding, bit);
  addsub_sse4_1(bf0[4], bf0[5], bf1 + 4, bf1 + 5, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[7], bf0[6], bf1 + 7, bf1 + 6, &clamp_lo, &clamp_hi);
  bf1[8] = bf0[8];
  bf1[9] =
      half_btf_sse4_1(&cospim16, &bf0[9], &cospi48, &bf0[14], &rounding, bit);
  bf1[10] =
      half_btf_sse4_1(&cospim48, &bf0[10], &cospim16, &bf0[13], &rounding, bit);
  bf1[11] = bf0[11];
  bf1[12] = bf0[12];
  bf1[13] =
      half_btf_sse4_1(&cospim16, &bf0[10], &cospi48, &bf0[13], &rounding, bit);
  bf1[14] =
      half_btf_sse4_1(&cospi48, &bf0[9], &cospi16, &bf0[14], &rounding, bit);
  bf1[15] = bf0[15];
  addsub_sse4_1(bf0[16], bf0[19], bf1 + 16, bf1 + 19, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[17], bf0[18], bf1 + 17, bf1 + 18, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[23], bf0[20], bf1 + 23, bf1 + 20, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[22], bf0[21], bf1 + 22, bf1 + 21, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[24], bf0[27], bf1 + 24, bf1 + 27, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[25], bf0[26], bf1 + 25, bf1 + 26, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[31], bf0[28], bf1 + 31, bf1 + 28, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[30], bf0[29], bf1 + 30, bf1 + 29, &clamp_lo, &clamp_hi);

  // stage 6
  addsub_sse4_1(bf1[0], bf1[3], bf0 + 0, bf0 + 3, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[1], bf1[2], bf0 + 1, bf0 + 2, &clamp_lo, &clamp_hi);
  bf0[4] = bf1[4];
  bf0[5] =
      half_btf_sse4_1(&cospim32, &bf1[5], &cospi32, &bf1[6], &rounding, bit);
  bf0[6] =
      half_btf_sse4_1(&cospi32, &bf1[5], &cospi32, &bf1[6], &rounding, bit);
  bf0[7] = bf1[7];
  addsub_sse4_1(bf1[8], bf1[11], bf0 + 8, bf0 + 11, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[9], bf1[10], bf0 + 9, bf0 + 10, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[15], bf1[12], bf0 + 15, bf0 + 12, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[14], bf1[13], bf0 + 14, bf0 + 13, &clamp_lo, &clamp_hi);
  bf0[16] = bf1[16];
  bf0[17] = bf1[17];
  bf0[18] =
      half_btf_sse4_1(&cospim16, &bf1[18], &cospi48, &bf1[29], &rounding, bit);
  bf0[19] =
      half_btf_sse4_1(&cospim16, &bf1[19], &cospi48, &bf1[28], &rounding, bit);
  bf0[20] =
      half_btf_sse4_1(&cospim48, &bf1[20], &cospim16, &bf1[27], &rounding, bit);
  bf0[21] =
      half_btf_sse4_1(&cospim48, &bf1[21], &cospim16, &bf1[26], &rounding, bit);
  bf0[22] = bf1[22];
  bf0[23] = bf1[23];
  bf0[24] = bf1[24];
  bf0[25] = bf1[25];
  bf0[26] =
      half_btf_sse4_1(&cospim16, &bf1[21], &cospi48, &bf1[26], &rounding, bit);
  bf0[27] =
      half_btf_sse4_1(&cospim16, &bf1[20], &cospi48, &bf1[27], &rounding, bit);
  bf0[28] =
      half_btf_sse4_1(&cospi48, &bf1[19], &cospi16, &bf1[28], &rounding, bit);
  bf0[29] =
      half_btf_sse4_1(&cospi48, &bf1[18], &cospi16, &bf1[29], &rounding, bit);
  bf0[30] = bf1[30];
  bf0[31] = bf1[31];

  // stage 7
  addsub_sse4_1(bf0[0], bf0[7], bf1 + 0, bf1 + 7, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[1], bf0[6], bf1 + 1, bf1 + 6, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[2], bf0[5], bf1 + 2, bf1 + 5, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[3], bf0[4], bf1 + 3, bf1 + 4, &clamp_lo, &clamp_hi);
  bf1[8] = bf0[8];
  bf1[9] = bf0[9];
  bf1[10] =
      half_btf_sse4_1(&cospim32, &bf0[10], &cospi32, &bf0[13], &rounding, bit);
  bf1[11] =
      half_btf_sse4_1(&cospim32, &bf0[11], &cospi32, &bf0[12], &rounding, bit);
  bf1[12] =
      half_btf_sse4_1(&cospi32, &bf0[11], &cospi32, &bf0[12], &rounding, bit);
  bf1[13] =
      half_btf_sse4_1(&cospi32, &bf0[10], &cospi32, &bf0[13], &rounding, bit);
  bf1[14] = bf0[14];
  bf1[15] = bf0[15];
  addsub_sse4_1(bf0[16], bf0[23], bf1 + 16, bf1 + 23, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[17], bf0[22], bf1 + 17, bf1 + 22, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[18], bf0[21], bf1 + 18, bf1 + 21, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[19], bf0[20], bf1 + 19, bf1 + 20, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[31], bf0[24], bf1 + 31, bf1 + 24, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[30], bf0[25], bf1 + 30, bf1 + 25, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[29], bf0[26], bf1 + 29, bf1 + 26, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[28], bf0[27], bf1 + 28, bf1 + 27, &clamp_lo, &clamp_hi);

  // stage 8
  addsub_sse4_1(bf1[0], bf1[15], bf0 + 0, bf0 + 15, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[1], bf1[14], bf0 + 1, bf0 + 14, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[2], bf1[13], bf0 + 2, bf0 + 13, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[3], bf1[12], bf0 + 3, bf0 + 12, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[4], bf1[11], bf0 + 4, bf0 + 11, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[5], bf1[10], bf0 + 5, bf0 + 10, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[6], bf1[9], bf0 + 6, bf0 + 9, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf1[7], bf1[8], bf0 + 7, bf0 + 8, &clamp_lo, &clamp_hi);
  bf0[16] = bf1[16];
  bf0[17] = bf1[17];
  bf0[18] = bf1[18];
  bf0[19] = bf1[19];
  bf0[20] =
      half_btf_sse4_1(&cospim32, &bf1[20], &cospi32, &bf1[27], &rounding, bit);
  bf0[21] =
      half_btf_sse4_1(&cospim32, &bf1[21], &cospi32, &bf1[26], &rounding, bit);
  bf0[22] =
      half_btf_sse4_1(&cospim32, &bf1[22], &cospi32, &bf1[25], &rounding, bit);
  bf0[23] =
      half_btf_sse4_1(&cospim32, &bf1[23], &cospi32, &bf1[24], &rounding, bit);
  bf0[24] =
      half_btf_sse4_1(&cospi32, &bf1[23], &cospi32, &bf1[24], &rounding, bit);
  bf0[25] =
      half_btf_sse4_1(&cospi32, &bf1[22], &cospi32, &bf1[25], &rounding, bit);
  bf0[26] =
      half_btf_sse4_1(&cospi32, &bf1[21], &cospi32, &bf1[26], &rounding, bit);
  bf0[27] =
      half_btf_sse4_1(&cospi32, &bf1[20], &cospi32, &bf1[27], &rounding, bit);
  bf0[28] = bf1[28];
  bf0[29] = bf1[29];
  bf0[30] = bf1[30];
  bf0[31] = bf1[31];

