| // Auto-generated file. Do not edit! |
| // Template: src/qs8-dwconv/unipass-scalar.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 <assert.h> |
| |
| #include <fp16.h> |
| |
| #include <xnnpack/dwconv.h> |
| #include <xnnpack/math.h> |
| |
| |
| void xnn_qc8_dwconv_minmax_fp32_ukernel_up4x9__scalar_fmagic( |
| size_t channels, |
| size_t output_width, |
| const int8_t** input, |
| const void* weights, |
| int8_t* output, |
| size_t input_stride, |
| size_t output_increment, |
| size_t input_offset, |
| const int8_t* zero, |
| const union xnn_qs8_minmax_params params[restrict XNN_MIN_ELEMENTS(1)]) |
| { |
| assert(channels != 0); |
| assert(output_width != 0); |
| |
| const float voutput_min_less_zero_point = params->scalar_fmagic.output_min_less_zero_point; |
| const float voutput_max_less_zero_point = params->scalar_fmagic.output_max_less_zero_point; |
| const float vmagic_bias = params->scalar_fmagic.magic_bias; |
| const int32_t vmagic_bias_less_output_zero_point = params->scalar_fmagic.magic_bias_less_output_zero_point; |
| do { |
| const int8_t* i0 = input[0]; |
| assert(i0 != NULL); |
| if XNN_UNPREDICTABLE(i0 != zero) { |
| i0 = (const int8_t*) ((uintptr_t) i0 + input_offset); |
| } |
| const int8_t* i1 = input[1]; |
| assert(i1 != NULL); |
| if XNN_UNPREDICTABLE(i1 != zero) { |
| i1 = (const int8_t*) ((uintptr_t) i1 + input_offset); |
| } |
| const int8_t* i2 = input[2]; |
| assert(i2 != NULL); |
| if XNN_UNPREDICTABLE(i2 != zero) { |
| i2 = (const int8_t*) ((uintptr_t) i2 + input_offset); |
| } |
| const int8_t* i3 = input[3]; |
| assert(i3 != NULL); |
| if XNN_UNPREDICTABLE(i3 != zero) { |
| i3 = (const int8_t*) ((uintptr_t) i3 + input_offset); |
| } |
| const int8_t* i4 = input[4]; |
| assert(i4 != NULL); |
| if XNN_UNPREDICTABLE(i4 != zero) { |
| i4 = (const int8_t*) ((uintptr_t) i4 + input_offset); |
| } |
| const int8_t* i5 = input[5]; |
| assert(i5 != NULL); |
| if XNN_UNPREDICTABLE(i5 != zero) { |
| i5 = (const int8_t*) ((uintptr_t) i5 + input_offset); |
| } |
| const int8_t* i6 = input[6]; |
| assert(i6 != NULL); |
| if XNN_UNPREDICTABLE(i6 != zero) { |
| i6 = (const int8_t*) ((uintptr_t) i6 + input_offset); |
| } |
| const int8_t* i7 = input[7]; |
| assert(i7 != NULL); |
| if XNN_UNPREDICTABLE(i7 != zero) { |
| i7 = (const int8_t*) ((uintptr_t) i7 + input_offset); |
| } |
| const int8_t* i8 = input[8]; |
| assert(i8 != NULL); |
| if XNN_UNPREDICTABLE(i8 != zero) { |
| i8 = (const int8_t*) ((uintptr_t) i8 + input_offset); |
| } |
| input = (const int8_t**) ((uintptr_t) input + input_stride); |
| |
| size_t c = channels; |
| const void* w = weights; |
| for (; c >= 4; c -= 4) { |
| int32_t vacc0 = ((const int32_t*) w)[0]; |
| int32_t vacc1 = ((const int32_t*) w)[1]; |
| int32_t vacc2 = ((const int32_t*) w)[2]; |
| int32_t vacc3 = ((const int32_t*) w)[3]; |
| |
| |
| const int32_t vi0x0 = (int32_t) i0[0]; |
| const int32_t vi0x1 = (int32_t) i0[1]; |
| const int32_t vi0x2 = (int32_t) i0[2]; |
| const int32_t vi0x3 = (int32_t) i0[3]; |
| i0 += 4; |
