xref: /aosp_15_r20/external/XNNPACK/src/x16-transposec/gen/8x8-multi-switch-sse2.c (revision 4bdc94577ba0e567308109d787f7fec7b531ce36)

// Auto-generated file. Do not edit!
//   Template: src/x32-transposec/sse2.c.in
//   Generator: tools/xngen
//
// Copyright 2021 Google LLC
//
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree.

#include <immintrin.h>

#include <assert.h>

#include <xnnpack/common.h>
#include <xnnpack/math.h>
#include <xnnpack/transpose.h>
#include <xnnpack/unaligned.h>


void xnn_x16_transposec_ukernel__8x8_multi_switch_sse2(
    const uint16_t* input,
    uint16_t* output,
    size_t input_stride,
    size_t output_stride,
    size_t block_width,
    size_t block_height) XNN_OOB_READS
{
  assert(output_stride >= block_height * sizeof(uint16_t));
  assert(input_stride >= block_width * sizeof(uint16_t));

  const size_t tile_height = 8;
  const size_t tile_width = 8;
  const size_t tile_hbytes = tile_height * sizeof(uint16_t);
  const size_t tile_wbytes = tile_width * sizeof(uint16_t);
  const size_t input_reset = tile_wbytes - round_down_po2(block_height, tile_height) * input_stride;
  const size_t input_offset = tile_height * input_stride;
  const size_t output_reset = tile_width * output_stride - round_down_po2(block_height, 2) * sizeof(uint16_t);

  const uint16_t* i0 = input;
  const uint16_t* i1 = (const uint16_t*) ((uintptr_t) i0 + input_stride);
  const uint16_t* i2 = (const uint16_t*) ((uintptr_t) i1 + input_stride);
  const uint16_t* i3 = (const uint16_t*) ((uintptr_t) i2 + input_stride);
  const uint16_t* i4 = (const uint16_t*) ((uintptr_t) i3 + input_stride);
  const uint16_t* i5 = (const uint16_t*) ((uintptr_t) i4 + input_stride);
  const uint16_t* i6 = (const uint16_t*) ((uintptr_t) i5 + input_stride);
  const uint16_t* i7 = (const uint16_t*) ((uintptr_t) i6 + input_stride);
  uint16_t* o = (uint16_t*) output;
  const size_t minus_output_stride = -output_stride;

  do {
    const size_t rem = min(block_width - 1, 7);
    const size_t oN_stride = rem * output_stride;
    size_t bh = block_height;
    for (; bh >= 8; bh -= 8) {
      const __m128i v3_0 = _mm_loadu_si128((const __m128i*) i0);
      i0 = (uint16_t*) ((uintptr_t) i0 + input_offset);
      const __m128i v3_1 = _mm_loadu_si128((const __m128i*) i1);
      i1 = (uint16_t*) ((uintptr_t) i1 + input_offset);
      const __m128i v3_2 = _mm_loadu_si128((const __m128i*) i2);
      i2 = (uint16_t*) ((uintptr_t) i2 + input_offset);
      const __m128i v3_3 = _mm_loadu_si128((const __m128i*) i3);
      i3 = (uint16_t*) ((uintptr_t) i3 + input_offset);
      const __m128i v3_4 = _mm_loadu_si128((const __m128i*) i4);
      i4 = (uint16_t*) ((uintptr_t) i4 + input_offset);
      const __m128i v3_5 = _mm_loadu_si128((const __m128i*) i5);
      i5 = (uint16_t*) ((uintptr_t) i5 + input_offset);
      const __m128i v3_6 = _mm_loadu_si128((const __m128i*) i6);
      i6 = (uint16_t*) ((uintptr_t) i6 + input_offset);
      const __m128i v3_7 = _mm_loadu_si128((const __m128i*) i7);
      i7 = (uint16_t*) ((uintptr_t) i7 + input_offset);