  // stage 9
  addsub_sse4_1(bf0[0], bf0[31], out + 0, out + 31, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[1], bf0[30], out + 1, out + 30, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[2], bf0[29], out + 2, out + 29, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[3], bf0[28], out + 3, out + 28, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[4], bf0[27], out + 4, out + 27, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[5], bf0[26], out + 5, out + 26, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[6], bf0[25], out + 6, out + 25, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[7], bf0[24], out + 7, out + 24, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[8], bf0[23], out + 8, out + 23, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[9], bf0[22], out + 9, out + 22, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[10], bf0[21], out + 10, out + 21, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[11], bf0[20], out + 11, out + 20, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[12], bf0[19], out + 12, out + 19, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[13], bf0[18], out + 13, out + 18, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[14], bf0[17], out + 14, out + 17, &clamp_lo, &clamp_hi);
  addsub_sse4_1(bf0[15], bf0[16], out + 15, out + 16, &clamp_lo, &clamp_hi);

  if (!do_cols) {
    const int log_range_out = AOMMAX(16, bd + 6);
    assert(log_range_out - 1 < 32);  // Ensure left-shift doesn't overflow.
    const __m128i clamp_lo_out = _mm_set1_epi32(-(1 << (log_range_out - 1)));
    const __m128i clamp_hi_out = _mm_set1_epi32((1 << (log_range_out - 1)) - 1);
    round_shift_8x8(out, out_shift);
    round_shift_8x8(out + 16, out_shift);
    highbd_clamp_epi32_sse4_1(out, out, &clamp_lo_out, &clamp_hi_out, 32);
  }
}

void av1_highbd_inv_txfm_add_8x8_sse4_1(const tran_low_t *input, uint16_t *dest,
                                        int stride,
                                        const TxfmParam *txfm_param) {
  int bd = txfm_param->bd;
  const TX_TYPE tx_type = txfm_param->tx_type;
  const int32_t *src = cast_to_int32(input);
  switch (tx_type) {
    case IDTX:
    case H_DCT:
    case H_ADST:
    case H_FLIPADST:
    case V_DCT:
    case V_ADST:
    case V_FLIPADST:
      av1_highbd_inv_txfm2d_add_universe_sse4_1(
          input, dest, stride, tx_type, txfm_param->tx_size,
          txfm_param->use_ddt, txfm_param->eob, bd);
      break;
    default:
      av1_inv_txfm2d_add_8x8_sse4_1(src, dest, stride, tx_type,
                                    txfm_param->use_ddt, bd);
      break;
  }
}
void av1_highbd_inv_txfm_add_4x4_sse4_1(const tran_low_t *input, uint16_t *dest,
                                        int stride,
                                        const TxfmParam *txfm_param) {
  assert(av1_ext_tx_used[txfm_param->tx_set_type][txfm_param->tx_type]);
  int eob = txfm_param->eob;
  int bd = txfm_param->bd;
  int lossless = txfm_param->lossless;
  const int32_t *src = cast_to_int32(input);
  const TX_TYPE tx_type = txfm_param->tx_type;
  if (lossless) {
#if CONFIG_LOSSLESS_DPCM && !CONFIG_IMPROVE_LOSSLESS_TXM
    assert(tx_type == DCT_DCT || tx_type == IDTX);
    if (tx_type == IDTX) {
      av1_inv_txfm2d_add_4x4_sse4_1(src, dest, stride, tx_type,
                                    txfm_param->use_ddt, bd);
    } else {
      av1_highbd_iwht4x4_add(input, dest, stride, eob, bd);
    }
#else
    assert(tx_type == DCT_DCT);
    av1_highbd_iwht4x4_add(input, dest, stride, eob, bd);
#endif  // CONFIG_LOSSLESS_DPCM && !CONFIG_IMPROVE_LOSSLESS_TXM
    return;
  }
  av1_inv_txfm2d_add_4x4_sse4_1(src, dest, stride, tx_type, txfm_param->use_ddt,
                                bd);
}
static void iidentity32_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                               int bd, int out_shift) {
  (void)bit;
  for (int i = 0; i < 32; i += 16) {
    out[i] = _mm_slli_epi32(in[i], 2);
    out[i + 1] = _mm_slli_epi32(in[i + 1], 2);
    out[i + 2] = _mm_slli_epi32(in[i + 2], 2);
    out[i + 3] = _mm_slli_epi32(in[i + 3], 2);
    out[i + 4] = _mm_slli_epi32(in[i + 4], 2);
    out[i + 5] = _mm_slli_epi32(in[i + 5], 2);
    out[i + 6] = _mm_slli_epi32(in[i + 6], 2);
    out[i + 7] = _mm_slli_epi32(in[i + 7], 2);
    out[i + 8] = _mm_slli_epi32(in[i + 8], 2);
    out[i + 9] = _mm_slli_epi32(in[i + 9], 2);
    out[i + 10] = _mm_slli_epi32(in[i + 10], 2);
    out[i + 11] = _mm_slli_epi32(in[i + 11], 2);
    out[i + 12] = _mm_slli_epi32(in[i + 12], 2);
    out[i + 13] = _mm_slli_epi32(in[i + 13], 2);
    out[i + 14] = _mm_slli_epi32(in[i + 14], 2);
    out[i + 15] = _mm_slli_epi32(in[i + 15], 2);
  }

  if (!do_cols) {
    const int log_range_out = AOMMAX(16, bd + 6);
    const __m128i clamp_lo_out = _mm_set1_epi32(-(1 << (log_range_out - 1)));
    const __m128i clamp_hi_out = _mm_set1_epi32((1 << (log_range_out - 1)) - 1);
    round_shift_8x8(out, out_shift);
    round_shift_8x8(out + 16, out_shift);
    highbd_clamp_epi32_sse4_1(out, out, &clamp_lo_out, &clamp_hi_out, 32);
  }
}
static const transform_1d_sse4_1
    highbd_txfm_all_1d_zeros_w8_arr_inter[TX_SIZES][ITX_TYPES_1D][4] = {
      {
          { idct4x4_sse4_1, NULL, NULL, NULL },
#if REPLACE_ADST4
          { iddt4x4_sse4_1, NULL, NULL, NULL },
#else
          { iadst4x4_sse4_1, NULL, NULL, NULL },
#endif
          { iidentity4_sse4_1, iidentity4_sse4_1, iidentity4_sse4_1, NULL },
      },
      {
          { idct8x8_low1_sse4_1, idct8x8_new_sse4_1, NULL, NULL },
#if REPLACE_ADST8
          { iddt8x8_low1_sse4_1, iddt8x8_new_sse4_1, NULL, NULL },
#else
          { iadst8x8_low1_sse4_1, iadst8x8_new_sse4_1, NULL, NULL },
#endif
          { iidentity8_sse4_1, iidentity8_sse4_1, NULL, NULL },
      },
      {
          { idct16x16_low1_sse4_1, idct16x16_low8_sse4_1, idct16x16_sse4_1,
            NULL },
#if REPLACE_ADST16
          { iddt16x16_low1_sse4_1, iddt16x16_low8_sse4_1, iddt16x16_sse4_1,
            NULL },
#else
          { iadst16x16_low1_sse4_1, iadst16x16_low8_sse4_1, iadst16x16_sse4_1,
            NULL },
#endif
          { iidentity16_sse4_1, NULL, iidentity16_sse4_1, NULL },
      },
      {
          { idct32x32_low1_sse4_1, idct32x32_low8_sse4_1,
            idct32x32_low16_sse4_1, idct32x32_sse4_1 },
          { NULL, NULL, NULL, NULL },
          { iidentity32_sse4_1, NULL, NULL, iidentity32_sse4_1 },
      },
      {
          { idct64x64_low1_sse4_1, idct64x64_low8_sse4_1,
            idct64x64_low16_sse4_1, idct64x64_sse4_1 },
          { NULL, NULL, NULL, NULL },
          { NULL, NULL, NULL, NULL },
      }
    };
static const transform_1d_sse4_1
    highbd_txfm_all_1d_zeros_w8_arr[TX_SIZES][ITX_TYPES_1D][4] = {
      {
          { idct4x4_sse4_1, NULL, NULL, NULL },
          { iadst4x4_sse4_1, NULL, NULL, NULL },
          { iidentity4_sse4_1, iidentity4_sse4_1, iidentity4_sse4_1, NULL },
      },
      {
          { idct8x8_low1_sse4_1, idct8x8_new_sse4_1, NULL, NULL },
          { iadst8x8_low1_sse4_1, iadst8x8_new_sse4_1, NULL, NULL },
          { iidentity8_sse4_1, iidentity8_sse4_1, NULL, NULL },
      },
      {
          { idct16x16_low1_sse4_1, idct16x16_low8_sse4_1, idct16x16_sse4_1,
            NULL },
          { iadst16x16_low1_sse4_1, iadst16x16_low8_sse4_1, iadst16x16_sse4_1,
            NULL },
          { iidentity16_sse4_1, NULL, iidentity16_sse4_1, NULL },
      },
      {
          { idct32x32_low1_sse4_1, idct32x32_low8_sse4_1,
            idct32x32_low16_sse4_1, idct32x32_sse4_1 },
          { NULL, NULL, NULL, NULL },
          { iidentity32_sse4_1, NULL, NULL, iidentity32_sse4_1 },
      },
      {
          { idct64x64_low1_sse4_1, idct64x64_low8_sse4_1,
            idct64x64_low16_sse4_1, idct64x64_sse4_1 },
          { NULL, NULL, NULL, NULL },
          { NULL, NULL, NULL, NULL },
      }
    };
static void highbd_inv_txfm2d_add_h_identity_ssse41(
    const int32_t *input, uint16_t *output, int stride, TX_TYPE tx_type,
    TX_SIZE tx_size, int use_ddt, int eob, const int bd) {
  __m128i buf1[64 * 16];
  int eobx, eoby;
  get_eobx_eoby_scan_v_identity(&eobx, &eoby, tx_size, eob);
  const int8_t *shift = av1_inv_txfm_shift_ls[tx_size];
  const int txw_idx = get_txw_idx(tx_size);
  const int txh_idx = get_txh_idx(tx_size);
  const int txfm_size_col = tx_size_wide[tx_size];
  const int txfm_size_row = tx_size_high[tx_size];
  const int input_stride = AOMMIN(32, txfm_size_col);
  const int buf_size_w_div4 = input_stride >> 2;
  const int buf_size_h_div8 = (eoby + 8) >> 3;
  const int rect_type = get_rect_tx_log_ratio(txfm_size_col, txfm_size_row);
  const int fun_idx = highbd_txfm_all_1d_zeros_idx[eoby];
  const transform_1d_sse4_1 row_txfm =
      use_ddt
          ? highbd_txfm_all_1d_zeros_w8_arr_inter[txw_idx][hitx_1d_tab[tx_type]]
                                                 [0]
          : highbd_txfm_all_1d_zeros_w8_arr[txw_idx][hitx_1d_tab[tx_type]][0];
  const transform_1d_sse4_1 col_txfm =
      use_ddt
          ? highbd_txfm_all_1d_zeros_w8_arr_inter[txh_idx][vitx_1d_tab[tx_type]]
                                                 [fun_idx]
          : highbd_txfm_all_1d_zeros_w8_arr[txh_idx][vitx_1d_tab[tx_type]]
                                           [fun_idx];
  int ud_flip, lr_flip;
  get_flip_cfg(tx_type, &ud_flip, &lr_flip);