| |
| const int32_t vk0x0 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[0]; |
| const int32_t vk0x1 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[1]; |
| const int32_t vk0x2 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[2]; |
| const int32_t vk0x3 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[3]; |
| |
| vacc0 += vi0x0 * vk0x0; |
| vacc1 += vi0x1 * vk0x1; |
| vacc2 += vi0x2 * vk0x2; |
| vacc3 += vi0x3 * vk0x3; |
| |
| const int32_t vi1x0 = (int32_t) i1[0]; |
| const int32_t vi1x1 = (int32_t) i1[1]; |
| const int32_t vi1x2 = (int32_t) i1[2]; |
| const int32_t vi1x3 = (int32_t) i1[3]; |
| i1 += 4; |
| |
| const int32_t vk1x0 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[4]; |
| const int32_t vk1x1 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[5]; |
| const int32_t vk1x2 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[6]; |
| const int32_t vk1x3 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[7]; |
| |
| vacc0 += vi1x0 * vk1x0; |
| vacc1 += vi1x1 * vk1x1; |
| vacc2 += vi1x2 * vk1x2; |
| vacc3 += vi1x3 * vk1x3; |
| |
| const int32_t vi2x0 = (int32_t) i2[0]; |
| const int32_t vi2x1 = (int32_t) i2[1]; |
| const int32_t vi2x2 = (int32_t) i2[2]; |
| const int32_t vi2x3 = (int32_t) i2[3]; |
| i2 += 4; |
| |
| const int32_t vk2x0 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[8]; |
| const int32_t vk2x1 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[9]; |
| const int32_t vk2x2 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[10]; |
| const int32_t vk2x3 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[11]; |
| |
| vacc0 += vi2x0 * vk2x0; |
| vacc1 += vi2x1 * vk2x1; |
| vacc2 += vi2x2 * vk2x2; |
| vacc3 += vi2x3 * vk2x3; |
| |
| const int32_t vi3x0 = (int32_t) i3[0]; |
| const int32_t vi3x1 = (int32_t) i3[1]; |
| const int32_t vi3x2 = (int32_t) i3[2]; |
| const int32_t vi3x3 = (int32_t) i3[3]; |
| i3 += 4; |
| |
| const int32_t vk3x0 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[12]; |
| const int32_t vk3x1 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[13]; |
| const int32_t vk3x2 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[14]; |
| const int32_t vk3x3 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[15]; |
| |
| vacc0 += vi3x0 * vk3x0; |
| vacc1 += vi3x1 * vk3x1; |
| vacc2 += vi3x2 * vk3x2; |
| vacc3 += vi3x3 * vk3x3; |
| |
| const int32_t vi4x0 = (int32_t) i4[0]; |
| const int32_t vi4x1 = (int32_t) i4[1]; |
| const int32_t vi4x2 = (int32_t) i4[2]; |
| const int32_t vi4x3 = (int32_t) i4[3]; |
| i4 += 4; |
| |
| const int32_t vk4x0 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[16]; |
| const int32_t vk4x1 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[17]; |
| const int32_t vk4x2 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[18]; |
| const int32_t vk4x3 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[19]; |
| |
| vacc0 += vi4x0 * vk4x0; |
| vacc1 += vi4x1 * vk4x1; |
| vacc2 += vi4x2 * vk4x2; |
| vacc3 += vi4x3 * vk4x3; |
| |
