xref: /aosp_15_r20/external/XNNPACK/src/qu8-gemm/gen/4x16c8-minmax-fp32-avx512skx.c (revision 4bdc94577ba0e567308109d787f7fec7b531ce36)

// Auto-generated file. Do not edit!
//   Template: src/qs8-gemm/MRx16c8-avx512skx.c.in
//   Generator: tools/xngen
//
// Copyright 2020 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 <assert.h>

#include <immintrin.h>

#include <xnnpack/gemm.h>
#include <xnnpack/intrinsics-polyfill.h>
#include <xnnpack/math.h>


void xnn_qu8_gemm_minmax_fp32_ukernel_4x16c8__avx512skx(
    size_t mr,
    size_t nc,
    size_t kc,
    const uint8_t* restrict a,
    size_t a_stride,
    const void* restrict w,
    uint8_t* restrict c,
    size_t cm_stride,
    size_t cn_stride,
    const union xnn_qu8_conv_minmax_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{
  assert(mr != 0);
  assert(mr <= 4);
  assert(nc != 0);
  assert(kc != 0);
  assert(kc % sizeof(uint8_t) == 0);
  assert(a != NULL);
  assert(w != NULL);
  assert(c != NULL);

  kc = round_up_po2(kc, 8);
  const uint8_t* a0 = a;
  uint8_t* c0 = c;
  const uint8_t* a1 = (const uint8_t*) ((uintptr_t) a0 + a_stride);
  uint8_t* c1 = (uint8_t*) ((uintptr_t) c0 + cm_stride);
  if XNN_UNPREDICTABLE(mr < 2) {
    a1 = a0;
    c1 = c0;
  }
  const uint8_t* a2 = (const uint8_t*) ((uintptr_t) a1 + a_stride);
  uint8_t* c2 = (uint8_t*) ((uintptr_t) c1 + cm_stride);
  if XNN_UNPREDICTABLE(mr <= 2) {
    a2 = a1;
    c2 = c1;
  }
  const uint8_t* a3 = (const uint8_t*) ((uintptr_t) a2 + a_stride);
  uint8_t* c3 = (uint8_t*) ((uintptr_t) c2 + cm_stride);
  if XNN_UNPREDICTABLE(mr != 4) {
    a3 = a2;
    c3 = c2;
  }

  const __mmask16 vbias_mask = _cvtu32_mask16(0x1111);
  const __m512 vscale = _mm512_load_ps(params->fp32_avx512.scale);
  const __m512 voutput_max_less_zero_point = _mm512_load_ps(params->fp32_avx512.output_max_less_zero_point);
  const __m512i voutput_zero_point = _mm512_load_si512(params->fp32_avx512.output_zero_point);
  const __m512i voutput_min = _mm512_load_si512(params->fp32_avx512.output_min);
  do {
    __m512i vacc0x0123 = _mm512_maskz_expandloadu_epi32(vbias_mask, w);
    __m512i vacc0x4567 = _mm512_maskz_expandloadu_epi32(vbias_mask, (const void*) ((const int32_t*) w + 4));
    __m512i vacc0x89AB = _mm512_maskz_expandloadu_epi32(vbias_mask, (const void*) ((const int32_t*) w + 8));
    __m512i vacc0xCDEF = _mm512_maskz_expandloadu_epi32(vbias_mask, (const void*) ((const int32_t*) w + 12));
    __m512i vacc1x0123 = vacc0x0123;
    __m512i vacc1x4567 = vacc0x4567;
    __m512i vacc1x89AB = vacc0x89AB;
    __m512i vacc1xCDEF = vacc0xCDEF;
    __m512i vacc2x0123 = vacc0x0123;
    __m512i vacc2x4567 = vacc0x4567;
    __m512i vacc2x89AB = vacc0x89AB;
    __m512i vacc2xCDEF = vacc0xCDEF;
    __m512i vacc3x0123 = vacc0x0123;
    __m512i vacc3x4567 = vacc0x4567;
    __m512i vacc3x89AB = vacc0x89AB;
    __m512i vacc3xCDEF = vacc0xCDEF;
    w = (const void*) ((const int32_t*) w + 16);