  for (int i = 0; i < (buf_size_h_div8 << 1); ++i) {
    __m128i buf0[64];
    const int32_t *input_row = input + i * input_stride * 4;
    for (int j = 0; j < buf_size_w_div4; ++j) {
      __m128i *buf0_cur = buf0 + j * 4;
      load_buffer_32bit_input(input_row + j * 4, input_stride, buf0_cur, 4);
    }
    if (abs(rect_type) % 2 == 1) {
      av1_round_shift_rect_array_32_sse4_1(buf0, buf0, input_stride, 0,
                                           NewInvSqrt2);
    }
    row_txfm(buf0, buf0, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
             -shift[0]);

    __m128i *_buf1 = buf1 + i * 4;

    for (int j = 0; j < buf_size_w_div4; ++j) {
      _buf1[j * txfm_size_row + 0] = buf0[j * 4 + 0];
      _buf1[j * txfm_size_row + 1] = buf0[j * 4 + 1];
      _buf1[j * txfm_size_row + 2] = buf0[j * 4 + 2];
      _buf1[j * txfm_size_row + 3] = buf0[j * 4 + 3];
    }
  }
  for (int i = 0; i < buf_size_w_div4; i++) {
    col_txfm(buf1 + i * txfm_size_row, buf1 + i * txfm_size_row,
             av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);

    av1_round_shift_array_32_sse4_1(buf1 + i * txfm_size_row,
                                    buf1 + i * txfm_size_row, txfm_size_row,
                                    -shift[1]);
  }

  // write to buffer
  for (int i = 0; i < (txfm_size_col >> 3); i++) {
    highbd_write_buffer_8xn_sse4_1(buf1 + i * txfm_size_row * 2, output + 8 * i,
                                   stride, ud_flip, txfm_size_row, bd);
  }
}
static void highbd_inv_txfm2d_add_v_identity_ssse41(
    const int32_t *input, uint16_t *output, int stride, TX_TYPE tx_type,
    TX_SIZE tx_size, int use_ddt, int eob, const int bd) {
  __m128i buf1[64 * 16];
  int eobx, eoby;
  get_eobx_eoby_scan_h_identity(&eobx, &eoby, tx_size, eob);
  const int8_t *shift = av1_inv_txfm_shift_ls[tx_size];
  const int txw_idx = get_txw_idx(tx_size);
  const int txh_idx = get_txh_idx(tx_size);
  const int txfm_size_col = tx_size_wide[tx_size];
  const int txfm_size_row = tx_size_high[tx_size];
  const int input_stride = AOMMIN(32, txfm_size_col);
  const int buf_size_w_div8 = txfm_size_col >> 2;
  const int row_max = AOMMIN(32, txfm_size_row);
  const int buf_size_nonzero_w_div8 = (eobx + 8) >> 3;
  const int rect_type = get_rect_tx_log_ratio(txfm_size_col, txfm_size_row);
  const int fun_idx = highbd_txfm_all_1d_zeros_idx[eobx];
  const transform_1d_sse4_1 row_txfm =
      use_ddt
          ? highbd_txfm_all_1d_zeros_w8_arr_inter[txw_idx][hitx_1d_tab[tx_type]]
                                                 [fun_idx]
          : highbd_txfm_all_1d_zeros_w8_arr[txw_idx][hitx_1d_tab[tx_type]]
                                           [fun_idx];
  const transform_1d_sse4_1 col_txfm =
      use_ddt
          ? highbd_txfm_all_1d_zeros_w8_arr_inter[txh_idx][vitx_1d_tab[tx_type]]
                                                 [0]
          : highbd_txfm_all_1d_zeros_w8_arr[txh_idx][vitx_1d_tab[tx_type]][0];
  int ud_flip, lr_flip;
  get_flip_cfg(tx_type, &ud_flip, &lr_flip);

  for (int i = 0; i < (row_max >> 2); ++i) {
    __m128i buf0[64];
    const int32_t *input_row = input + i * input_stride * 4;
    for (int j = 0; j < (buf_size_nonzero_w_div8 << 1); ++j) {
      __m128i *buf0_cur = buf0 + j * 4;
      load_buffer_32bit_input(input_row + j * 4, input_stride, buf0_cur, 4);

      TRANSPOSE_4X4(buf0_cur[0], buf0_cur[1], buf0_cur[2], buf0_cur[3],
                    buf0_cur[0], buf0_cur[1], buf0_cur[2], buf0_cur[3]);
    }
    if (abs(rect_type) % 2 == 1) {
      av1_round_shift_rect_array_32_sse4_1(
          buf0, buf0, (buf_size_nonzero_w_div8 << 3), 0, NewInvSqrt2);
    }
    row_txfm(buf0, buf0, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
             -shift[0]);

    __m128i *_buf1 = buf1 + i * 4;
    if (lr_flip) {
      for (int j = 0; j < buf_size_w_div8; ++j) {
        TRANSPOSE_4X4(buf0[4 * j + 3], buf0[4 * j + 2], buf0[4 * j + 1],
                      buf0[4 * j],
                      _buf1[txfm_size_row * (buf_size_w_div8 - 1 - j) + 0],
                      _buf1[txfm_size_row * (buf_size_w_div8 - 1 - j) + 1],
                      _buf1[txfm_size_row * (buf_size_w_div8 - 1 - j) + 2],
                      _buf1[txfm_size_row * (buf_size_w_div8 - 1 - j) + 3]);
      }
    } else {
      for (int j = 0; j < buf_size_w_div8; ++j) {
        TRANSPOSE_4X4(
            buf0[j * 4 + 0], buf0[j * 4 + 1], buf0[j * 4 + 2], buf0[j * 4 + 3],
            _buf1[j * txfm_size_row + 0], _buf1[j * txfm_size_row + 1],
            _buf1[j * txfm_size_row + 2], _buf1[j * txfm_size_row + 3]);
      }
    }
  }
  for (int i = 0; i < buf_size_w_div8; i++) {
    col_txfm(buf1 + i * txfm_size_row, buf1 + i * txfm_size_row,
             av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);

    av1_round_shift_array_32_sse4_1(buf1 + i * txfm_size_row,
                                    buf1 + i * txfm_size_row, txfm_size_row,
                                    -shift[1]);
  }