      const __m128i v2_0 = _mm_unpacklo_epi16(v3_0, v3_1);
      const __m128i v2_1 = _mm_unpackhi_epi16(v3_0, v3_1);
      const __m128i v2_2 = _mm_unpacklo_epi16(v3_2, v3_3);
      const __m128i v2_3 = _mm_unpackhi_epi16(v3_2, v3_3);
      const __m128i v2_4 = _mm_unpacklo_epi16(v3_4, v3_5);
      const __m128i v2_5 = _mm_unpackhi_epi16(v3_4, v3_5);
      const __m128i v2_6 = _mm_unpacklo_epi16(v3_6, v3_7);
      const __m128i v2_7 = _mm_unpackhi_epi16(v3_6, v3_7);

      const __m128i v1_0 = _mm_unpacklo_epi32(v2_0, v2_2);
      const __m128i v1_1 = _mm_unpackhi_epi32(v2_0, v2_2);
      const __m128i v1_2 = _mm_unpacklo_epi32(v2_1, v2_3);
      const __m128i v1_3 = _mm_unpackhi_epi32(v2_1, v2_3);
      const __m128i v1_4 = _mm_unpacklo_epi32(v2_4, v2_6);
      const __m128i v1_5 = _mm_unpackhi_epi32(v2_4, v2_6);
      const __m128i v1_6 = _mm_unpacklo_epi32(v2_5, v2_7);
      const __m128i v1_7 = _mm_unpackhi_epi32(v2_5, v2_7);

      const __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_4);
      const __m128i v0_1 = _mm_unpackhi_epi64(v1_0, v1_4);
      const __m128i v0_2 = _mm_unpacklo_epi64(v1_1, v1_5);
      const __m128i v0_3 = _mm_unpackhi_epi64(v1_1, v1_5);
      const __m128i v0_4 = _mm_unpacklo_epi64(v1_2, v1_6);
      const __m128i v0_5 = _mm_unpackhi_epi64(v1_2, v1_6);
      const __m128i v0_6 = _mm_unpacklo_epi64(v1_3, v1_7);
      const __m128i v0_7 = _mm_unpackhi_epi64(v1_3, v1_7);


      uint16_t* oN = (uint16_t*) ((uintptr_t) o + oN_stride);
      switch (rem) {
        case 7:
          _mm_storeu_si128((__m128i*) oN, v0_7);
          oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
        case 6:
          _mm_storeu_si128((__m128i*) oN, v0_6);
          oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
        case 5:
          _mm_storeu_si128((__m128i*) oN, v0_5);
          oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
        case 4:
          _mm_storeu_si128((__m128i*) oN, v0_4);
          oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
        case 3:
          _mm_storeu_si128((__m128i*) oN, v0_3);
          oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
        case 2:
          _mm_storeu_si128((__m128i*) oN, v0_2);
          oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
        case 1:
          _mm_storeu_si128((__m128i*) oN, v0_1);
        case 0:
          _mm_storeu_si128((__m128i*) o, v0_0);
          o = (uint16_t*) ((uintptr_t) o + tile_hbytes);
          break;
        default:
          XNN_UNREACHABLE;
      }
    }
    if (bh != 0) {
      const __m128i v3_0 = _mm_loadu_si128((const __m128i*) i0);
      if XNN_UNPREDICTABLE(bh < 2) {
        i1 = i0;
      }
      const __m128i v3_1 = _mm_loadu_si128((const __m128i*) i1);
      if XNN_UNPREDICTABLE(bh <= 2) {
        i2 = i0;
      }
      const __m128i v3_2 = _mm_loadu_si128((const __m128i*) i2);
      if XNN_UNPREDICTABLE(bh < 4) {
        i3 = i0;
      }
      const __m128i v3_3 = _mm_loadu_si128((const __m128i*) i3);
      if XNN_UNPREDICTABLE(bh <= 4) {
        i4 = i0;
      }
      const __m128i v3_4 = _mm_loadu_si128((const __m128i*) i4);
      if XNN_UNPREDICTABLE(bh < 6) {
        i5 = i0;
      }
      const __m128i v3_5 = _mm_loadu_si128((const __m128i*) i5);
      if XNN_UNPREDICTABLE(bh <= 6) {
        i6 = i0;
      }
      const __m128i v3_6 = _mm_loadu_si128((const __m128i*) i6);
      const __m128i v3_7 = _mm_undefined_si128();