    size_t k = 0;
    const __m512i vb_zero_point = _mm512_load_si512(params->fp32_avx512.kernel_zero_point);
    while (k < kc) {
      const __m512i va0 = _mm512_broadcast_i32x4(_mm_cvtepu8_epi16(_mm_loadl_epi64((const __m128i*) a0)));
      a0 += 8;
      const __m512i va1 = _mm512_broadcast_i32x4(_mm_cvtepu8_epi16(_mm_loadl_epi64((const __m128i*) a1)));
      a1 += 8;
      const __m512i va2 = _mm512_broadcast_i32x4(_mm_cvtepu8_epi16(_mm_loadl_epi64((const __m128i*) a2)));
      a2 += 8;
      const __m512i va3 = _mm512_broadcast_i32x4(_mm_cvtepu8_epi16(_mm_loadl_epi64((const __m128i*) a3)));
      a3 += 8;

      const __m512i vb0123 = _mm512_sub_epi16(_mm512_cvtepu8_epi16(_mm256_load_si256((const __m256i*) w)), vb_zero_point);

      vacc0x0123 = _mm512_add_epi32(vacc0x0123, _mm512_madd_epi16(va0, vb0123));
      vacc1x0123 = _mm512_add_epi32(vacc1x0123, _mm512_madd_epi16(va1, vb0123));
      vacc2x0123 = _mm512_add_epi32(vacc2x0123, _mm512_madd_epi16(va2, vb0123));
      vacc3x0123 = _mm512_add_epi32(vacc3x0123, _mm512_madd_epi16(va3, vb0123));
      const __m512i vb4567 = _mm512_sub_epi16(_mm512_cvtepu8_epi16(_mm256_load_si256((const __m256i*) ((const uint8_t*) w + 32))), vb_zero_point);

      vacc0x4567 = _mm512_add_epi32(vacc0x4567, _mm512_madd_epi16(va0, vb4567));
      vacc1x4567 = _mm512_add_epi32(vacc1x4567, _mm512_madd_epi16(va1, vb4567));
      vacc2x4567 = _mm512_add_epi32(vacc2x4567, _mm512_madd_epi16(va2, vb4567));
      vacc3x4567 = _mm512_add_epi32(vacc3x4567, _mm512_madd_epi16(va3, vb4567));
      const __m512i vb89AB = _mm512_sub_epi16(_mm512_cvtepu8_epi16(_mm256_load_si256((const __m256i*) ((const uint8_t*) w + 64))), vb_zero_point);

      vacc0x89AB = _mm512_add_epi32(vacc0x89AB, _mm512_madd_epi16(va0, vb89AB));
      vacc1x89AB = _mm512_add_epi32(vacc1x89AB, _mm512_madd_epi16(va1, vb89AB));
      vacc2x89AB = _mm512_add_epi32(vacc2x89AB, _mm512_madd_epi16(va2, vb89AB));
      vacc3x89AB = _mm512_add_epi32(vacc3x89AB, _mm512_madd_epi16(va3, vb89AB));
      const __m512i vbCDEF = _mm512_sub_epi16(_mm512_cvtepu8_epi16(_mm256_load_si256((const __m256i*) ((const uint8_t*) w + 96))), vb_zero_point);

      vacc0xCDEF = _mm512_add_epi32(vacc0xCDEF, _mm512_madd_epi16(va0, vbCDEF));
      vacc1xCDEF = _mm512_add_epi32(vacc1xCDEF, _mm512_madd_epi16(va1, vbCDEF));
      vacc2xCDEF = _mm512_add_epi32(vacc2xCDEF, _mm512_madd_epi16(va2, vbCDEF));
      vacc3xCDEF = _mm512_add_epi32(vacc3xCDEF, _mm512_madd_epi16(va3, vbCDEF));

      w = (const void*) ((const uint8_t*) w + 128);
      k += 8 * sizeof(uint8_t);
    }