  // write to buffer
  {
    for (int i = 0; i < (txfm_size_col >> 3); i++) {
      highbd_write_buffer_8xn_sse4_1(buf1 + i * txfm_size_row * 2,
                                     output + 8 * i, stride, ud_flip,
                                     txfm_size_row, bd);
    }
  }
}
static void highbd_inv_txfm2d_add_idtx_ssse41(const int32_t *input,
                                              uint16_t *output, int stride,
                                              TX_TYPE tx_type, TX_SIZE tx_size,
                                              int eob, const int bd) {
  (void)eob;
  __m128i buf1[64 * 4];
  const int8_t *shift = av1_inv_txfm_shift_ls[tx_size];
  const int txw_idx = get_txw_idx(tx_size);
  const int txh_idx = get_txh_idx(tx_size);
  const int txfm_size_col = tx_size_wide[tx_size];
  const int txfm_size_row = tx_size_high[tx_size];
  const int input_stride = AOMMIN(32, txfm_size_col);
  const int row_max = AOMMIN(32, txfm_size_row);
  const int rect_type = get_rect_tx_log_ratio(txfm_size_col, txfm_size_row);
  const transform_1d_sse4_1 row_txfm =
      highbd_txfm_all_1d_zeros_w8_arr[txw_idx][hitx_1d_tab[tx_type]][0];
  const transform_1d_sse4_1 col_txfm =
      highbd_txfm_all_1d_zeros_w8_arr[txh_idx][vitx_1d_tab[tx_type]][0];

  for (int i = 0; i < (row_max >> 2); ++i) {
    __m128i buf0[32];
    const int32_t *input_row = input + i * input_stride * 4;
    for (int j = 0; j < (input_stride >> 2); ++j) {
      __m128i *buf0_cur = buf0 + j * 4;
      load_buffer_32bit_input(input_row + j * 4, input_stride, buf0_cur, 4);
    }
    if (abs(rect_type) % 2 == 1) {
      av1_round_shift_rect_array_32_sse4_1(buf0, buf0, input_stride, 0,
                                           NewInvSqrt2);
    }
    row_txfm(buf0, buf0, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
             -shift[0]);

    __m128i *_buf1 = buf1 + i * 4;
    for (int j = 0; j < (input_stride >> 2); ++j) {
      _buf1[j * txfm_size_row + 0] = buf0[j * 4 + 0];
      _buf1[j * txfm_size_row + 1] = buf0[j * 4 + 1];
      _buf1[j * txfm_size_row + 2] = buf0[j * 4 + 2];
      _buf1[j * txfm_size_row + 3] = buf0[j * 4 + 3];
    }
  }
  for (int i = 0; i < (input_stride >> 2); i++) {
    col_txfm(buf1 + i * txfm_size_row, buf1 + i * txfm_size_row,
             av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);

    av1_round_shift_array_32_sse4_1(buf1 + i * txfm_size_row,
                                    buf1 + i * txfm_size_row, txfm_size_row,
                                    -shift[1]);
  }

  // write to buffer
  {
    for (int i = 0; i < (txfm_size_col >> 3); i++) {
      highbd_write_buffer_8xn_sse4_1(buf1 + i * txfm_size_row * 2,
                                     output + 8 * i, stride, 0, txfm_size_row,
                                     bd);
    }
  }
}
static void highbd_inv_txfm2d_add_no_identity_sse41(
    const int32_t *input, uint16_t *output, int stride, TX_TYPE tx_type,
    TX_SIZE tx_size, int use_ddt, int eob, const int bd) {
  __m128i buf1[64 * 16];
  int eobx, eoby;
  get_eobx_eoby_scan_default(&eobx, &eoby, tx_size, eob);
  const int8_t *shift = av1_inv_txfm_shift_ls[tx_size];
  const int txw_idx = get_txw_idx(tx_size);
  const int txh_idx = get_txh_idx(tx_size);
  const int txfm_size_col = tx_size_wide[tx_size];
  const int txfm_size_row = tx_size_high[tx_size];
  const int buf_size_w_div8 = txfm_size_col >> 2;
  const int buf_size_nonzero_w_div8 = (eobx + 8) >> 3;
  const int buf_size_nonzero_h_div8 = (eoby + 8) >> 3;
  const int input_stride = AOMMIN(32, txfm_size_col);
  const int rect_type = get_rect_tx_log_ratio(txfm_size_col, txfm_size_row);

  const int fun_idx_x = highbd_txfm_all_1d_zeros_idx[eobx];
  const int fun_idx_y = highbd_txfm_all_1d_zeros_idx[eoby];
  const transform_1d_sse4_1 row_txfm =
      use_ddt
          ? highbd_txfm_all_1d_zeros_w8_arr_inter[txw_idx][hitx_1d_tab[tx_type]]
                                                 [fun_idx_x]
          : highbd_txfm_all_1d_zeros_w8_arr[txw_idx][hitx_1d_tab[tx_type]]
                                           [fun_idx_x];
  const transform_1d_sse4_1 col_txfm =
      use_ddt
          ? highbd_txfm_all_1d_zeros_w8_arr_inter[txh_idx][vitx_1d_tab[tx_type]]
                                                 [fun_idx_y]
          : highbd_txfm_all_1d_zeros_w8_arr[txh_idx][vitx_1d_tab[tx_type]]
                                           [fun_idx_y];

  assert(col_txfm != NULL);
  assert(row_txfm != NULL);
  int ud_flip, lr_flip;
  get_flip_cfg(tx_type, &ud_flip, &lr_flip);

  // 1st stage: column transform
  for (int i = 0; i < buf_size_nonzero_h_div8 << 1; i++) {
    __m128i buf0[64];
    const int32_t *input_row = input + i * input_stride * 4;
    for (int j = 0; j < buf_size_nonzero_w_div8 << 1; ++j) {
      __m128i *buf0_cur = buf0 + j * 4;
      load_buffer_32bit_input(input_row + j * 4, input_stride, buf0_cur, 4);

      TRANSPOSE_4X4(buf0_cur[0], buf0_cur[1], buf0_cur[2], buf0_cur[3],
                    buf0_cur[0], buf0_cur[1], buf0_cur[2], buf0_cur[3]);
    }
    // rect special code where size difference is a factor of 2
    if (abs(rect_type) % 2 == 1) {
      av1_round_shift_rect_array_32_sse4_1(
          buf0, buf0, buf_size_nonzero_w_div8 << 3, 0, NewInvSqrt2);
    }
    row_txfm(buf0, buf0, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
             -shift[0]);

    __m128i *_buf1 = buf1 + i * 4;
    if (lr_flip) {
      for (int j = 0; j < buf_size_w_div8; ++j) {
        TRANSPOSE_4X4(buf0[4 * j + 3], buf0[4 * j + 2], buf0[4 * j + 1],
                      buf0[4 * j],
                      _buf1[txfm_size_row * (buf_size_w_div8 - 1 - j) + 0],
                      _buf1[txfm_size_row * (buf_size_w_div8 - 1 - j) + 1],
                      _buf1[txfm_size_row * (buf_size_w_div8 - 1 - j) + 2],
                      _buf1[txfm_size_row * (buf_size_w_div8 - 1 - j) + 3]);
      }
    } else {
      for (int j = 0; j < buf_size_w_div8; ++j) {
        TRANSPOSE_4X4(
            buf0[j * 4 + 0], buf0[j * 4 + 1], buf0[j * 4 + 2], buf0[j * 4 + 3],
            _buf1[j * txfm_size_row + 0], _buf1[j * txfm_size_row + 1],
            _buf1[j * txfm_size_row + 2], _buf1[j * txfm_size_row + 3]);
      }
    }
  }
  // 2nd stage: column transform
  for (int i = 0; i < buf_size_w_div8; i++) {
    col_txfm(buf1 + i * txfm_size_row, buf1 + i * txfm_size_row,
             av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);

    av1_round_shift_array_32_sse4_1(buf1 + i * txfm_size_row,
                                    buf1 + i * txfm_size_row, txfm_size_row,
                                    -shift[1]);
  }

  // write to buffer
  {
    for (int i = 0; i < (txfm_size_col >> 3); i++) {
      highbd_write_buffer_8xn_sse4_1(buf1 + i * txfm_size_row * 2,
                                     output + 8 * i, stride, ud_flip,
                                     txfm_size_row, bd);
    }
  }
}

static void highbd_inv_txfm2d_add_4x8_sse41(const int32_t *input,
                                            uint16_t *output, int stride,
                                            TX_TYPE tx_type, TX_SIZE tx_size,
                                            int use_ddt, int eob,
                                            const int bd) {
  (void)eob;
  __m128i buf1[8];
  const int8_t *shift = av1_inv_txfm_shift_ls[tx_size];
  const int txw_idx = get_txw_idx(tx_size);
  const int txh_idx = get_txh_idx(tx_size);
  const int txfm_size_col = tx_size_wide[tx_size];
  const int txfm_size_row = tx_size_high[tx_size];
  const transform_1d_sse4_1 row_txfm =
      use_ddt
          ? highbd_txfm_all_1d_zeros_w8_arr_inter[txw_idx][hitx_1d_tab[tx_type]]
                                                 [0]
          : highbd_txfm_all_1d_zeros_w8_arr[txw_idx][hitx_1d_tab[tx_type]][0];
  const transform_1d_sse4_1 col_txfm =
      use_ddt
          ? highbd_txfm_all_1d_zeros_w8_arr_inter[txh_idx][vitx_1d_tab[tx_type]]
                                                 [1]
          : highbd_txfm_all_1d_zeros_w8_arr[txh_idx][vitx_1d_tab[tx_type]][1];
  const int input_stride = AOMMIN(32, txfm_size_col);

  assert(col_txfm != NULL);
  assert(row_txfm != NULL);
  int ud_flip, lr_flip;
  get_flip_cfg(tx_type, &ud_flip, &lr_flip);