      const __m128i v2_0 = _mm_unpacklo_epi16(v3_0, v3_1);
      const __m128i v2_1 = _mm_unpackhi_epi16(v3_0, v3_1);
      const __m128i v2_2 = _mm_unpacklo_epi16(v3_2, v3_3);
      const __m128i v2_3 = _mm_unpackhi_epi16(v3_2, v3_3);
      const __m128i v2_4 = _mm_unpacklo_epi16(v3_4, v3_5);
      const __m128i v2_5 = _mm_unpackhi_epi16(v3_4, v3_5);
      const __m128i v2_6 = _mm_unpacklo_epi16(v3_6, v3_7);
      const __m128i v2_7 = _mm_unpackhi_epi16(v3_6, v3_7);

      const __m128i v1_0 = _mm_unpacklo_epi32(v2_0, v2_2);
      const __m128i v1_1 = _mm_unpackhi_epi32(v2_0, v2_2);
      const __m128i v1_2 = _mm_unpacklo_epi32(v2_1, v2_3);
      const __m128i v1_3 = _mm_unpackhi_epi32(v2_1, v2_3);
      const __m128i v1_4 = _mm_unpacklo_epi32(v2_4, v2_6);
      const __m128i v1_5 = _mm_unpackhi_epi32(v2_4, v2_6);
      const __m128i v1_6 = _mm_unpacklo_epi32(v2_5, v2_7);
      const __m128i v1_7 = _mm_unpackhi_epi32(v2_5, v2_7);

      __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_4);
      __m128i v0_1 = _mm_unpackhi_epi64(v1_0, v1_4);
      __m128i v0_2 = _mm_unpacklo_epi64(v1_1, v1_5);
      __m128i v0_3 = _mm_unpackhi_epi64(v1_1, v1_5);
      __m128i v0_4 = _mm_unpacklo_epi64(v1_2, v1_6);
      __m128i v0_5 = _mm_unpackhi_epi64(v1_2, v1_6);
      __m128i v0_6 = _mm_unpacklo_epi64(v1_3, v1_7);
      __m128i v0_7 = _mm_unpackhi_epi64(v1_3, v1_7);


      if (bh & 4) {
        uint16_t* oN = (uint16_t*) ((uintptr_t) o + oN_stride);
        switch (rem) {
          case 7:
            _mm_storel_epi64((__m128i*) oN, v0_7);
            oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
          case 6:
            _mm_storel_epi64((__m128i*) oN, v0_6);
            oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
          case 5:
            _mm_storel_epi64((__m128i*) oN, v0_5);
            oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
          case 4:
            _mm_storel_epi64((__m128i*) oN, v0_4);
            oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
          case 3:
            _mm_storel_epi64((__m128i*) oN, v0_3);
            oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
          case 2:
            _mm_storel_epi64((__m128i*) oN, v0_2);
            oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
          case 1:
            _mm_storel_epi64((__m128i*) oN, v0_1);
          case 0:
            _mm_storel_epi64((__m128i*) o, v0_0);
            break;
          default:
            XNN_UNREACHABLE;
        }
        o += 4;
        v0_0 = _mm_unpackhi_epi64(v0_0, v0_0);
        v0_1 = _mm_unpackhi_epi64(v0_1, v0_1);
        v0_2 = _mm_unpackhi_epi64(v0_2, v0_2);
        v0_3 = _mm_unpackhi_epi64(v0_3, v0_3);
        v0_4 = _mm_unpackhi_epi64(v0_4, v0_4);
        v0_5 = _mm_unpackhi_epi64(v0_5, v0_5);
        v0_6 = _mm_unpackhi_epi64(v0_6, v0_6);
        v0_7 = _mm_unpackhi_epi64(v0_7, v0_7);
      }