    const __m512i vacc0x04152637 = _mm512_add_epi32(_mm512_unpacklo_epi32(vacc0x0123, vacc0x4567), _mm512_unpackhi_epi32(vacc0x0123, vacc0x4567));
    const __m512i vacc0x8C9DAEBF = _mm512_add_epi32(_mm512_unpacklo_epi32(vacc0x89AB, vacc0xCDEF), _mm512_unpackhi_epi32(vacc0x89AB, vacc0xCDEF));
    const __m512i vacc1x04152637 = _mm512_add_epi32(_mm512_unpacklo_epi32(vacc1x0123, vacc1x4567), _mm512_unpackhi_epi32(vacc1x0123, vacc1x4567));
    const __m512i vacc1x8C9DAEBF = _mm512_add_epi32(_mm512_unpacklo_epi32(vacc1x89AB, vacc1xCDEF), _mm512_unpackhi_epi32(vacc1x89AB, vacc1xCDEF));
    const __m512i vacc2x04152637 = _mm512_add_epi32(_mm512_unpacklo_epi32(vacc2x0123, vacc2x4567), _mm512_unpackhi_epi32(vacc2x0123, vacc2x4567));
    const __m512i vacc2x8C9DAEBF = _mm512_add_epi32(_mm512_unpacklo_epi32(vacc2x89AB, vacc2xCDEF), _mm512_unpackhi_epi32(vacc2x89AB, vacc2xCDEF));
    const __m512i vacc3x04152637 = _mm512_add_epi32(_mm512_unpacklo_epi32(vacc3x0123, vacc3x4567), _mm512_unpackhi_epi32(vacc3x0123, vacc3x4567));
    const __m512i vacc3x8C9DAEBF = _mm512_add_epi32(_mm512_unpacklo_epi32(vacc3x89AB, vacc3xCDEF), _mm512_unpackhi_epi32(vacc3x89AB, vacc3xCDEF));

    __m512i vacc0x084C195D2A6E3B7F = _mm512_add_epi32(_mm512_unpacklo_epi32(vacc0x04152637, vacc0x8C9DAEBF), _mm512_unpackhi_epi32(vacc0x04152637, vacc0x8C9DAEBF));
    __m512i vacc1x084C195D2A6E3B7F = _mm512_add_epi32(_mm512_unpacklo_epi32(vacc1x04152637, vacc1x8C9DAEBF), _mm512_unpackhi_epi32(vacc1x04152637, vacc1x8C9DAEBF));
    __m512i vacc2x084C195D2A6E3B7F = _mm512_add_epi32(_mm512_unpacklo_epi32(vacc2x04152637, vacc2x8C9DAEBF), _mm512_unpackhi_epi32(vacc2x04152637, vacc2x8C9DAEBF));
    __m512i vacc3x084C195D2A6E3B7F = _mm512_add_epi32(_mm512_unpacklo_epi32(vacc3x04152637, vacc3x8C9DAEBF), _mm512_unpackhi_epi32(vacc3x04152637, vacc3x8C9DAEBF));

    __m512 vscaled0x084C195D2A6E3B7F = _mm512_cvtepi32_ps(vacc0x084C195D2A6E3B7F);
    __m512 vscaled1x084C195D2A6E3B7F = _mm512_cvtepi32_ps(vacc1x084C195D2A6E3B7F);
    __m512 vscaled2x084C195D2A6E3B7F = _mm512_cvtepi32_ps(vacc2x084C195D2A6E3B7F);
    __m512 vscaled3x084C195D2A6E3B7F = _mm512_cvtepi32_ps(vacc3x084C195D2A6E3B7F);

    vscaled0x084C195D2A6E3B7F = _mm512_mul_ps(vscaled0x084C195D2A6E3B7F, vscale);
    vscaled1x084C195D2A6E3B7F = _mm512_mul_ps(vscaled1x084C195D2A6E3B7F, vscale);
    vscaled2x084C195D2A6E3B7F = _mm512_mul_ps(vscaled2x084C195D2A6E3B7F, vscale);
    vscaled3x084C195D2A6E3B7F = _mm512_mul_ps(vscaled3x084C195D2A6E3B7F, vscale);

    vscaled0x084C195D2A6E3B7F = _mm512_min_ps(vscaled0x084C195D2A6E3B7F, voutput_max_less_zero_point);
    vscaled1x084C195D2A6E3B7F = _mm512_min_ps(vscaled1x084C195D2A6E3B7F, voutput_max_less_zero_point);
    vscaled2x084C195D2A6E3B7F = _mm512_min_ps(vscaled2x084C195D2A6E3B7F, voutput_max_less_zero_point);
    vscaled3x084C195D2A6E3B7F = _mm512_min_ps(vscaled3x084C195D2A6E3B7F, voutput_max_less_zero_point);

