#include <assert.h>
#include <math.h>
#include <stddef.h>
#include <stdint.h>
#include <immintrin.h>
#include "xnnpack/avgpool.h"
#include "xnnpack/common.h"
#include "xnnpack/gavgpool.h"
#include "xnnpack/intrinsics-polyfill.h"
#include "xnnpack/math.h"
#include "xnnpack/maxpool.h"
#include "xnnpack/microparams.h"
#include "xnnpack/prelu.h"
#include "xnnpack/reduce.h"
#include "xnnpack/unaligned.h"
#include "xnnpack/vbinary.h"
#include "xnnpack/vcvt.h"
#include "xnnpack/vunary.h"
void xnn_f16_avgpool_minmax_ukernel_9p8x__f16c_c8(
size_t output_pixels,
size_t kernel_elements,
size_t channels,
const void** input,
size_t input_offset,
const void* zero,
void* buffer,
void* output,
size_t input_increment,
size_t output_increment,
const union xnn_f16_scaleminmax_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_avgpool_minmax_ukernel_9x__f16c_c8(
size_t output_pixels,
size_t kernel_elements,
size_t channels,
const void** input,
size_t input_offset,
const void* zero,
void* output,
size_t input_increment,
size_t output_increment,
const union xnn_f16_scaleminmax_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_f32_vcvt_ukernel__f16c_u16(
size_t batch,
const void* input,
float* output,
const union xnn_f16_f32_cvt_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_f32acc_rdsum_ukernel_7p7x__f16c_c32(
size_t rows,
size_t channels,
const void* input,
size_t input_stride,
const void* zero,
float* output,
const union xnn_f16_f32acc_scale_params params[restrict XNN_MIN_ELEMENTS(1)])
{ … }
void xnn_f16_f32acc_rsum_ukernel__f16c_u32_acc4(
size_t batch,
const void* input,
float* output,
const union xnn_f16_f32acc_scale_params params[restrict XNN_MIN_ELEMENTS(1)])
{ … }
void xnn_f16_gavgpool_minmax_ukernel_7p7x__f16c_c8(
size_t rows,
size_t channels,
const void* input,
size_t input_stride,
const void* zero,
void* buffer,
void* output,
const union xnn_f16_scaleminmax_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_gavgpool_minmax_ukernel_7x__f16c_c8(
size_t rows,
size_t channels,
const void* input,
size_t input_stride,
const void* zero,
void* output,
const union xnn_f16_scaleminmax_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_maxpool_minmax_ukernel_9p8x__f16c_c8(
size_t output_pixels,
size_t kernel_elements,
size_t channels,
const void** input,
size_t input_offset,
void* output,
size_t input_increment,
size_t output_increment,
const union xnn_f16_minmax_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_prelu_ukernel__f16c_2x16(
size_t rows,
size_t channels,
const void* restrict input,
size_t input_stride,
const void* restrict weights,
void* restrict output,
size_t output_stride) XNN_OOB_READS
{ … }
void xnn_f16_rmax_ukernel__f16c_u32(
size_t batch,
const void* input,
void* output,
const union xnn_f16_default_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vadd_minmax_ukernel__f16c_u16(
size_t batch,
const void* restrict input_a,
const void* restrict input_b,
void* restrict output,
const union xnn_f16_minmax_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vaddc_minmax_ukernel__f16c_u16(
size_t batch,
const void* restrict input_a,
const void* restrict input_b,
void* restrict output,
const union xnn_f16_minmax_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vdiv_minmax_ukernel__f16c_u8(
size_t batch,
const void* restrict input_a,
const void* restrict input_b,
void* restrict output,
const union xnn_f16_minmax_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vdivc_minmax_ukernel__f16c_u8(
size_t batch,
const void* restrict input_a,
const void* restrict input_b,
void* restrict output,
const union xnn_f16_minmax_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vmax_ukernel__f16c_u16(
size_t batch,
const void* restrict input_a,
const void* restrict input_b,
void* restrict output,
const union xnn_f16_default_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vmaxc_ukernel__f16c_u16(
size_t batch,
const void* restrict input_a,
const void* restrict input_b,
void* restrict output,
const union xnn_f16_default_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vmin_ukernel__f16c_u16(
