Line data    Source code

       1             : /*
       2             :  * Copyright (c) 2016, Alliance for Open Media. All rights reserved
       3             :  *
       4             :  * This source code is subject to the terms of the BSD 2 Clause License and
       5             :  * the Alliance for Open Media Patent License 1.0. If the BSD 2 Clause License
       6             :  * was not distributed with this source code in the LICENSE file, you can
       7             :  * obtain it at www.aomedia.org/license/software. If the Alliance for Open
       8             :  * Media Patent License 1.0 was not distributed with this source code in the
       9             :  * PATENTS file, you can obtain it at www.aomedia.org/license/patent.
      10             :  */
      11             : 
      12             : #ifndef AOM_DSP_X86_FWD_TXFM_AVX2_H
      13             : #define AOM_DSP_X86_FWD_TXFM_AVX2_H
      14             : 
      15             : #include "./aom_config.h"
      16             : 
      17           0 : static INLINE void storeu_output_avx2(const __m256i *coeff, tran_low_t *out) {
      18             : #if CONFIG_HIGHBITDEPTH
      19           0 :   const __m256i zero = _mm256_setzero_si256();
      20           0 :   const __m256i sign = _mm256_cmpgt_epi16(zero, *coeff);
      21             : 
      22           0 :   __m256i x0 = _mm256_unpacklo_epi16(*coeff, sign);
      23           0 :   __m256i x1 = _mm256_unpackhi_epi16(*coeff, sign);
      24             : 
      25           0 :   __m256i y0 = _mm256_permute2x128_si256(x0, x1, 0x20);
      26           0 :   __m256i y1 = _mm256_permute2x128_si256(x0, x1, 0x31);
      27             : 
      28             :   _mm256_storeu_si256((__m256i *)out, y0);
      29           0 :   _mm256_storeu_si256((__m256i *)(out + 8), y1);
      30             : #else
      31             :   _mm256_storeu_si256((__m256i *)out, *coeff);
      32             : #endif
      33           0 : }
      34             : 
      35             : #endif  // AOM_DSP_X86_FWD_TXFM_AVX2_H