  // 1st stage: column transform
  __m128i buf0[8];
  const int32_t *input_row = input;
  __m128i *buf0_cur = buf0;
  load_buffer_32bit_input(input_row, input_stride, buf0_cur, txfm_size_row);
  av1_round_shift_rect_array_32_sse4_1(buf0, buf0, txfm_size_row, 0,
                                       NewInvSqrt2);
  row_txfm(buf0, buf0, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd, -shift[0]);
  row_txfm(buf0 + 4, buf0 + 4, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd,
           -shift[0]);

  if (lr_flip) {
    TRANSPOSE_4X4(buf0[3], buf0[2], buf0[1], buf0[0], buf1[0], buf1[1], buf1[2],
                  buf1[3]);

    TRANSPOSE_4X4(buf0[7], buf0[6], buf0[5], buf0[4], buf1[4], buf1[5], buf1[6],
                  buf1[7]);
  } else {
    TRANSPOSE_4X4(buf0[0], buf0[1], buf0[2], buf0[3], buf1[0], buf1[1], buf1[2],
                  buf1[3]);

    TRANSPOSE_4X4(buf0[4], buf0[5], buf0[6], buf0[7], buf1[4], buf1[5], buf1[6],
                  buf1[7]);
  }

  // 2nd stage: column transform
  col_txfm(buf1, buf1, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);

  av1_round_shift_array_32_sse4_1(buf1, buf1, txfm_size_row, -shift[1]);

  // write to buffer
  highbd_write_buffer_4xn_sse4_1(buf1, output, stride, ud_flip, txfm_size_row,
                                 bd);
}

static void highbd_inv_txfm2d_add_8x4_sse41(const int32_t *input,
                                            uint16_t *output, int stride,
                                            TX_TYPE tx_type, TX_SIZE tx_size,
                                            int use_ddt, int eob,
                                            const int bd) {
  (void)eob;
  __m128i buf1[8];
  const int8_t *shift = av1_inv_txfm_shift_ls[tx_size];
  const int txw_idx = get_txw_idx(tx_size);
  const int txh_idx = get_txh_idx(tx_size);
  const int txfm_size_col = tx_size_wide[tx_size];
  const int txfm_size_row = tx_size_high[tx_size];
  const transform_1d_sse4_1 row_txfm =
      use_ddt
          ? highbd_txfm_all_1d_zeros_w8_arr_inter[txw_idx][hitx_1d_tab[tx_type]]
                                                 [1]
          : highbd_txfm_all_1d_zeros_w8_arr[txw_idx][hitx_1d_tab[tx_type]][1];
  const transform_1d_sse4_1 col_txfm =
      use_ddt
          ? highbd_txfm_all_1d_zeros_w8_arr_inter[txh_idx][vitx_1d_tab[tx_type]]
                                                 [0]
          : highbd_txfm_all_1d_zeros_w8_arr[txh_idx][vitx_1d_tab[tx_type]][0];

  assert(col_txfm != NULL);
  assert(row_txfm != NULL);
  int ud_flip, lr_flip;
  get_flip_cfg(tx_type, &ud_flip, &lr_flip);

  // 1st stage: column transform
  __m128i buf0[8];
  const int32_t *input_row = input;
  load_buffer_32bit_input(input_row, 4, buf0, txfm_size_col);

  TRANSPOSE_4X4(buf0[0], buf0[2], buf0[4], buf0[6], buf1[0], buf1[1], buf1[2],
                buf1[3]);
  TRANSPOSE_4X4(buf0[1], buf0[3], buf0[5], buf0[7], buf1[4], buf1[5], buf1[6],
                buf1[7]);

  av1_round_shift_rect_array_32_sse4_1(buf1, buf0, txfm_size_col, 0,
                                       NewInvSqrt2);
  row_txfm(buf0, buf0, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd, -shift[0]);

  __m128i *buf1_ptr;
  if (lr_flip) {
    flip_buf_sse2(buf0, buf1, txfm_size_col);
    buf1_ptr = buf1;
  } else {
    buf1_ptr = buf0;
  }

  // 2nd stage: column transform
  for (int i = 0; i < 2; i++) {
    col_txfm(buf1_ptr + i * txfm_size_row, buf1_ptr + i * txfm_size_row,
             av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);
  }
  av1_round_shift_array_32_sse4_1(buf1_ptr, buf1_ptr, txfm_size_col, -shift[1]);
  // write to buffer
  highbd_write_buffer_8xn_sse4_1(buf1_ptr, output, stride, ud_flip,
                                 txfm_size_row, bd);
}

static void highbd_inv_txfm2d_add_4x16_sse4_1(const int32_t *input,
                                              uint16_t *output, int stride,
                                              TX_TYPE tx_type, TX_SIZE tx_size,
                                              int use_ddt, int eob,
                                              const int bd) {
  (void)eob;
  __m128i buf1[16];
  const int8_t *shift = av1_inv_txfm_shift_ls[tx_size];
  const int txw_idx = get_txw_idx(tx_size);
  const int txh_idx = get_txh_idx(tx_size);
  const int txfm_size_col = tx_size_wide[tx_size];
  const int txfm_size_row = tx_size_high[tx_size];
  const int buf_size_h_div8 = txfm_size_row >> 2;
  const transform_1d_sse4_1 row_txfm =
      use_ddt
          ? highbd_txfm_all_1d_zeros_w8_arr_inter[txw_idx][hitx_1d_tab[tx_type]]
                                                 [0]
          : highbd_txfm_all_1d_zeros_w8_arr[txw_idx][hitx_1d_tab[tx_type]][0];
  const transform_1d_sse4_1 col_txfm =
      use_ddt
          ? highbd_txfm_all_1d_zeros_w8_arr_inter[txh_idx][vitx_1d_tab[tx_type]]
                                                 [2]
          : highbd_txfm_all_1d_zeros_w8_arr[txh_idx][vitx_1d_tab[tx_type]][2];
  const int input_stride = AOMMIN(32, txfm_size_col);

  assert(col_txfm != NULL);
  assert(row_txfm != NULL);
  int ud_flip, lr_flip;
  get_flip_cfg(tx_type, &ud_flip, &lr_flip);

  // 1st stage: column transform
  __m128i buf0[16];
  const int32_t *input_row = input;
  __m128i *buf0_cur = buf0;
  load_buffer_32bit_input(input_row, input_stride, buf0_cur, txfm_size_row);
  for (int i = 0; i < (txfm_size_row >> 2); i++) {
    row_txfm(buf0 + (i << 2), buf0 + (i << 2),
             av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd, -shift[0]);
  }

  if (lr_flip) {
    for (int j = 0; j < buf_size_h_div8; ++j) {
      TRANSPOSE_4X4(buf0[4 * j + 3], buf0[4 * j + 2], buf0[4 * j + 1],
                    buf0[4 * j], buf1[4 * j], buf1[4 * j + 1], buf1[4 * j + 2],
                    buf1[4 * j + 3]);
    }
  } else {
    for (int j = 0; j < buf_size_h_div8; ++j) {
      TRANSPOSE_4X4(buf0[4 * j], buf0[4 * j + 1], buf0[4 * j + 2],
                    buf0[4 * j + 3], buf1[4 * j], buf1[4 * j + 1],
                    buf1[4 * j + 2], buf1[4 * j + 3]);
    }
  }

  // 2nd stage: column transform
  col_txfm(buf1, buf1, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);

  av1_round_shift_array_32_sse4_1(buf1, buf1, txfm_size_row, -shift[1]);

  // write to buffer
  highbd_write_buffer_4xn_sse4_1(buf1, output, stride, ud_flip, txfm_size_row,
                                 bd);
}

static void highbd_inv_txfm2d_add_16x4_sse4_1(const int32_t *input,
                                              uint16_t *output, int stride,
                                              TX_TYPE tx_type, TX_SIZE tx_size,
                                              int use_ddt, int eob,
                                              const int bd) {
  (void)eob;
  __m128i buf1[16];
  const int8_t *shift = av1_inv_txfm_shift_ls[tx_size];
  const int txw_idx = get_txw_idx(tx_size);
  const int txh_idx = get_txh_idx(tx_size);
  const int txfm_size_col = tx_size_wide[tx_size];
  const int txfm_size_row = tx_size_high[tx_size];
  const int buf_size_w_div8 = txfm_size_col >> 2;
  const transform_1d_sse4_1 row_txfm =
      use_ddt
          ? highbd_txfm_all_1d_zeros_w8_arr_inter[txw_idx][hitx_1d_tab[tx_type]]
                                                 [2]
          : highbd_txfm_all_1d_zeros_w8_arr[txw_idx][hitx_1d_tab[tx_type]][2];
  const transform_1d_sse4_1 col_txfm =
      use_ddt
          ? highbd_txfm_all_1d_zeros_w8_arr_inter[txh_idx][vitx_1d_tab[tx_type]]
                                                 [0]
          : highbd_txfm_all_1d_zeros_w8_arr[txh_idx][vitx_1d_tab[tx_type]][0];

  assert(col_txfm != NULL);
  assert(row_txfm != NULL);
  int ud_flip, lr_flip;
  get_flip_cfg(tx_type, &ud_flip, &lr_flip);