| const int32_t vi5x0 = (int32_t) i5[0]; |
| const int32_t vi5x1 = (int32_t) i5[1]; |
| const int32_t vi5x2 = (int32_t) i5[2]; |
| const int32_t vi5x3 = (int32_t) i5[3]; |
| i5 += 4; |
| |
| const int32_t vk5x0 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[20]; |
| const int32_t vk5x1 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[21]; |
| const int32_t vk5x2 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[22]; |
| const int32_t vk5x3 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[23]; |
| |
| vacc0 += vi5x0 * vk5x0; |
| vacc1 += vi5x1 * vk5x1; |
| vacc2 += vi5x2 * vk5x2; |
| vacc3 += vi5x3 * vk5x3; |
| |
| const int32_t vi6x0 = (int32_t) i6[0]; |
| const int32_t vi6x1 = (int32_t) i6[1]; |
| const int32_t vi6x2 = (int32_t) i6[2]; |
| const int32_t vi6x3 = (int32_t) i6[3]; |
| i6 += 4; |
| |
| const int32_t vk6x0 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[24]; |
| const int32_t vk6x1 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[25]; |
| const int32_t vk6x2 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[26]; |
| const int32_t vk6x3 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[27]; |
| |
| vacc0 += vi6x0 * vk6x0; |
| vacc1 += vi6x1 * vk6x1; |
| vacc2 += vi6x2 * vk6x2; |
| vacc3 += vi6x3 * vk6x3; |
| |
| const int32_t vi7x0 = (int32_t) i7[0]; |
| const int32_t vi7x1 = (int32_t) i7[1]; |
| const int32_t vi7x2 = (int32_t) i7[2]; |
| const int32_t vi7x3 = (int32_t) i7[3]; |
| i7 += 4; |
| |
| const int32_t vk7x0 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[28]; |
| const int32_t vk7x1 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[29]; |
| const int32_t vk7x2 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[30]; |
| const int32_t vk7x3 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[31]; |
| |
| vacc0 += vi7x0 * vk7x0; |
| vacc1 += vi7x1 * vk7x1; |
| vacc2 += vi7x2 * vk7x2; |
| vacc3 += vi7x3 * vk7x3; |
| |
| const int32_t vi8x0 = (int32_t) i8[0]; |
| const int32_t vi8x1 = (int32_t) i8[1]; |
| const int32_t vi8x2 = (int32_t) i8[2]; |
| const int32_t vi8x3 = (int32_t) i8[3]; |
| i8 += 4; |
| |
| const int32_t vk8x0 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[32]; |
| const int32_t vk8x1 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[33]; |
| const int32_t vk8x2 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[34]; |
| const int32_t vk8x3 = (int32_t) ((const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)))[35]; |
| |
| vacc0 += vi8x0 * vk8x0; |
| vacc1 += vi8x1 * vk8x1; |
| vacc2 += vi8x2 * vk8x2; |
| vacc3 += vi8x3 * vk8x3; |
| |
| w = (const void*) ((uintptr_t) w + 4 * sizeof(int32_t) + 36 * sizeof(int8_t)); |
| |
| float vfpacc0 = (float) vacc0; |
| float vfpacc1 = (float) vacc1; |
| float vfpacc2 = (float) vacc2; |
| float vfpacc3 = (float) vacc3; |
| |
| const float vscale0 = ((const float*) w)[0]; |
| const float vscale1 = ((const float*) w)[1]; |
| const float vscale2 = ((const float*) w)[2]; |
| const float vscale3 = ((const float*) w)[3]; |
| w = (const void*) ((const float*) w + 4); |