    vacc0x084C195D2A6E3B7F = _mm512_cvtps_epi32(vscaled0x084C195D2A6E3B7F);
    vacc1x084C195D2A6E3B7F = _mm512_cvtps_epi32(vscaled1x084C195D2A6E3B7F);
    vacc2x084C195D2A6E3B7F = _mm512_cvtps_epi32(vscaled2x084C195D2A6E3B7F);
    vacc3x084C195D2A6E3B7F = _mm512_cvtps_epi32(vscaled3x084C195D2A6E3B7F);

    const __m512i vacc01x084Cx195Dx2A6Ex3B7F = _mm512_adds_epi16(_mm512_packs_epi32(vacc0x084C195D2A6E3B7F, vacc1x084C195D2A6E3B7F), voutput_zero_point);
    const __m512i vacc23x084Cx195Dx2A6Ex3B7F = _mm512_adds_epi16(_mm512_packs_epi32(vacc2x084C195D2A6E3B7F, vacc3x084C195D2A6E3B7F), voutput_zero_point);

    __m512i vout0123x084Cx195Dx2A6Ex3B7F = _mm512_packus_epi16(vacc01x084Cx195Dx2A6Ex3B7F, vacc23x084Cx195Dx2A6Ex3B7F);
    vout0123x084Cx195Dx2A6Ex3B7F = _mm512_permutexvar_epi32(_mm512_set_epi32(15, 11, 7, 3, 14, 10, 6, 2, 13, 9, 5, 1, 12, 8, 4, 0), vout0123x084Cx195Dx2A6Ex3B7F);
    __m512i vout0123x0123456789ABCDEF = _mm512_shuffle_epi8(vout0123x084Cx195Dx2A6Ex3B7F, _mm512_set_epi8(15, 11, 7, 3, 13, 9, 5, 1, 14, 10, 6, 2, 12, 8, 4, 0, 15, 11, 7, 3, 13, 9, 5, 1, 14, 10, 6, 2, 12, 8, 4, 0, 15, 11, 7, 3, 13, 9, 5, 1, 14, 10, 6, 2, 12, 8, 4, 0, 15, 11, 7, 3, 13, 9, 5, 1, 14, 10, 6, 2, 12, 8, 4, 0));
    vout0123x0123456789ABCDEF = _mm512_max_epu8(vout0123x0123456789ABCDEF, voutput_min);

    if (nc >= 16) {
      _mm_storeu_si128((__m128i*) c0, _mm512_castsi512_si128(vout0123x0123456789ABCDEF));
      _mm_storeu_si128((__m128i*) c1, _mm512_extracti32x4_epi32(vout0123x0123456789ABCDEF, 1));
      _mm_storeu_si128((__m128i*) c2, _mm512_extracti32x4_epi32(vout0123x0123456789ABCDEF, 2));
      _mm_storeu_si128((__m128i*) c3, _mm512_extracti32x4_epi32(vout0123x0123456789ABCDEF, 3));

      a0 = (const uint8_t*) ((uintptr_t) a0 - k);
      a1 = (const uint8_t*) ((uintptr_t) a1 - k);
      a2 = (const uint8_t*) ((uintptr_t) a2 - k);
      a3 = (const uint8_t*) ((uintptr_t) a3 - k);

      c0 = (uint8_t*) ((uintptr_t) c0 + cn_stride);
      c1 = (uint8_t*) ((uintptr_t) c1 + cn_stride);
      c2 = (uint8_t*) ((uintptr_t) c2 + cn_stride);
      c3 = (uint8_t*) ((uintptr_t) c3 + cn_stride);

      nc -= 16;
    } else {
      // Prepare mask for valid 8-bit elements (depends on nc).
      __mmask64 vmask = _cvtu64_mask64((uint64_t) ((UINT32_C(1) << nc) - UINT32_C(1)));

      _mm512_mask_storeu_epi8(c0, vmask, vout0123x0123456789ABCDEF);
      vmask = _kshiftli_mask64(vmask, 16);
      _mm512_mask_storeu_epi8(c1 - 16, vmask, vout0123x0123456789ABCDEF);
      vmask = _kshiftli_mask64(vmask, 16);
      _mm512_mask_storeu_epi8(c2 - 32, vmask, vout0123x0123456789ABCDEF);
      vmask = _kshiftli_mask64(vmask, 16);
      _mm512_mask_storeu_epi8(c3 - 48, vmask, vout0123x0123456789ABCDEF);

      nc = 0;
    }
  } while (nc != 0);
}