  // 1st stage: column transform
  __m128i buf0[16];
  const int32_t *input_row = input;
  load_buffer_32bit_input(input_row, 4, buf0, txfm_size_col);

  for (int j = 0; j < buf_size_w_div8; j++) {
    TRANSPOSE_4X4(buf0[j], buf0[j + 4], buf0[j + 8], buf0[j + 12], buf1[4 * j],
                  buf1[4 * j + 1], buf1[4 * j + 2], buf1[4 * j + 3]);
  }
  row_txfm(buf1, buf0, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd, -shift[0]);

  __m128i *buf1_ptr;
  if (lr_flip) {
    flip_buf_sse2(buf0, buf1, txfm_size_col);
    buf1_ptr = buf1;
  } else {
    buf1_ptr = buf0;
  }

  // 2nd stage: column transform
  for (int i = 0; i < buf_size_w_div8; i++) {
    col_txfm(buf1_ptr + i * txfm_size_row, buf1_ptr + i * txfm_size_row,
             av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);
  }
  av1_round_shift_array_32_sse4_1(buf1_ptr, buf1_ptr, txfm_size_col, -shift[1]);

  // write to buffer
  for (int i = 0; i < (txfm_size_col >> 3); i++) {
    highbd_write_buffer_8xn_sse4_1(buf1_ptr + i * txfm_size_row * 2,
                                   output + 8 * i, stride, ud_flip,
                                   txfm_size_row, bd);
  }
}

static void highbd_inv_txfm2d_add_4x32_sse4_1(const int32_t *input,
                                              uint16_t *output, int stride,
                                              TX_TYPE tx_type, TX_SIZE tx_size,
                                              int use_ddt, int eob,
                                              const int bd) {
  (void)eob;
  __m128i buf1[32];
  const int8_t *shift = av1_inv_txfm_shift_ls[tx_size];
  const int txw_idx = get_txw_idx(tx_size);
  const int txh_idx = get_txh_idx(tx_size);
  const int txfm_size_col = tx_size_wide[tx_size];
  const int txfm_size_row = tx_size_high[tx_size];
  const int buf_size_h_div8 = txfm_size_row >> 2;
  const transform_1d_sse4_1 row_txfm =
      (use_ddt && REPLACE_ADST4)
          ? highbd_txfm_all_1d_zeros_w8_arr_inter[txw_idx][hitx_1d_tab[tx_type]]
                                                 [0]
          : highbd_txfm_all_1d_zeros_w8_arr[txw_idx][hitx_1d_tab[tx_type]][0];
  const transform_1d_sse4_1 col_txfm =
      highbd_txfm_all_1d_zeros_w8_arr[txh_idx][vitx_1d_tab[tx_type]][3];
  const int input_stride = AOMMIN(32, txfm_size_col);

  assert(col_txfm != NULL);
  assert(row_txfm != NULL);
  int ud_flip, lr_flip;
  get_flip_cfg(tx_type, &ud_flip, &lr_flip);

  // 1st stage: column transform
  __m128i buf0[32];
  const int32_t *input_row = input;
  __m128i *buf0_cur = buf0;
  load_buffer_32bit_input(input_row, input_stride, buf0_cur, txfm_size_row);
  av1_round_shift_rect_array_32_sse4_1(buf0, buf0, txfm_size_row, 0,
                                       NewInvSqrt2);
  for (int i = 0; i < (txfm_size_row >> 2); i++) {
    row_txfm(buf0 + (i << 2), buf0 + (i << 2),
             av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd, -shift[0]);
  }

  if (lr_flip) {
    for (int j = 0; j < buf_size_h_div8; ++j) {
      TRANSPOSE_4X4(buf0[4 * j + 3], buf0[4 * j + 2], buf0[4 * j + 1],
                    buf0[4 * j], buf1[4 * j], buf1[4 * j + 1], buf1[4 * j + 2],
                    buf1[4 * j + 3]);
    }
  } else {
    for (int j = 0; j < buf_size_h_div8; ++j) {
      TRANSPOSE_4X4(buf0[4 * j], buf0[4 * j + 1], buf0[4 * j + 2],
                    buf0[4 * j + 3], buf1[4 * j], buf1[4 * j + 1],
                    buf1[4 * j + 2], buf1[4 * j + 3]);
    }
  }

  // 2nd stage: column transform
  col_txfm(buf1, buf1, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);

  av1_round_shift_array_32_sse4_1(buf1, buf1, txfm_size_row, -shift[1]);

  // write to buffer
  highbd_write_buffer_4xn_sse4_1(buf1, output, stride, ud_flip, txfm_size_row,
                                 bd);
}

static void highbd_inv_txfm2d_add_32x4_sse4_1(const int32_t *input,
                                              uint16_t *output, int stride,
                                              TX_TYPE tx_type, TX_SIZE tx_size,
                                              int use_ddt, int eob,
                                              const int bd) {
  (void)eob;
  __m128i buf1[32];
  const int8_t *shift = av1_inv_txfm_shift_ls[tx_size];
  const int txw_idx = get_txw_idx(tx_size);
  const int txh_idx = get_txh_idx(tx_size);
  const int txfm_size_col = tx_size_wide[tx_size];
  const int txfm_size_row = tx_size_high[tx_size];
  const int buf_size_w_div8 = txfm_size_col >> 2;
  const transform_1d_sse4_1 row_txfm =
      highbd_txfm_all_1d_zeros_w8_arr[txw_idx][hitx_1d_tab[tx_type]][3];
  const transform_1d_sse4_1 col_txfm =
      (use_ddt && REPLACE_ADST4)
          ? highbd_txfm_all_1d_zeros_w8_arr_inter[txh_idx][vitx_1d_tab[tx_type]]
                                                 [0]
          : highbd_txfm_all_1d_zeros_w8_arr[txh_idx][vitx_1d_tab[tx_type]][0];

  assert(col_txfm != NULL);
  assert(row_txfm != NULL);
  int ud_flip, lr_flip;
  get_flip_cfg(tx_type, &ud_flip, &lr_flip);

  // 1st stage: column transform
  __m128i buf0[32];
  const int32_t *input_row = input;
  load_buffer_32bit_input(input_row, 4, buf0, txfm_size_col);
  for (int j = 0; j < buf_size_w_div8; j++) {
    TRANSPOSE_4X4(buf0[j], buf0[j + 8], buf0[j + 16], buf0[j + 24], buf1[4 * j],
                  buf1[4 * j + 1], buf1[4 * j + 2], buf1[4 * j + 3]);
  }

  av1_round_shift_rect_array_32_sse4_1(buf1, buf1, txfm_size_col, 0,
                                       NewInvSqrt2);
  row_txfm(buf1, buf0, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd, -shift[0]);

  __m128i *buf1_ptr;
  if (lr_flip) {
    flip_buf_sse2(buf0, buf1, txfm_size_col);
    buf1_ptr = buf1;
  } else {
    buf1_ptr = buf0;
  }

  // 2nd stage: column transform
  for (int i = 0; i < buf_size_w_div8; i++) {
    col_txfm(buf1_ptr + i * txfm_size_row, buf1_ptr + i * txfm_size_row,
             av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);
  }
  av1_round_shift_array_32_sse4_1(buf1_ptr, buf1_ptr, txfm_size_col, -shift[1]);