      if (bh & 2) {
        uint16_t* oN = (uint16_t*) ((uintptr_t) o + oN_stride);
        switch (rem) {
          case 7:
            unaligned_store_u32(oN, (uint32_t) _mm_cvtsi128_si32(v0_7));
            oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
          case 6:
            unaligned_store_u32(oN, (uint32_t) _mm_cvtsi128_si32(v0_6));
            oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
          case 5:
            unaligned_store_u32(oN, (uint32_t) _mm_cvtsi128_si32(v0_5));
            oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
          case 4:
            unaligned_store_u32(oN, (uint32_t) _mm_cvtsi128_si32(v0_4));
            oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
          case 3:
            unaligned_store_u32(oN, (uint32_t) _mm_cvtsi128_si32(v0_3));
            oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
          case 2:
            unaligned_store_u32(oN, (uint32_t) _mm_cvtsi128_si32(v0_2));
            oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
          case 1:
            unaligned_store_u32(oN, (uint32_t) _mm_cvtsi128_si32(v0_1));
          case 0:
            unaligned_store_u32(o, (uint32_t) _mm_cvtsi128_si32(v0_0));
            break;
          default:
            XNN_UNREACHABLE;
        }
        o += 2;
        v0_0 = _mm_srli_epi64(v0_0, 32);
        v0_1 = _mm_srli_epi64(v0_1, 32);
        v0_2 = _mm_srli_epi64(v0_2, 32);
        v0_3 = _mm_srli_epi64(v0_3, 32);
        v0_4 = _mm_srli_epi64(v0_4, 32);
        v0_5 = _mm_srli_epi64(v0_5, 32);
        v0_6 = _mm_srli_epi64(v0_6, 32);
        v0_7 = _mm_srli_epi64(v0_7, 32);
      }
      if (bh & 1) {
        uint16_t* oN = (uint16_t*) ((uintptr_t) o + oN_stride);
        switch (rem) {
          case 7:
            unaligned_store_u16(oN, (uint16_t) _mm_cvtsi128_si32(v0_7));
            oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
          case 6:
            unaligned_store_u16(oN, (uint16_t) _mm_cvtsi128_si32(v0_6));
            oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
          case 5:
            unaligned_store_u16(oN, (uint16_t) _mm_cvtsi128_si32(v0_5));
            oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
          case 4:
            unaligned_store_u16(oN, (uint16_t) _mm_cvtsi128_si32(v0_4));
            oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
          case 3:
            unaligned_store_u16(oN, (uint16_t) _mm_cvtsi128_si32(v0_3));
            oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
          case 2:
            unaligned_store_u16(oN, (uint16_t) _mm_cvtsi128_si32(v0_2));
            oN = (uint16_t*) ((uintptr_t) oN + minus_output_stride);
          case 1:
            unaligned_store_u16(oN, (uint16_t) _mm_cvtsi128_si32(v0_1));
          case 0:
            unaligned_store_u16(o, (uint16_t) _mm_cvtsi128_si32(v0_0));
            break;
          default:
            XNN_UNREACHABLE;
        }
      }
    }

    i0 = (const uint16_t*) ((uintptr_t) i0 + input_reset);
    i1 = (const uint16_t*) ((uintptr_t) i0 + input_stride);
    i2 = (const uint16_t*) ((uintptr_t) i1 + input_stride);
    i3 = (const uint16_t*) ((uintptr_t) i2 + input_stride);
    i4 = (const uint16_t*) ((uintptr_t) i3 + input_stride);
    i5 = (const uint16_t*) ((uintptr_t) i4 + input_stride);
    i6 = (const uint16_t*) ((uintptr_t) i5 + input_stride);
    i7 = (const uint16_t*) ((uintptr_t) i6 + input_stride);
    o = (uint16_t*) ((uintptr_t) o + output_reset);
    block_width = doz(block_width, tile_width);
  } while (block_width != 0);
}