chromium/third_party/libvpx/source/libvpx/vpx_dsp/x86/convolve_avx2.h

/*
 *  Copyright (c) 2017 The WebM project authors. All Rights Reserved.
 *
 *  Use of this source code is governed by a BSD-style license
 *  that can be found in the LICENSE file in the root of the source
 *  tree. An additional intellectual property rights grant can be found
 *  in the file PATENTS.  All contributing project authors may
 *  be found in the AUTHORS file in the root of the source tree.
 */

#ifndef VPX_VPX_DSP_X86_CONVOLVE_AVX2_H_
#define VPX_VPX_DSP_X86_CONVOLVE_AVX2_H_

#include <immintrin.h>  // AVX2

#include "./vpx_config.h"

#if defined(__clang__)
#if (__clang_major__ > 0 && __clang_major__ < 3) ||            \
    (__clang_major__ == 3 && __clang_minor__ <= 3) ||          \
    (defined(__APPLE__) && defined(__apple_build_version__) && \
     ((__clang_major__ == 4 && __clang_minor__ <= 2) ||        \
      (__clang_major__ == 5 && __clang_minor__ == 0)))
#define MM256_BROADCASTSI128_SI256 …
#else  // clang > 3.3, and not 5.0 on macosx.
#define MM256_BROADCASTSI128_SI256 …
#endif  // clang <= 3.3
#elif defined(__GNUC__)
#if __GNUC__ < 4 || (__GNUC__ == 4 && __GNUC_MINOR__ <= 6)
#define MM256_BROADCASTSI128_SI256 …
#elif __GNUC__ == 4 && __GNUC_MINOR__ == 7
#define MM256_BROADCASTSI128_SI256 …
#else  // gcc > 4.7
#define MM256_BROADCASTSI128_SI256 …
#endif  // gcc <= 4.6
#else   // !(gcc || clang)
#define MM256_BROADCASTSI128_SI256 …
#endif  // __clang__

static INLINE void shuffle_filter_avx2(const int16_t *const filter,
                                       __m256i *const f) { … }

static INLINE __m256i convolve8_16_avx2(const __m256i *const s,
                                        const __m256i *const f) { … }

static INLINE __m128i convolve8_8_avx2(const __m256i *const s,
                                       const __m256i *const f) { … }

static INLINE __m256i mm256_loadu2_si128(const void *lo, const void *hi) { … }

static INLINE __m256i mm256_loadu2_epi64(const void *lo, const void *hi) { … }

static INLINE void mm256_store2_si128(__m128i *const dst_ptr_1,
                                      __m128i *const dst_ptr_2,
                                      const __m256i *const src) { … }

static INLINE void mm256_storeu2_epi64(__m128i *const dst_ptr_1,
                                       __m128i *const dst_ptr_2,
                                       const __m256i *const src) { … }

static INLINE void mm256_storeu2_epi32(__m128i *const dst_ptr_1,
                                       __m128i *const dst_ptr_2,
                                       const __m256i *const src) { … }

static INLINE __m256i mm256_round_epi32(const __m256i *const src,
                                        const __m256i *const half_depth,
                                        const int depth) { … }

static INLINE __m256i mm256_round_epi16(const __m256i *const src,
                                        const __m256i *const half_depth,
                                        const int depth) { … }

static INLINE __m256i mm256_madd_add_epi32(const __m256i *const src_0,
                                           const __m256i *const src_1,
                                           const __m256i *const ker_0,
                                           const __m256i *const ker_1) { … }

#undef MM256_BROADCASTSI128_SI256

#endif  // VPX_VPX_DSP_X86_CONVOLVE_AVX2_H_