  // write to buffer
  for (int i = 0; i < (txfm_size_col >> 3); i++) {
    highbd_write_buffer_8xn_sse4_1(buf1_ptr + i * txfm_size_row * 2,
                                   output + 8 * i, stride, ud_flip,
                                   txfm_size_row, bd);
  }
}

static void highbd_inv_txfm2d_add_4x64_sse4_1(const int32_t *input,
                                              uint16_t *output, int stride,
                                              TX_TYPE tx_type, TX_SIZE tx_size,
                                              int use_ddt, int eob,
                                              const int bd) {
  (void)eob;
  __m128i buf1[64];
  const int8_t *shift = av1_inv_txfm_shift_ls[tx_size];
  const int txw_idx = get_txw_idx(tx_size);
  const int txh_idx = get_txh_idx(tx_size);
  const int txfm_size_col = tx_size_wide[tx_size];
  const int txfm_size_row = tx_size_high[tx_size];
  const int buf_size_h_div8 = txfm_size_row >> 2;
  const transform_1d_sse4_1 row_txfm =
      (use_ddt && REPLACE_ADST4)
          ? highbd_txfm_all_1d_zeros_w8_arr_inter[txw_idx][hitx_1d_tab[tx_type]]
                                                 [0]
          : highbd_txfm_all_1d_zeros_w8_arr[txw_idx][hitx_1d_tab[tx_type]][0];
  const transform_1d_sse4_1 col_txfm =
      highbd_txfm_all_1d_zeros_w8_arr[txh_idx][vitx_1d_tab[tx_type]][3];
  const int input_stride = AOMMIN(32, txfm_size_col);

  assert(col_txfm != NULL);
  assert(row_txfm != NULL);
  int ud_flip, lr_flip;
  get_flip_cfg(tx_type, &ud_flip, &lr_flip);

  // 1st stage: column transform
  __m128i buf0[64];
  const int32_t *input_row = input;
  __m128i *buf0_cur = buf0;
  load_buffer_32bit_input(input_row, input_stride, buf0_cur, txfm_size_row);

  for (int i = 0; i < (txfm_size_row >> 2); i++) {
    row_txfm(buf0 + (i << 2), buf0 + (i << 2),
             av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd, -shift[0]);
  }

  if (lr_flip) {
    for (int j = 0; j < buf_size_h_div8; ++j) {
      TRANSPOSE_4X4(buf0[4 * j + 3], buf0[4 * j + 2], buf0[4 * j + 1],
                    buf0[4 * j], buf1[4 * j], buf1[4 * j + 1], buf1[4 * j + 2],
                    buf1[4 * j + 3]);
    }
  } else {
    for (int j = 0; j < buf_size_h_div8; ++j) {
      TRANSPOSE_4X4(buf0[4 * j], buf0[4 * j + 1], buf0[4 * j + 2],
                    buf0[4 * j + 3], buf1[4 * j], buf1[4 * j + 1],
                    buf1[4 * j + 2], buf1[4 * j + 3]);
    }
  }

  // 2nd stage: column transform
  col_txfm(buf1, buf1, av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);

  av1_round_shift_array_32_sse4_1(buf1, buf1, txfm_size_row, -shift[1]);

  // write to buffer
  highbd_write_buffer_4xn_sse4_1(buf1, output, stride, ud_flip, txfm_size_row,
                                 bd);
}

static void highbd_inv_txfm2d_add_64x4_sse4_1(const int32_t *input,
                                              uint16_t *output, int stride,
                                              TX_TYPE tx_type, TX_SIZE tx_size,
                                              int use_ddt, int eob,
                                              const int bd) {
  (void)eob;
  __m128i buf1[64];
  const int8_t *shift = av1_inv_txfm_shift_ls[tx_size];
  const int txw_idx = get_txw_idx(tx_size);
  const int txh_idx = get_txh_idx(tx_size);
  const int txfm_size_col = tx_size_wide[tx_size];
  const int txfm_size_row = tx_size_high[tx_size];
  const int buf_size_w_div8 = txfm_size_col >> 2;
  const int input_stride = AOMMIN(32, txfm_size_col);
  const transform_1d_sse4_1 row_txfm =
      highbd_txfm_all_1d_zeros_w8_arr[txw_idx][hitx_1d_tab[tx_type]][3];
  const transform_1d_sse4_1 col_txfm =
      (use_ddt && REPLACE_ADST4)
          ? highbd_txfm_all_1d_zeros_w8_arr_inter[txh_idx][vitx_1d_tab[tx_type]]
                                                 [0]
          : highbd_txfm_all_1d_zeros_w8_arr[txh_idx][vitx_1d_tab[tx_type]][0];

  assert(col_txfm != NULL);
  assert(row_txfm != NULL);
  int ud_flip, lr_flip;
  get_flip_cfg(tx_type, &ud_flip, &lr_flip);

  // 1st stage: column transform
  __m128i buf0[64];
  const int32_t *input_row = input;
  load_buffer_32bit_input(input_row, 4, buf0, input_stride);

  for (int j = 0; j < buf_size_w_div8 >> 1; j++) {
    TRANSPOSE_4X4(buf0[j], buf0[j + 8], buf0[j + 16], buf0[j + 24], buf1[4 * j],
                  buf1[4 * j + 1], buf1[4 * j + 2], buf1[4 * j + 3]);
  }

  row_txfm(buf1, buf0, av1_inv_cos_bit_row[txw_idx][txh_idx], 0, bd, -shift[0]);

  __m128i *buf1_ptr;
  if (lr_flip) {
    flip_buf_sse2(buf0, buf1, txfm_size_col);
    buf1_ptr = buf1;
  } else {
    buf1_ptr = buf0;
  }

  // 2nd stage: column transform
  for (int i = 0; i < buf_size_w_div8; i++) {
    col_txfm(buf1_ptr + i * txfm_size_row, buf1_ptr + i * txfm_size_row,
             av1_inv_cos_bit_col[txw_idx][txh_idx], 1, bd, 0);
  }
  av1_round_shift_array_32_sse4_1(buf1_ptr, buf1_ptr, txfm_size_col, -shift[1]);

  // write to buffer
  for (int i = 0; i < (txfm_size_col >> 3); i++) {
    highbd_write_buffer_8xn_sse4_1(buf1_ptr + i * txfm_size_row * 2,
                                   output + 8 * i, stride, ud_flip,
                                   txfm_size_row, bd);
  }
}

void av1_highbd_inv_txfm2d_add_universe_sse4_1(const int32_t *input,
                                               uint16_t *output, int stride,
                                               TX_TYPE tx_type, TX_SIZE tx_size,
                                               int use_ddt, int eob,
                                               const int bd) {
  switch (tx_type) {
    case DCT_DCT:
    case ADST_DCT:
    case DCT_ADST:
    case ADST_ADST:
    case FLIPADST_DCT:
    case DCT_FLIPADST:
    case FLIPADST_FLIPADST:
    case ADST_FLIPADST:
    case FLIPADST_ADST:
      highbd_inv_txfm2d_add_no_identity_sse41(input, output, stride, tx_type,
                                              tx_size, use_ddt, eob, bd);
      break;
    case V_DCT:
    case V_ADST:
    case V_FLIPADST:
      highbd_inv_txfm2d_add_h_identity_ssse41(input, output, stride, tx_type,
                                              tx_size, use_ddt, eob, bd);
      break;
    case H_DCT:
    case H_ADST:
    case H_FLIPADST:
      highbd_inv_txfm2d_add_v_identity_ssse41(input, output, stride, tx_type,
                                              tx_size, use_ddt, eob, bd);
      break;
    case IDTX:
      highbd_inv_txfm2d_add_idtx_ssse41(input, output, stride, tx_type, tx_size,
                                        eob, bd);
      break;
    default: assert(0); break;
  }
}

void av1_highbd_inv_txfm_add_4x8_sse4_1(const tran_low_t *input, uint16_t *dest,
                                        int stride,
                                        const TxfmParam *txfm_param) {
  int bd = txfm_param->bd;
  const TX_TYPE tx_type = txfm_param->tx_type;
  const TX_SIZE tx_size = txfm_param->tx_size;
  int eob = txfm_param->eob;
  highbd_inv_txfm2d_add_4x8_sse41(input, dest, stride, tx_type, tx_size,
                                  txfm_param->use_ddt, eob, bd);
}

void av1_highbd_inv_txfm_add_8x4_sse4_1(const tran_low_t *input, uint16_t *dest,
                                        int stride,
                                        const TxfmParam *txfm_param) {
  int bd = txfm_param->bd;
  const TX_TYPE tx_type = txfm_param->tx_type;
  const TX_SIZE tx_size = txfm_param->tx_size;
  int eob = txfm_param->eob;
  highbd_inv_txfm2d_add_8x4_sse41(input, dest, stride, tx_type, tx_size,
                                  txfm_param->use_ddt, eob, bd);
}

void av1_highbd_inv_txfm_add_4x16_sse4_1(const tran_low_t *input,
                                         uint16_t *dest, int stride,
                                         const TxfmParam *txfm_param) {
  int bd = txfm_param->bd;
  const TX_TYPE tx_type = txfm_param->tx_type;
  const TX_SIZE tx_size = txfm_param->tx_size;
  int eob = txfm_param->eob;
  highbd_inv_txfm2d_add_4x16_sse4_1(input, dest, stride, tx_type, tx_size,
                                    txfm_param->use_ddt, eob, bd);
}