size_t batch,
const void* restrict input_a,
const void* restrict input_b,
void* restrict output,
const union xnn_f16_default_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vminc_ukernel__f16c_u16(
size_t batch,
const void* restrict input_a,
const void* restrict input_b,
void* restrict output,
const union xnn_f16_default_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vmul_minmax_ukernel__f16c_u16(
size_t batch,
const void* restrict input_a,
const void* restrict input_b,
void* restrict output,
const union xnn_f16_minmax_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vmulc_minmax_ukernel__f16c_u16(
size_t batch,
const void* restrict input_a,
const void* restrict input_b,
void* restrict output,
const union xnn_f16_minmax_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vrdivc_minmax_ukernel__f16c_u8(
size_t batch,
const void* restrict input_a,
const void* restrict input_b,
void* restrict output,
const union xnn_f16_minmax_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vrsubc_minmax_ukernel__f16c_u16(
size_t batch,
const void* restrict input_a,
const void* restrict input_b,
void* restrict output,
const union xnn_f16_minmax_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vsqrdiff_ukernel__f16c_u16(
size_t batch,
const void* restrict input_a,
const void* restrict input_b,
void* restrict output,
const union xnn_f16_default_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vsqrdiffc_ukernel__f16c_u16(
size_t batch,
const void* restrict input_a,
const void* restrict input_b,
void* restrict output,
const union xnn_f16_default_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vsub_minmax_ukernel__f16c_u16(
size_t batch,
const void* restrict input_a,
const void* restrict input_b,
void* restrict output,
const union xnn_f16_minmax_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vsubc_minmax_ukernel__f16c_u16(
size_t batch,
const void* restrict input_a,
const void* restrict input_b,
void* restrict output,
const union xnn_f16_minmax_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vclamp_ukernel__f16c_u16(
size_t batch,
const void* restrict input,
void* restrict output,
const union xnn_f16_minmax_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vhswish_ukernel__f16c_u16(
size_t batch,
const void* restrict input,
void* restrict output,
const union xnn_f16_hswish_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vlrelu_ukernel__f16c_u16(
size_t batch,
const void* input,
void* output,
const union xnn_f16_lrelu_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vrndd_ukernel__f16c_u16(
size_t batch,
const void* input,
void* output,
const union xnn_f16_rnd_params params[restrict XNN_MIN_ELEMENTS(1)])
{ … }
void xnn_f16_vrndne_ukernel__f16c_u16(
size_t batch,
const void* input,
void* output,
const union xnn_f16_rnd_params params[restrict XNN_MIN_ELEMENTS(1)])
{ … }
void xnn_f16_vrndu_ukernel__f16c_u16(
size_t batch,
const void* input,
void* output,
const union xnn_f16_rnd_params params[restrict XNN_MIN_ELEMENTS(1)])
{ … }
void xnn_f16_vrndz_ukernel__f16c_u16(
size_t batch,
const void* input,
void* output,
const union xnn_f16_rnd_params params[restrict XNN_MIN_ELEMENTS(1)])
{ … }
void xnn_f16_vrsqrt_ukernel__f16c_rsqrt_u32(
size_t batch,
const void* input,
void* output,
const union xnn_f16_rsqrt_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vsqrt_ukernel__f16c_rsqrt_u32(
size_t batch,
const void* input,
void* output,
const union xnn_f16_sqrt_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vtanh_ukernel__f16c_expm1minus_rr1_p3h2ts_rcp_u72(
size_t batch,
const void* input,
void* output,
const union xnn_f16_tanh_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f16_vsqr_ukernel__f16c_u16(
size_t batch,
const void* input,
void* output,
const union xnn_f16_default_params params[restrict XNN_MIN_ELEMENTS(1)]) XNN_OOB_READS
{ … }
void xnn_f32_f16_vcvt_ukernel__f16c_u16(
size_t batch,
const float* input,
void* output,
const union xnn_f32_f16_cvt_params params[restrict XNN_MIN_ELEMENTS(1)])
{ … }