| |
| vfpacc0 *= vscale0; |
| vfpacc1 *= vscale1; |
| vfpacc2 *= vscale2; |
| vfpacc3 *= vscale3; |
| |
| vfpacc0 = math_max_f32(vfpacc0, voutput_min_less_zero_point); |
| vfpacc1 = math_max_f32(vfpacc1, voutput_min_less_zero_point); |
| vfpacc2 = math_max_f32(vfpacc2, voutput_min_less_zero_point); |
| vfpacc3 = math_max_f32(vfpacc3, voutput_min_less_zero_point); |
| |
| vfpacc0 = math_min_f32(vfpacc0, voutput_max_less_zero_point); |
| vfpacc1 = math_min_f32(vfpacc1, voutput_max_less_zero_point); |
| vfpacc2 = math_min_f32(vfpacc2, voutput_max_less_zero_point); |
| vfpacc3 = math_min_f32(vfpacc3, voutput_max_less_zero_point); |
| |
| vfpacc0 += vmagic_bias; |
| vfpacc1 += vmagic_bias; |
| vfpacc2 += vmagic_bias; |
| vfpacc3 += vmagic_bias; |
| |
| int32_t vout0 = (int32_t) fp32_to_bits(vfpacc0) - vmagic_bias_less_output_zero_point; |
| int32_t vout1 = (int32_t) fp32_to_bits(vfpacc1) - vmagic_bias_less_output_zero_point; |
| int32_t vout2 = (int32_t) fp32_to_bits(vfpacc2) - vmagic_bias_less_output_zero_point; |
| int32_t vout3 = (int32_t) fp32_to_bits(vfpacc3) - vmagic_bias_less_output_zero_point; |
| |
| output[0] = (int8_t) vout0; |
| output[1] = (int8_t) vout1; |
| output[2] = (int8_t) vout2; |
| output[3] = (int8_t) vout3; |
| output += 4; |
| } |
| if XNN_UNLIKELY(c != 0) { |
| const int8_t* k = (const int8_t*) ((uintptr_t) w + 4 * sizeof(int32_t)); |
| do { |
| int32_t vacc = *((const int32_t*) w); |
| w = (const void*) ((uintptr_t) w + sizeof(int32_t)); |
| |
| const int32_t vi0 = (int32_t) *i0++; |
| const int32_t vk0 = (int32_t) k[0]; |
| vacc += vi0 * vk0; |
| const int32_t vi1 = (int32_t) *i1++; |
| const int32_t vk1 = (int32_t) k[4]; |
| vacc += vi1 * vk1; |
| const int32_t vi2 = (int32_t) *i2++; |
| const int32_t vk2 = (int32_t) k[8]; |
| vacc += vi2 * vk2; |
| const int32_t vi3 = (int32_t) *i3++; |
| const int32_t vk3 = (int32_t) k[12]; |
| vacc += vi3 * vk3; |
| const int32_t vi4 = (int32_t) *i4++; |
| const int32_t vk4 = (int32_t) k[16]; |
| vacc += vi4 * vk4; |
| const int32_t vi5 = (int32_t) *i5++; |
| const int32_t vk5 = (int32_t) k[20]; |
| vacc += vi5 * vk5; |
| const int32_t vi6 = (int32_t) *i6++; |
| const int32_t vk6 = (int32_t) k[24]; |
| vacc += vi6 * vk6; |
| const int32_t vi7 = (int32_t) *i7++; |
| const int32_t vk7 = (int32_t) k[28]; |
| vacc += vi7 * vk7; |
| const int32_t vi8 = (int32_t) *i8++; |
| const int32_t vk8 = (int32_t) k[32]; |
| vacc += vi8 * vk8; |
| k += 1; |
| |
| const float vscale = *((const float*) ((uintptr_t) w + 3 * sizeof(int32_t) + 36 * sizeof(int8_t))); |
| float vfpacc = (float) vacc * vscale; |
| vfpacc = math_max_f32(vfpacc, voutput_min_less_zero_point); |
| vfpacc = math_min_f32(vfpacc, voutput_max_less_zero_point); |
| vfpacc += vmagic_bias; |
| int32_t vout = (int32_t) fp32_to_bits(vfpacc) - vmagic_bias_less_output_zero_point; |
| |
| *output++ = (int8_t) vout; |
| } while (--c != 0); |
| } |
| |
| output = (int8_t*) ((uintptr_t) output + output_increment); |
| } while (--output_width != 0); |
| } |