void av1_highbd_inv_txfm_add_16x4_sse4_1(const tran_low_t *input,
                                         uint16_t *dest, int stride,
                                         const TxfmParam *txfm_param) {
  int bd = txfm_param->bd;
  const TX_TYPE tx_type = txfm_param->tx_type;
  const TX_SIZE tx_size = txfm_param->tx_size;
  int eob = txfm_param->eob;
  highbd_inv_txfm2d_add_16x4_sse4_1(input, dest, stride, tx_type, tx_size,
                                    txfm_param->use_ddt, eob, bd);
}

void av1_highbd_inv_txfm_add_4x32_sse4_1(const tran_low_t *input,
                                         uint16_t *dest, int stride,
                                         const TxfmParam *txfm_param) {
  int bd = txfm_param->bd;
  const TX_TYPE tx_type = txfm_param->tx_type;
  const TX_SIZE tx_size = txfm_param->tx_size;
  int eob = txfm_param->eob;
  highbd_inv_txfm2d_add_4x32_sse4_1(input, dest, stride, tx_type, tx_size,
                                    txfm_param->use_ddt, eob, bd);
}

void av1_highbd_inv_txfm_add_32x4_sse4_1(const tran_low_t *input,
                                         uint16_t *dest, int stride,
                                         const TxfmParam *txfm_param) {
  int bd = txfm_param->bd;
  const TX_TYPE tx_type = txfm_param->tx_type;
  const TX_SIZE tx_size = txfm_param->tx_size;
  int eob = txfm_param->eob;
  highbd_inv_txfm2d_add_32x4_sse4_1(input, dest, stride, tx_type, tx_size,
                                    txfm_param->use_ddt, eob, bd);
}

void av1_highbd_inv_txfm_add_4x64_sse4_1(const tran_low_t *input,
                                         uint16_t *dest, int stride,
                                         const TxfmParam *txfm_param) {
  int bd = txfm_param->bd;
  const TX_TYPE tx_type = txfm_param->tx_type;
  const TX_SIZE tx_size = txfm_param->tx_size;
  int eob = txfm_param->eob;
  highbd_inv_txfm2d_add_4x64_sse4_1(input, dest, stride, tx_type, tx_size,
                                    txfm_param->use_ddt, eob, bd);
}

void av1_highbd_inv_txfm_add_64x4_sse4_1(const tran_low_t *input,
                                         uint16_t *dest, int stride,
                                         const TxfmParam *txfm_param) {
  int bd = txfm_param->bd;
  const TX_TYPE tx_type = txfm_param->tx_type;
  const TX_SIZE tx_size = txfm_param->tx_size;
  int eob = txfm_param->eob;
  highbd_inv_txfm2d_add_64x4_sse4_1(input, dest, stride, tx_type, tx_size,
                                    txfm_param->use_ddt, eob, bd);
}

void av1_highbd_inv_txfm_add_sse4_1(const tran_low_t *input, uint16_t *dest,
                                    int stride, const TxfmParam *txfm_param) {
  assert(av1_ext_tx_used[txfm_param->tx_set_type][txfm_param->tx_type]);
#if CONFIG_CORE_TX
  inv_txfm_c(input, dest, stride, txfm_param);
#else
  const TX_SIZE tx_size = txfm_param->tx_size;
  switch (tx_size) {
    case TX_8X8:
      av1_highbd_inv_txfm_add_8x8_sse4_1(input, dest, stride, txfm_param);
      break;
    case TX_4X8:
      av1_highbd_inv_txfm_add_4x8_sse4_1(input, dest, stride, txfm_param);
      break;
    case TX_8X4:
      av1_highbd_inv_txfm_add_8x4_sse4_1(input, dest, stride, txfm_param);
      break;
    case TX_4X4:
      av1_highbd_inv_txfm_add_4x4_sse4_1(input, dest, stride, txfm_param);
      break;
    case TX_16X4:
      av1_highbd_inv_txfm_add_16x4_sse4_1(input, dest, stride, txfm_param);
      break;
    case TX_4X16:
      av1_highbd_inv_txfm_add_4x16_sse4_1(input, dest, stride, txfm_param);
      break;
    case TX_4X32:
      av1_highbd_inv_txfm_add_4x32_sse4_1(input, dest, stride, txfm_param);
      break;
    case TX_32X4:
      av1_highbd_inv_txfm_add_32x4_sse4_1(input, dest, stride, txfm_param);
      break;
    case TX_4X64:
      av1_highbd_inv_txfm_add_4x64_sse4_1(input, dest, stride, txfm_param);
      break;
    case TX_64X4:
      av1_highbd_inv_txfm_add_64x4_sse4_1(input, dest, stride, txfm_param);
      break;
    default:
      av1_highbd_inv_txfm2d_add_universe_sse4_1(
          input, dest, stride, txfm_param->tx_type, tx_size,
          txfm_param->use_ddt, txfm_param->eob, txfm_param->bd);
      break;
  }
#endif  // CONFIG_CORE_TX
}

// 128bit intrinsic implementation of ROUND_POWER_OF_TWO_SIGNED.
static INLINE __m128i round_power_of_two_signed_epi32(__m128i val, int bits) {
  const __m128i v_bias_d = _mm_set1_epi32((1 << bits) >> 1);
  const __m128i v_sign_d = _mm_srai_epi32(val, 31);
  const __m128i v_tmp_d = _mm_add_epi32(_mm_add_epi32(val, v_bias_d), v_sign_d);
  return _mm_srai_epi32(v_tmp_d, bits);
}

// Inverse secondary transform
void inv_stxfm_sse4_1(tran_low_t *src, tran_low_t *dst,
                      const PREDICTION_MODE mode, const uint8_t stx_idx,
                      const int size, const int bd) {
  assert(stx_idx < 4);
#if CONFIG_E124_IST_REDUCE_METHOD4
  const int16_t *kernel = (size == 0) ? ist_4x4_kernel[mode][stx_idx][0]
                                      : ist_8x8_kernel[mode][stx_idx][0];
#else
  const int16_t *kernel = (size == 4) ? ist_4x4_kernel[mode][stx_idx][0]
                                      : ist_8x8_kernel[mode][stx_idx][0];
#endif  // CONFIG_E124_IST_REDUCE_METHOD4

  int reduced_width, reduced_height;
#if CONFIG_E124_IST_REDUCE_METHOD4
  if (size == 0) {
    reduced_height = IST_4x4_HEIGHT;
    reduced_width = IST_4x4_WIDTH;
  } else {
#if CONFIG_F105_IST_MEM_REDUCE
    reduced_height = (size == 1)
                         ? IST_8x8_HEIGHT_RED
                         : ((size == 3) ? IST_ADST_NZ_CNT : IST_8x8_HEIGHT);
#else
    reduced_height = (size == 1) ? IST_8x8_HEIGHT_RED : IST_8x8_HEIGHT;
#endif  // CONFIG_F105_IST_MEM_REDUCE
    reduced_width = IST_8x8_WIDTH;
  }
#else
  if (size == 4) {
    reduced_height = IST_4x4_HEIGHT;
    reduced_width = IST_4x4_WIDTH;
  } else {
    reduced_height = IST_8x8_HEIGHT;
    reduced_width = IST_8x8_WIDTH;
  }
#endif  // CONFIG_E124_IST_REDUCE_METHOD4
  for (int j = 0; j < reduced_height; j++) {
    const int16_t *kernel_tmp = kernel;
    int *srcPtr = src;
    int *out = dst;
    __m128i tmpCoeff = _mm_set1_epi32(srcPtr[j]);
    __m128i *tmpBlock = (__m128i *)out;
    for (int i = 0; i < reduced_width; i += 4, tmpBlock++) {
      __m128i tmp =
          _mm_cvtepi16_epi32(_mm_loadl_epi64((__m128i *)(kernel_tmp + i)));
      __m128i sum = _mm_loadu_si128(tmpBlock);
      tmp = _mm_mullo_epi32(tmpCoeff, tmp);
      tmp = _mm_add_epi32(tmp, sum);
      _mm_storeu_si128(tmpBlock, tmp);
    }
    kernel += reduced_width;
  }
  int *out = dst;
  __m128i *tmpBlock = (__m128i *)out;
  const __m128i max_value = _mm_set1_epi32((1 << (7 + bd)) - 1);
  const __m128i min_value = _mm_set1_epi32(-(1 << (7 + bd)));
  for (int j = 0; j < reduced_width; j += 4, tmpBlock++) {
    __m128i tmp = _mm_loadu_si128(tmpBlock);
    tmp = round_power_of_two_signed_epi32(tmp, 7);
    tmp = _mm_min_epi32(_mm_max_epi32(tmp, min_value), max_value);
    _mm_storeu_si128(tmpBlock, tmp);
  }
}